Science and Technology of Advanced Materials ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tsta20 Magneto-ionic suppression of magnetic vortices Yu Chen, Aliona Nicolenco, Pau Molet, Agustin Mihi, Eva Pellicer & Jordi Sort To cite this article: Yu Chen, Aliona Nicolenco, Pau Molet, Agustin Mihi, Eva Pellicer & Jordi Sort (2021) Magneto-ionic suppression of magnetic vortices, Science and Technology of Advanced Materials, 22:1, 972-984, DOI: 10.1080/14686996.2021.1988830 To link to this article: https://doi.org/10.1080/14686996.2021.1988830 © 2021 The Author(s). Published by National Institute for Materials Science in partnership with Taylor & Francis Group. View supplementary material Published online: 04 Nov 2021. Submit your article to this journal Article views: 1767 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=tsta20 SCIENCE AND TECHNOLOGY OF ADVANCED MATERIALS 2021, VOL. 22, NO. 1, 972–984 https://doi.org/10.1080/14686996.2021.1988830 Magneto-ionic suppression of magnetic vortices a a b b a a,c Yu Chen , Aliona Nicolenco , Pau Molet , Agustin Mihi , Eva Pellicer and Jordi Sort a Departament de Física, Universitat Autònoma de Barcelona, Bellaterra, 08193, Spain; b Institute of Materials Science of Barcelona, Bellaterra, 08193, Spain; c Institució Catalana de Recerca i Estudis Avançats, Pg. Lluís Companys 23, Barcelona, 08010, Spain ABSTRACT ARTICLE HISTORY Magneto-ionics refers to the non-volatile control of the magnetic properties of materials by Received 19 July 2021 voltage-driven ion migration. This phenomenon constitutes one of the most important mag­ Revised 28 September 2021 netoelectric mechanisms and, so far, it has been employed to modify the magnetic easy axis of Accepted 29 September 2021 thin films, their coercivity or their net magnetization. Herein, a novel magneto-ionic effect is KEYWORDS demonstrated: the transition from vortex to coherent rotation states, caused by voltage- Magnetoelectric; magneto- induced ion motion, in arrays of patterned nanopillars. Electrolyte-gated Co/GdOx bilayered ionics; vortex; voltage nanopillars are chosen as a model system. Electron microscopy observations reveal that, upon control of magnetism voltage application, oxygen ions diffuse from GdOx to Co, resulting in the development of CLASSIFICATIONS paramagnetic oxide phases (CoOx) along sporadic diffusion channels. This breaks up the initial 40 Optical; magnetic and magnetization configuration of the ferromagnetic pillars (i.e. vortex states) and leads to the electronic device materials; formation of small ferromagnetic nanoclusters, embedded in the CoOx matrix, which behave as 102 Porous / Nanoporous / single-domain nanoparticles. As a result, a decrease in the net magnetic moment is observed, Nanostructured materials < together with a drastic change in the shape of the hysteresis loop. Micromagnetic simulations 100 Materials; 106 Metallic are used to interpret these findings. These results pave the way towards a new potential materials < 100 Materials; 203 application of magnetoelectricity: the magneto-ionic control of magnetic vortex states. Magnetics / Spintronics / Superconductors < 200 Applications; 212 Surface and interfaces < 200 Applications; 503 TEM, STEM, SEM < 500 Characterization 1. Introduction extend far beyond memory applications, and also Magnetic vortex is one of fundamental magnetization include oscillators for effective microwave generation states that occurs in micro-/nano sized ferromagnetic [6], energy storage [7], or biomedical applications structures, e.g. disks, ellipses or nanowires, and so on, (theragnostics) where the magnetic vortices are cap­ due to geometrical spin confinement [1–4]. This state able of damaging the membrane of cancer cells or is highly appealing for non-volatile magnetic mem­ promote drug delivery [8,9]. In turn, single-domain ories and spintronic devices since the magnetic infor­ magnetic nanoparticles have found applications in, mation can be stored by encoding both (i) the vortex e.g., high-density recording media, magnetic biosen­ chirality, i.e. the direction of in-plane magnetization sing, magnetic imaging, or hyperthermia therapy, rotation (clockwise or counter-clockwise), and (ii) the among others [10]. polarity, i.e. the out-of-plane magnetization of the Traditionally, the magnetic behavior of nanostruc­ nanoscale vortex core (up or down) [5]. Hence, each tures is controlled during sample preparation by vary­ magnetic vortex can store four bits of information. ing the aspect ratio of the motifs [1,11,12]. Namely, by Nonetheless, the applications of magnetic vortices increasing the thickness of a circular nanostructure CONTACT Aliona Nicolenco Aliona.Nicolenco@uab.cat; Jordi Sort Jordi.Sort@uab.cat Science faculty, Building Cc, Departament de Física, Universitat Autònoma, Bellaterra 08193, Spain Supplemental data for this article can be accessed here. © 2021 The Author(s). Published by National Institute for Materials Science in partnership with Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Sci. Technol. Adv. Mater. 22 (2021) 973 Y. CHEN et al. while keeping a fixed diameter, several ferromagnetic resistance states, also magneto-ionically or by strain states can be generated, such as in-plane single [39,40], and (iv) anisotropy easy axis magneto-ionic domain, vortex, flower, or out-of-plane single- switching, from in-plane to out-of-plane directions domain states, respectively. The occurrence of mag­ [41,42]. So far, magneto-ionic switching from vortex netic vortices, particularly in soft magnetic materials to single-domain states has not been demonstrated. (e.g. Ni, NiFe, CoFe, FeGa, Co, etc.), can be predicted ME manipulation of magnetic vortex properties has from well-established phase diagrams. Thus, arrays of been achieved using strain transfer from an under­ patterned dots with desired magnetic states can be lying ferroelectric material. Micromagnetic simula­ prepared on-demand according to these data tions first predicted the possibility to switch between [1,11,13–15]. However, the magnetic properties of the magnetic vortex and polar states in FeGa nanodots the as-prepared materials cannot be straightforwardly grown onto a ferroelectric PZT substrate [15]. Later modified after sample preparation. This limits the on, experimental works showed annihilation of mag­ versatility and potential functional applications of netic vortices in Ni disks via uniaxial compressive these nanostructures. To overcome this limitation, strain transferred from PMN-PT [43] or BaTiO3 significant research effort has been put forward over [44]. Electric-field-assisted switching of magnetic vor­ the years to tune the properties of magnetic vortices by tex chirality was also shown in Co/PMN-PT [45]. external means, such as spin-torque effects using spin- While these results are very appealing for voltage- polarized currents [16,17], laser pulse excitation control of vortex states, the complicated behavior of [18,19], mechanical actuation (e.g. unidirectional the ferroelectric domains of these substrates (includ­ compression) [20] or exchange coupling with an adja­ ing ferroelectric relaxor phenomena [46] and presence cent antiferromagnetic layer (after suitable heat treat­ of surface skin layers [47]) poses some challenges to ments and field cooling procedures) [2,21,22]. obtain uniform magnetization reversal modes in all Nonetheless, these methods are energetically costly disks comprised patterned arrays. This adds to other and sometimes only induce volatile changes in the important issues of strain-mediated ME systems, such vortex properties. as clamping effects, mechanical fatigue, high fabrica­ In parallel, during the last few years, a number of tion costs (in the case of PMN-PT) or need of rela­ investigations have demonstrated the possibility to tively large applied voltages (e.g. 400 V) [43]. tailor the magnetic properties of materials at the ‘post- Among all converse ME effects, magneto-ionics is synthesis’ stage using electric fields through the so- gaining attention because voltage-driven ion transport called magnetoelectric (ME) effect. This is very appeal­ between the magnetic material of interest and an ion ing because the use of voltage (instead of electrical source/sink allows for a stringent non-volatile control currents) minimizes Joule heating effects and it can of interfacial magnetism to an unprecedented extent thus be considered as a highly energy-efficient actua­ [25,48,49]. Upon electrolyte gating, the ions may form tion protocol [23]. Voltage has been shown to induce diffusion channels or even uniform migration fronts changes in the coercivity [24], net magnetization [25], (eventually developing new interfaces within the actu­ remanence [26], exchange bias [27], Curie tempera­ ated films), resulting in unique magnetic characteris­ ture [28] and magnetoresistance [29,30], of the inves­ tics (in e.g. oxides or nitrides) [25,48]. Magneto-ionic tigated materials. Several ME mechanisms have been effects can involve the transport of various types of ion identified: intrinsic ME coupling in single-phase mul­ species, such as alkali metal cations (Li+, Na+ and K+) tiferroics, electric surface charging in ultra-thin films, [50–52], H+ [49,53], F− [54], N3- [48], and O2-. The magneto-ionics (voltage-driven ion migration), and latter remains the most studied up to date. Interesting strain-mediated effects in piezoelectric/magnetostric­ effects have been reported both in transition-metal tive composites [31–33]. ME approaches have rapidly oxide single layers, where the oxygen ions are already gained significant interest in the design of ultra low- present in the original material structure [25,55,56], heat-dissipating memories, logic devices, neuro­ and in ME heterostructures where the oxygen ions are morphic computing or lab-on-chip platforms, among supplied to the magnetic material of interest from an others. adjacent high ionic mobility ion reservoirs, such as Up to date, ME effects have allowed experimental Gd2O3 [34,49,57–60], HfO2 [61–63], or Al2O3 realization of several types of ‘magnetic switches’: (i) [41,64,65] layers. OFF/ON switching of magnetization, when the start­ In this work, we investigate a new room- ing material is non-magnetic but a magnetic phase is temperature magneto-ionic effect that is induced at generated upon electric field application (and vice the nanoscale: the transition from ‘magnetic vortex versa), via magneto-ionics, strain, or surface charging state’ to ‘coherent rotation’ in electrolyte-gated mechanisms [25,34–36]; (ii) switching between super­ Co/GdOx bilayered nanopillars. Upon voltage applica­ paramagnetic and single-domain states in nanocrys­ tion, oxygen diffuses from the GdOx solid ionic con­ tals, either magneto-ionically or via strain [37,38]; (iii) ductor towards metallic Co, forming a non-magnetic magneto-resistive switching between low and high CoOx matrix with embedded Co nanoclusters which Sci. Technol. Adv. Mater. 22 (2021) 974 Y. CHEN et al. enable vortex suppression. The experimental results Electrodeposition was done from an electrolytic bath are interpreted with the aid of micromagnetic simula­ with the following composition: 0.2 M CoSO4∙7H2O, tions to model magnetization behavior in the pristine 0.65 M H3BO3, 0.25 M Na3C6H5O7 and 0.04 M C6H8 pillars (at 0 V) and after voltage actuation. The chal­ O7 [69]. The pH was left at its unadjusted value (4.6). lenges and future research directions for further opti­ Co nanopillars were grown potentiostatically at mization and control of this newly observed magneto- −1.00 V under stagnant conditions at room tempera­ ionic effect are described. ture. From the representative current–time curve shown in Supporting Figure S1 it can be concluded that Co grows inside the template for ~100 s and then 2. Experimental section overgrows, tending to form a continuous layer for Cobalt nanopillars were grown by electrodeposition of longer deposition times. A preliminary magnetic char­ the metal on Au (80 nm)/Ti (20 nm)/Si substrates acterization study revealed that magnetic vortex beha­ coated with a patterned resist film (Figure 1(a)) featur­ vior (i.e. constricted hysteresis loops) was observed in ing a square array of cylindrical holes of 200 nm in Co pillars produced after depositing for 20–60 seconds, diameter, 400 nm pitch and 280 nm in height. The which corresponds to 60–120 nm thickness. This is in patterns were obtained by nanoimprinting 280– agreement with the vortex vs. coherent rotation 300 nm thick film of SU8 2000.5 (Microchem, experimental phase diagram for Co disks/pillars Germany) photoresist with pre-patterned polydi­ reported in the literature, which predicts vortex state methylsiloxane (PDMS) molds [66]. This technique for this range of thicknesses-to-diameter aspect ratios has been proven capable of producing extended [1,11]. For the magneto-ionic actuation, nanopillars nanostructured areas with high fidelity, rapidity, with 80 nm thickness were selected. Note that nanos­ cleanroom-free and at a low cost [67,68]. After the tructures with lower thickness and flat surface are pattern is imprinted on the resist, the Au seed-layer on difficult to produce by electrodeposition, due to the the substrate was exposed on the patterned areas by complex nucleation and growth mechanisms which performing reactive ion etching using O2 plasma. result in a rough topography and local thickness var­ The resulting pattern was used as a template for iation. Next, a 40 nm ± 5 nm layer of GdOx was electrodeposition of Co pillars. Electrodeposition was deposited on top of the template partially filled with performed in a three-electrode single compartment Co using magnetron sputtering (Figure 1(a)) at room cell connected to a PGSTAT302N Autolab potentio­ temperature with a base pressure around 10−8 Torr. stat/galvanostat (Metrohm-Autolab, The Netherlands) The flows of high-purity Ar and O2 were 17 sccm and . A double junction Ag/AgCl reference electrode 10 sccm, respectively, with a total pressure of (Metrohm AG, Switzerland) was used with 3 M KCl 3 × 10−3 mbar. The distance between the target and inner solution and 1 M Na2SO4 outer solution, and the substrate was around 8 cm, and the deposition rate a Pt wire served as a counter electrode. was about 0.36 Å s−1. The photoresist template was not Figure 1. (a) Schematics of the synthesis procedure to grow bilayered nanopillars, wherein Co is first electrodeposited on a nanoimprinted substrate and GdOx is subsequently sputtered on top (note that for sake of clarity the GdOx that grew on top of the SU-8 resist is not shown). (b) Representative SEM image of an array of electrodeposited Co nanopillars embedded in SU-8 template (before sputtering GdOx). (c) Reconstructed 3D profile of an individual as-deposited Co nanopillar (without GdOx) obtained after processing the SEM micrograph with ImageJ software. The inset in panel (c) shows a FE-SEM image of a single as- grown Co pillar. Sci. Technol. Adv. Mater. 22 (2021) 975 Y. CHEN et al. removed at any stage of sample preparation in order lower than for bulk Co, as it has been reported for not to distort the Co/GdOx bilayer structure of the electrodeposited Co due to the presence of impurities) pillars (i.e. to maintain one horizontal interface). [1,74]. The cell size was set to 5 nm. To reproduce the Surface morphology and composition of the samples nanopillars with upper rough surface, the geometry were characterized on a Zeiss Merlin (Zeiss, Germany) was defined as a cylinder with disordered arrays of field emission scanning electron microscope (FE-SEM) little cones with different sizes and pitch. The energy equipped with an energy-dispersive X-ray (EDX) detec­ minimization evolver was adopted to obtain the mag­ tor, and the ‘Image J’ software [70] was used to recon­ netization configuration at different values of applied struct the topography of the as-deposited nanopillars. magnetic field. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and electron 3. Results and discussion energy loss spectroscopy (EELS) were performed using a JEOL JEM2010F microscope (Jeol, Japan) equipped Magnetic nanopillars comprising electrodeposited Co with a field emission gun, operated at 200 kV and and sputtered GdOx layers were prepared onto metal­ coupled to a Gatan GIF spectrometer. lized Si substrates covered with a nanoimprinted A MicroSense (LOT-QuantumDesign, United resist, as schematically illustrated in Figure 1(a). States) Vibrating Sample Magnetometer (VSM) was Figure 1(b) shows a representative top-view SEM used to perform magnetoelectric measurements. The micrograph of the as-deposited Co pillars inside the voltage was applied in-situ using a home-made two- template. The diameter of the pillars is given by the electrode electrochemical cell connected to the Agilent size of the pores and is around 200 nm. The pillars B2902A (Agilent Technologies, United States) power exhibit some roughness (Figure 1(c)) provided by the supply and attached to a VSM sample holder, as polycrystalline grain growth during electrodeposition. described in detail in our previous works [71]. The It is worth mentioning that the polymeric template sample with the Co/GdOx nanopillars of approxi­ was not removed at any stage of the experiment. The mately 0.25 cm2 size acted as a working electrode SU-8 polymer (epoxy-based, negative tone resist) is and the contact was made to the underlying Au seed commonly used in various micro-/nanoelectromecha­ layer in a capacitor-like configuration [72]. A Pt wire nical systems (MEMS/NEMS) due to its outstanding was used as counter electrode. The electrodes were chemical and thermal resistivity, optical transparency placed in a small Eppendorf tube (1 ml) filled with and easy fabrication of high-aspect-ratio features. In an anhydrous electrolyte. The electrolyte solution was addition, it is an electrical insulator (breakdown field propylene carbonate treated with metallic sodium to ~108 V m−1). Thanks to such properties, SU-8 is an remove any traces of water. The sodium treatment left ideal material to be fully integrated in electronics, a small amount of Na+ and OH− ions, and no other functional MEMS/NEMS devices and lab-on-chip salts were added. The use of a liquid electrolyte is microsystems [75]. Since the Co pillars do not com­ advantageous because it allows the generation of ultra- pletely fill the pores of the template, sputtering GdOx high electric fields (of the order of hundreds of on top of such structures renders an array of bilayered MV cm−1), without electric pinholes or defects in nanopillars with a Co/GdOx interface roughly parallel a thin layer of GdOx. All magnetic measurements to the substrate. were taken at room temperature in an in-plane con­ The magnetic properties of the nanostructures were figuration, while the voltage was applied out-of-plane. measured at room temperature using VSM. Figure 2 Prior to measuring each hysteresis loop a specific (a) shows the evolution of the M-H hysteresis loops of negative voltage was applied for 30 min following the Co/GdOx nanopillars as a function of applied this sequence: 0–60 V with a step of 10 V, and 80– positive voltage. In the initial state at 0 V, the shape 280 V with a step of 20 V. In all cases, voltage was of the loop indicates magnetization reversal via for­ maintained during the measurement (each hysteresis mation of a vortex state, with its characteristic central loop takes 30 min to be completed), which resulted in constriction, negligible coercivity, small remanent a total voltage application time of 1 h at each specific magnetization and clear open lobes at high magnetic condition. A waiting time of 15 min was allowed fields. Applying positive voltage to the Co/GdOx het­ between each voltage switch. erostructures significantly alters the shape of the hys­ Micromagnetic modeling was performed using the teresis loop, as well as the magnetic moment at micromagnetic simulator Object Oriented saturation (Figure 2(b)). The magnetic moment MicroMagnetic Framework (OOMMF) [73]. The fol­ decreases gradually up to 100 V and then remains lowing parameters were used for hexagonal close- rather constant even when the voltage is further packed (hcp)-Co: exchange constant A = 30 pJ m−1, increased. The decrease of magnetization is in accor­ anisotropy K1 = 430 kJ m−3 chosen to be along random dance with magneto-ionic phenomena. Upon applica­ directions among cells (polycrystalline Co) and satura­ tion of positive voltage, the negatively charged oxygen tion magnetization Ms = 1100 emu cm−3 (somewhat ions migrate from GdOx towards the underlying Co Sci. Technol. Adv. Mater. 22 (2021) 976 Y. CHEN et al. Figure 2. Magnetoelectric actuation of Co/GdOx nanopillars. (a) Representative room-temperature hysteresis loops for gradually increasing values of positive voltage from 0 to 280 V. The loops are normalized to 1 to facilitate comparison of their shapes. Dependence of the (b) change in magnetic moment with respect to 0 V, (c) squareness ratio and (d) normalized area of the loops as a function of the applied voltage where 1 is the area of the loop at 0 V. layers, driven by the electric field which is induced (c)), and the shape of the hysteresis loop (i.e. the perpendicularly to the sample surface. This triggers normalized integrated area in Figure 2(d)) continue two redox reactions that occur simultaneously: (i) to evolve even for higher applied positive voltages. partial Co oxidation by O2- to produce CoOx; and Therefore, although the magnetic moment remains (ii) partial GdOx reduction to metallic Gd. Since rather constant for ∆V > 100 V, the system does not both, Gd and CoOx phases (both CoO and Co3O4), seem to reach a steady state yet. At ∆V = 120 V, the are paramagnetic at room temperature, these redox hysteresis loop still shows small open lobes character­ reactions are responsible for the observed decrease in istic for a magnetic vortex state. Remarkably, neither the magnetic moment. the nucleation field nor the annihilation field of the The plateau in Figure 2(b) reached at ~100 V sug­ vortex varies considerably as a function of applied gests that, most probably, already at this voltage, most voltage. However, the irreversible hysteresis loss (i.e. of the oxygen ions have been transported towards Co. the area enclosed by the hysteresis loop) progressively Hence, further paramagnetic phase fraction does not decreases up to ∆V = 200 V. Above this voltage, the form. However, the squareness ratio, Mr/Ms (Figure 2 central constriction and the open lobes of the loops are Sci. Technol. Adv. Mater. 22 (2021) 977 Y. CHEN et al. no longer observed, suggesting that the vortex state is (which correspond to arrays of hundreds or thousands suppressed. At such high voltages, the loop shape of Co pillars), i.e. more rounded and sheared hyster­ resembles that of polycrystalline cobalt. Since migra­ esis loops. tion of O2- from GdOx is not expected above Figure 3 illustrates the evolution of the simulated ∆V = 100 V, other processes should be responsible spin configurations corresponding to different for the observed loss of the vortex stability. Indeed, values of applied magnetic field for rough Co nano­ recent magneto-ionic studies in CoOx single layers, pillars with the geometry depicted in Supporting involving positron annihilation experiments, have Figure S2(f) (i.e. magnetization reversal occurring revealed a complicated voltage-driven ion transport through vortex spin structure). When the magnetic involving not only O2- ion migration but also Co2+ field is decreased from saturation (1), the magnetic cations redistribution while preserving the total net vortex nucleates at around 1000 Oe (2–3). This is magnetic moment [25,48]. accompanied by an abrupt decrease in the average Micromagnetic simulations were performed in dot magnetization. As the magnetic field is further order to get a better understanding of the magnetiza­ decreased the vortex core progressively displaces tion reversal mechanisms in the voltage-actuated towards the center of the disk and at zero magnetic arrays of Co/GdOx nanopillars, both at 0 V and field all the spins follow the circular geometry of the after application of positive voltage. Supporting disk (4). When the magnetic field reaches the anni­ Figure S2(a) compares the experimental loop at 0 V hilation field (5), the vortex vanishes completely (6). with the simulated hysteresis loop of an individual This occurs at around 2020 Oe. This process stabi­ ‘perfect’ disk-shaped Co pillar with 200 nm width lizes the single-domain state with negative and 80 nm thickness. The nucleation and annihila­ orientation. tion fields obtained from the simulations are in rather As experimentally demonstrated and shown in good agreement with the experimental data. Figure 2, voltage actuation significantly changes the However, the experimental loop is significantly magnetic response of the system. Assuming the vol­ more tilted than the theoretical prediction. To tage-driven oxygen transport towards the substrate, improve the agreement between experimental and a large fraction of metallic Co adjacent to GdOx should computed loops, the effect of surface roughness was be converted into paramagnetic CoOx, thus resulting also modeled (by incorporating surface imperfec­ in a decrease in the effective thickness of ferromag­ tions, such as holes or hills). The results, presented netic Co counterpart. Micromagnetic simulations in Supporting Figure S2(b,c) indicate that, indeed, have been employed to model the shape and magne­ the magnetic behavior of the Co nanopillars is influ­ tization reversal behavior of the Co nanopillars after enced, to some extent, by the surface topological positive (e.g. ∆V = 200 V) voltage actuation. Assuming characteristics. Since electrodeposited Co pillars fea­ that the oxygen moving front is horizontal (parallel to ture noticeable surface inhomogeneities (Figure 1), the substrate), one may expect to form a flat CoOx/Co incorporation of a certain degree of random rough­ interface, where the Co counterpart may retain the ness in the simulations, resembling the topography of initial disk shape and CoOx would be on top. electrodeposited Co, is justified. As shown in Figure 4(a) shows a simulated hysteresis loop of an Supporting Figure S2(d), the occurrence of surface individual ‘perfect’ Co disk with a thickness of 25 nm roughness induces variations in the values of nuclea­ (which roughly corresponds to 1/3 of the initial thick­ tion and annihilation fields (i.e. a distribution of the ness of the Co pillar, in agreement with the observed switching fields among pillars). This explains why the decrease in magnetic moment (Figure 2(b))). nucleation and annihilation events in the experimen­ According to the modeling, this disk-shaped pillar tal loops are not as sharp as in the simulations. still results in magnetization reversal via vortex spin Next, arrays of four cylindrical disks having vari­ arrangement, with clear open lobes at high fields and able thickness (Supporting Figure S2(e)) and rough­ zero net magnetization at remanence. This is actually ness (Figure S2(f)) were also simulated. This allows to in agreement with the calculated ground state phase better assess the effects of averaging local variations in diagram for polycrystalline hcp-Co sub-micrometer the nucleation/annihilation fields among dots, as well dots, wherein the vortex-to-single domain transition as possible effects from inter-dot magnetic dipolar occurs at above 20 nm thickness [11]. However, the interactions. In all cases, for the selected geometries experimental loop shows a very different behavior, (thickness range between 60 and 90 nm and roughness with no constriction in its central part, indicating inhomogeneity with peak-to-valley distances up to that reversal is not by vortex-state formation. To see 30 nm), vortex-like hysteresis loops were obtained. whether a transition from vortex to in-plane single- Interestingly, adding these local inhomogeneities domain state occurred during the magneto-ionic among pillars allows obtaining a better agreement experiments, we then simulated the magnetization between the simulations and the experimental loops reversal of a thinner (i.e. 15 nm) Co pillar (Figure 4 Sci. Technol. Adv. Mater. 22 (2021) 978 Y. CHEN et al. in-plane magnetic anisotropy). Obviously, none of the two simulated loops agree with the experimentally obtained data. Several magneto-ionic studies indicated that oxygen- ion transport is a complex process and occurs via, at least, two-fold mechanism: (i) homogeneous O2- migra­ tion toward the electrode with the opposite charge (due to Coulombic attraction), and (ii) localized oxygen ion migration along fast diffusion channels [25,48]. Grain boundaries promote the development of sporadic diffu­ sion paths for the oxygen ions, rather than horizontal migration fronts, thus leading to development of multi­ ple CoOx/Co interfaces, as it is schematically illustrated in the inset of Figure 4(c). Remarkably, micromagnetic simulations using such pseudo-random arrangement of irregularly shaped Co nanoparticles embedded in the paramagnetic CoOx matrix, each of them having its own magnetic easy axis, led to computed hysteresis loops which capture well the experimental one. More specifically, the cluster array was defined by two sets of particles with different periods, shifts and sizes to imi­ tate a kind of random shapes (Figure 5). Due to their small lateral dimensions (10–30 nm), magnetization reversal in each of these Co nanoclusters occurs by coherent rotation, as revealed by the spin configuration modeling at different magnetic fields. Note that some of the Co clusters could easily have even lower sizes and could therefore be superparamagnetic at room temperature. To shed light on the nanostructural changes that take place under voltage actuation, the composite nanopillars were inspected using STEM and EELS techniques. Cross-section lamellae from the pristine and voltage- treated nanopillars were prepared. Figure 6(a) shows the STEM image of a non-treated bilayered nanopillar, at interface region between Co and GdOx. Figure 6(b) shows a similar zoom after treating the pillar at +50 V for 1 hour. To locally quantify the oxygen distribution, EELS Figure 3. Micromagnetic modelling of the magnetic behavior mappings near the Co/GdOx interface were conducted. of the as-prepared nanopillars. The hysteresis loop was mod­ The corresponding inspected areas are highlighted with elled using an array of four rough Co nanopillars with geome­ white rectangles on panels (a) and (b). EELS mappings of try mimicking the experimentally observed one. Characteristic Co, Gd and O are shown in Supporting Figure S3. spin configurations, at different values of applied magnetic Figure 6(c) shows the oxygen EELS mapping obtained field. [(1)–(6)] correspond to the magnetic structure of an individual rough Co nanopillar from the array. For each con­ from the pristine nanopillar (0 V, on the left) and in the figuration the horizontal cross-section through the center (at voltage-treated sample (+50 V, on the right). At 0 V half the height) and the vertical cross-section through the (non-treated sample), clear red (oxygen) and black (no diameter of the pillar are shown in a circle and the rectangle oxygen) regions are observed, corresponding to GdOx below, respectively. The diameter of the pillars is 200 nm and and oxygen-free (metallic) cobalt. In contrast to the as- the thickness is 60 nm with 20 nm added roughness. The black grown sample, the EELS mapping of a similar region in arrows represent the in-plane orientations of the magnetic moments. In the cross-section configurations, the red color is the voltage-treated sample suggests oxygen redistribu­ defined by the magnetization pointing to the right side, and tion across the whole inspected area. These results con­ the cyan color is for magnetization pointing left. firm the magneto-ionic mechanism in voltage actuated Co/GdOx nanopillars, i.e. under the action of electric (b)). In this case, the hysteresis loop becomes squarer field the O2- ions are driven from GdOx towards Co, and it opens in the central part, resulting in higher with oxygen content ~20 at.% in the region below the coercive field (i.e. tending to single-domain state with Co/GdOx interface. Sci. Technol. Adv. Mater. 22 (2021) 979 Y. CHEN et al. Figure 4. Simulated hysteresis loops of an individual Co nanopillar after positive voltage application (e.g. ∆V = 200 V) considering the formation of (a) a flat Co disk of 25 nm thickness, (b) a flat Co disk of 15 nm thickness, and (c) array of Co clusters with random anisotropy easy axis embedded in a CoOx matrix. The insets of the panels a, b and c show schematic drawings where grey color represents non-magnetic CoOx and red color is for the simulated Co entities. The Co/GdOx nanopillars treated at +50 V for hysteresis loop implies a ferromagnetic behavior with 1 hour were further investigated by STEM. Figure 7 a non-zero coercivity (Figure 2(a)). Hence, most likely, (a) shows the selected area electron diffraction (SAED) there is a distribution of lateral sizes in the Co clusters, pattern of a region belonging to the Co layer, at about probably being larger towards the substrate, i.e. in 30 nm from the interface with GdOx. The diffraction regions farther away from the oxygen source. In any pattern corresponding to this area indicates the pre­ case, note that the HR-TEM analysis does not easily sence of several phases, including hcp-Co and GdOx allow to evidence the presence of large clusters of (as in the pristine sample, Figure S4(a,b)), and also either Co or CoOx because at lower magnification Co3O4 as well as metallic Gd, in agreement with the the interplanar distances cannot be resolved. EELS mapping. This implies that indeed oxygen mag­ Large voltages of opposite polarity were applied to neto-ionics results in a partial reduction of GdOx and the sample in order to try to recover the original oxidation of metallic Co to an oxide or a mixture of vortex state. The experimental results demonstrate oxides (some of them may be non-stoichiometric that, unfortunately, the observed effect appears to be especially when larger voltage is applied) [25]. irreversible. No changes in magnetic moment or the An even smaller region, indicated by the white shape of the loop could be observed upon voltage square in Figure 7(a), was inspected in more detail. gating at ∆V = −200 V, or even larger negative vol­ Fast Fourier Transform (FFT) analysis of this area tages. Poor reversibly and cyclability are among the shows spots with interplanar distances that corre­ main disadvantages of the magneto-ionic mechanism spond to metallic Co and Co3O4 nanoscale phases [23]. Generally, the reason for this is the severe vol­ (Figure 7(b,c)). Figure 7(d) reveals that this region is tage-induced phase transformations in pristine mate­ actually a zoom of an interface area between the metal rials caused by oxidation/reduction reactions that lead and the oxide phases. In order to separate the contri­ to the development of new phases with distinct struc­ butions from each of these two phases, either Co or tural properties. It has been reported that Co reaction CoOx reflections were masked via inverse Fourier with oxygen proceeds along the forward direction, transform image processing. The area highlighted in while the reversible reaction is much less energetically red color in Figure 7(e) corresponds to a region with favorable [57]. Furthermore, the propylene carbonate well-distinguishable planes with an average interpla­ electrolyte used in this work as a dielectric material nar distance of ~0.19 nm, which can be indexed to tends to polymerize at positive voltages (i.e. at the hcp-Co(101). Likewise, Figure 7(f) shows a region of same polarity required to drive the oxygen ions from fcc-Co3O4 (311) with a larger interatomic distance of GdOx towards Co) [77]. Experimentally, it had been ~0.24 nm, shadowed in yellow color. This data con­ observed that polymerization starts already at +14 V firms that, for positive applied voltages, oxygen [56], but the system investigated here requires much migrates from GdOx and leaves behind irregular larger positive voltages to switch between the magnetic metallic Co nano-clusters of different sizes. Note that states. Unfortunately, propylene carbonate cannot be the size of the individual Co clusters in Figure S4(c) is depolymerized with voltage of opposite polarity. Thus, below 5 nm. In this particular case, such Co nanoclus­ the organic products accumulated at the surface of the ters would be superparamagnetic, since the critical magnetic material significantly reduce the effective cluster size that renders superparamagnetic behavior electric field and therefore hamper magneto-ionic at room temperature for cobalt is below 13 nm [76]. cycling. Another factor that should also be considered However, the shape of the experimentally obtained is the ionic conductivity of the oxygen source. For Sci. Technol. Adv. Mater. 22 (2021) 980 Y. CHEN et al. Figure 6. Structural and compositional characterization, by HAADF-STEM and EELS, respectively. (a) HAADF of a Co/GdOx nanopillar prior to voltage actuation, i.e. at 0 V, (b) HAADF of the interface between Co and GdOx after positive voltage bias at ∆V = +50 V for 1 hour. The scale bar is 40 nm. The white rectangles in panels (a) and (b) correspond to the area inspected by EELS. (c) Characteristic oxygen EELS mappings: left was obtained from (a) and it corresponds to 0 V, whereas (c) right was obtained from panel (b) after applying +50 V. White arrow shows the direction of oxygen ions diffu­ sion towards the substrate. been reported that, under certain conditions, Gd2O3 Figure 5. Micromagnetic modelling of characteristic spin con­ is more prone to transmit oxygen rather than sink it figurations, at different values of applied magnetic field, of an [57]. Under positive voltages, AlOx may exhibit better individual rough Co nanopillar after positive voltage bias. [(1)– oxygen donor capabilities but it tends to dissolve/ (6)] Correspond to the magnetic structure of irregularly detach for voltages larger than 50 V. In contrast, shaped Co nanoclusters embedded in paramagnetic CoOx matrix (shown as grey background). The black arrows repre­ HfO2 is more prone to receive oxygen than to donate sent the orientations of the magnetic moments. The horizontal it [65], so it does not work as well as Gd2O3 to induce cross-section at the center and the vertical cross-section oxidation of adjacent metals. through the diameter of the pillar are shown as a circle and Finally, it is also worth mentioning that, although a rectangle, respectively. The diameter of the disks is 200 nm switching between vortex and coherent rotation states and those of the clusters in the range of 10–30 nm. The red color is defined by the magnetization pointing to the right in Co nanopillars by voltage-driven oxygen migration side, the cyan color is for magnetization pointing left. is experimentally feasible, the complicated oxygen-ion transport and high voltages required make the system overall difficult to control. However, recent studies on voltage-actuated CoN thin films have demonstrated oxygen magneto-ionics, the most commonly utilized the occurrence of uniform horizontal ion migration solid electrolytes are Gd2O3 [34,49,57–59], HfO2 fronts in this system, without occurrence of grain- [61,62], or Al2O3 [64]. All these materials are charac­ boundary diffusion channels during voltage bias [48]. terized by high ionic conductivity. However, it has In addition, nitrogen magneto-ionics requires lower Sci. Technol. Adv. Mater. 22 (2021) 981 Y. CHEN et al. Figure 7. Structural characterization by STEM of the sample treated with positive voltage (+50 V for 1 hour): (a) STEM image of an area located at the middle region of the Co pillar; (b) fast Fourier transform of the small region indicated by the white square in the panel (a); (c) average interplanar distances corresponding to the identified hcp-Co(101) and fcc-Co3O4 (311) phases obtained from the panels (e) and (f); (d) magnified STEM image of the sample marked with a white square in the panel (a); (e) hcp-Co planes imaged by masking the contribution from the Co3O4 spots in the FFT; and (f) Co3O4 planes images by masking the hcp-Co reflections. threshold voltages and has better cyclability than oxy­ requires relatively high voltages. Nevertheless, the gen magneto-ionics. Thus, the way towards reversible concept can be potentially extended to room- control of magnetic vortices will most likely lie beyond temperature nitrogen magneto-ionics (e.g. in CoN) oxygen magneto-ionics. or other magneto-ionic systems with a better cyclability and more controllable ion-migration mechanism, thereby allowing for the development 4. Conclusions of a unique energy-efficient actuation protocol to control the magnetization reversal mechanism of Our work demonstrates a new magneto-ionic effect: patterned structures after sample growth, just by a proof of concept of voltage-driven magneto-ionic applying suitable voltage protocols. From switching from vortex to single-domain magnetiza­ a technological viewpoint, voltage-assisted transitions tion reversal states in heterostructured Co/GdOx from vortex to single-domain states are interesting nanopillars. Large areas can be fabricated at ambient for potential application in stochastic computing (i.e. conditions and at low cost by combining electrode­ if the direction of the magnetization in the Co clus­ position with nanoimprint lithography. A positive ters is random), multi-state memories or neuro­ voltage drives the O2- ions from the GdOx towards morphic computing (where cumulative effects upon the substrate thus resulting in partial oxidation of successive voltage application are needed), provided metallic Co to paramagnetic CoOx. Due to the com­ the good reversibility and durability of the system. plicated oxygen-ion transport mechanism, which most likely involves stochastic diffusion channels associated with grain boundaries, oxygen magneto- ionics promotes the formation of irregularly shaped Acknowledgments Co clusters randomly distributed within the CoOx This work has received funding from the European Union’s matrix. Micromagnetic modelling of the system in Horizon 2020 research and innovation programme under the initial and voltage-actuated states reveals that such Marie Skłodowska-Curie grant agreement Nº 892661– MAGNUS. Financial support by the European Research clusters behave as single-domain particles whose Council (MAGIC-SWITCH 2019-Proof of Concept Grant spins rotate coherently with the applied magnetic (Nº 875018)), the Spanish Ministerio de Ciencia e Innovación field. Under the experimental conditions of this through grants, MAT2017-86357-C3-1-R, PID2019- work, the induced switching is irreversible and 106860GB-I00, PID2020-116844RB-C21 and The Spanish Sci. Technol. Adv. Mater. 22 (2021) 982 Y. CHEN et al. Severo Ochoa Centre of Excellence program FUNFUTURE [8] Kim DH, Rozhkova EA, Ulasov IV, et al. (CEX2019-000917-S as well as associated FEDER. The Biofunctionalized magnetic-vortex microdiscs for Generalitat de Catalunya (2017-SGR-292 and 2018-LLAV targeted cancer-cell destruction. Nat Mater. -00032) is also acknowledged. P.M acknowledges financial 2010;9:165–171. support from an FPI contract (2017) of the MICINN (Spain) [9] Kim DH, Karavayev P, Rozhkova EA, et al. cofounded by the ESF and the UAB. Mechanoresponsive system based on sub-micron chitosan-functionalized ferromagnetic disks. J Mater Chem. 2011;21:8422–8426. [10] Wu K, Su D, Liu J, et al. Magnetic nanoparticles in Disclosure statement nanomedicine: a review of recent advances. No potential conflict of interest was reported by the Nanotechnology. 2019;30:502003. author(s). [11] Prejbeanu IL, Natali M, Buda LD, et al. In-plane reversal mechanisms in circular Co dots. J Appl Phys. 2002;91:7343–7345. [12] Agramunt-Puig S, Del-Valle N, Navau C, et al. Funding Controlling vortex chirality and polarity by geome­ try in magnetic nanodots. Appl Phys Lett. This work was supported by the Generalitat de Catalunya 2014;104:12407. [2017-SGR-292,2018-LLAV-00032]; European Research [13] Goiriena-Goikoetxea M, Guslienko KY, Rouco M, Council [875018]; Marie Sklodowska-Curie [892661]; and et al. Magnetization reversal in circular vortex dots Spanish Ministry of Science and Innovation [CEX2019- of small radius. Nanoscale. 2017;9:11269–11278. 000917-S,MAT2017-86357-C3-1-R,PID2019-106860GB- [14] Liu XL, Yang Y, Wu JP, et al. Novel magnetic vortex I00, PID2020-116844RB-C21]. nanorings/nanodiscs: synthesis and theranostic applications. Chin Phys B. 2015;24:127505. [15] Sheng Q, Liu XL, Chen WJ, et al. Phase diagrams of ORCID magnetic state transformations in multiferroic com­ posites controlled by size, shape and interfacial cou­ Yu Chen http://orcid.org/0000-0003-1651-3698 pling strain. AIP Adv. 2017;7:105221. Aliona Nicolenco http://orcid.org/0000-0003-4624-2163 [16] Yamada K, Kasai S, Nakatani Y, et al. Electrical Pau Molet http://orcid.org/0000-0003-3693-0309 switching of the vortex core in a magnetic disk. Nat Agustin Mihi http://orcid.org/0000-0003-3821-7881 Mater. 2007;6:270–273. Eva Pellicer http://orcid.org/0000-0002-8901-0998 [17] Liu Y, Gliga S, Hertel R, et al. Current-induced mag­ Jordi Sort http://orcid.org/0000-0003-1213-3639 netic vortex core switching in a permalloy nanodisk. Appl Phys Lett. 2007;91:112501. [18] Fu X, Pollard SD, Chen B, et al. Optical manipulation Data availability of magnetic vortex visualized in situ by 4D electron microscopy. Sci Adv. 2018;4:eaat3077. The raw/processed data required to reproduce these findings [19] Le Guyader L, El Moussaoui S, Buzzi M, et al. cannot be shared at this time as the data also form part of an Demonstration of laser induced magnetization rever­ ongoing study. The data that support the findings of this study sal in GdFeCo nanostructures. Appl Phys Lett. are available from the corresponding authors on request. 2012;101:22410. [20] Wang J. Mechanical control of magnetic order: from phase transition to skyrmions. Ann Rev Mater Res. References 2019;49:361–388. [21] Sort J, Hoffmann A, Chung S-H, et al. Magnetization [1] Ross CA, Hwang M, Shima M, et al. Micromagnetic reversal in submicron disks: exchange biased vortices. behavior of electrodeposited cylinder arrays. Phys Rev Phys Rev Lett. 2005;95:67201. B. 2020;65:144417. [22] Gilbert DA, Ye L, Varea A, et al. A new reversal mode [2] Sort J, Buchanan KS, Novosad V, et al. Imprinting in exchange coupled antiferromagnetic/ferromag­ vortices into antiferromagnets. Phys Rev Lett. netic disks: distorted viscous vortex. Nanoscale. 2006;97:67201. 2015;7:9878–9885. [3] Ivanov YP, Chuvilin A, Vivas LG, et al. Single crystal­ [23] Hu JM, Nan CW. Opportunities and challenges for line cylindrical nanowires-toward dense 3D arrays of magnetoelectric devices. APL Mater. 2019;7:80905. magnetic vortices. Sci Rep. 2016;6:23844. [24] Quintana A, Zhang J, Isarain-Chávez E, et al. Voltage- [4] Yan M, Leaf G, Kaper H, et al. Spin-wave modes in induced coercivity reduction in nanoporous alloy a cobalt square vortex: micromagnetic simulations. films: a boost toward energy-efficient magnetic Phys Rev B. 2006;73:14425. actuation. Adv Funct Mater. 2017;27:1701904. [5] Cowburn RP. Change of direction. Nat Mater. [25] Quintana A, Menéndez E, Liedke MO, et al. Voltage- 2007;6:255–256. controlled ON-OFF ferromagnetism at room tem­ [6] Dussaux A, Georges B, Grollier J, et al. Large micro­ perature in a single metal oxide film. ACS Nano. wave generation from current-driven magnetic vortex 2018;12:10291. oscillators in magnetic tunnel junctions. Nat [26] Zehner J, Soldatov I, Schneider S, et al. Voltage- Commun. 2010;1:8. controlled deblocking of magnetization reversal in [7] Jones D, Zou J, Zhang S, et al. Energy storage in thin films by tunable domain wall interactions magnetic textures driven by vorticity flow. Phys Rev and pinning sites. Adv Electron Mater. B. 2020;102:140411(R). 2020;6:2000406. Sci. Technol. Adv. Mater. 22 (2021) 983 Y. CHEN et al. [27] Zehner J, Huhnstock R, Oswald S, et al. Nonvolatile [47] Kong S, Kumar N, Checchia S, et al. Defect-driven electric control of exchange bias by a redox transfor­ structural distortions at the surface of relaxor mation of the ferromagnetic layer. Adv Electron ferroelectrics. Adv Funct Mater. 2019;29:1900344. Mater. 2019;5:1900296. [48] De Rojas J, Quintana A, Lopeandía A, et al. Voltage- [28] Shimamura K, Chiba D, Ono S, et al. Electrical con­ driven motion of nitrogen ions: a new paradigm for trol of Curie temperature in cobalt using an ionic magneto-ionics. Nat Commun. 2020;11:5871. liquid film. Appl Phys Lett. 2012;100:122402. [49] Tan AJ, Huang M, Avci CO, et al. Magneto-ionic [29] Li X, Lee A, Razavi SA, et al. Voltage-controlled control of magnetism using a solid-state proton magnetoelectric memory and logic devices. MRS pump. Nat Mater. 2019;18:35–41. Bull. 2018;43:970–977. [50] Zhang Q, Luo X, Wang L, et al. Lithium-ion battery [30] Su W, Wang Z, Chen Y, et al. Reconfigurable magne­ cycling for magnetism control. Nano Lett. toresistive sensor based on magnetoelectric coupling. 2016;16:583–587. Adv Electron Mater. 2020;6:1901061. [51] Dasgupta S, Das B, Knapp M, et al. Intercalation- [31] Molinari A, Hahn H, Kruk R. Voltage-control of driven reversible control of magnetism in bulk magnetism in all-solid-state and solid/liquid magne­ ferromagnets. Adv Mater. 2014;26:4639–4644. toelectric composites. Adv Mater. 2019;31:1806662. [52] Zhou W, Tong WY, Wang Y, et al. Joint theoretical [32] Navarro-Senent C, Quintana A, Menéndez E, et al. and experimental study on the effects of the salts in Electrolyte-gated magnetoelectric actuation: phe­ the graphite-based dual-ion batteries. J Phys Chem C. nomenology, materials, mechanisms, and prospective 2019;123:18132−18141. applications. APL Mater. 2019;7:30701. [53] Gößler M, Topolovec S, Krenn H, et al. Nanoporous [33] Duschek K, Petr A, Zehner J, et al. All- Pd1-xCox for hydrogen-intercalation magneto-ionics. electrochemical voltage-control of magnetization in APL Mater. 2021;9:41101. metal oxide/metal nanoislands. J Mater Chem C. [54] Vasala S, Jakob A, Wissel K, et al. Reversible tuning of 2018;6:8411–8417. magnetization in a ferromagnetic Ruddlesden– [34] Bi C, Liu Y, Newhouse-Illige T, et al. Reversible con­ Popper-type manganite by electrochemical fluoride- trol of Co magnetism by voltage-induced oxidation. ion intercalation. Adv Electron Mater. 2020;6: Phys Rev Lett. 2014;113:267202. 1900974. [35] Cherifi RO, Ivanovskaya V, Phillips LC, et al. Electric- [55] Robbennolt S, Menéndez E, Quintana A, et al. field control of magnetic order above room Reversible, electric-field induced magneto-ionic con­ temperature. Nat Mater. 2014;13:345–351. trol of magnetism in mesoporous cobalt ferrite thin [36] Molinari A, Hahn H, Kruk R. Voltage-controlled on/ films. Sci Rep. 2019;9:10804. off switching of ferromagnetism in manganite [56] Navarro-Senent C, Fornell J, Isarain-Chávez E, et al. supercapacitors. Adv Mater. 2018;30:1703908. Large magnetoelectric effects in electrodeposited [37] Kim HKD, Schelhas LT, Keller S, et al. nanoporous microdisks driven by effective surface Magnetoelectric control of superparamagnetism. charging and magneto-ionics. ACS Appl Mater Nano Lett. 2013;13:884–888. Interfaces. 2018;10:44897–44905. [38] Gößler M, Albu M, Klinser G, et al. Magneto-ionic [57] Gilbert DA, Grutter AJ, Arenholz E, et al. Structural switching of superparamagnetism. Small. 2019;15:190 and magnetic depth profiles of magneto-ionic hetero­ 4523. structures beyond the interface limit. Nat Commun. [39] Waser R, Dittmann R, Staikov C, et al. Redox-based 2016;7:12264. resistive switching memories nanoionic mechanisms, [58] Mishra R, Kumar D, Yang H. Oxygen-migration- prospects, and challenges. Adv Mater. 2009;21:26 based spintronic device emulating a biological 32–2663. synapse. Phys Rev Appl. 2019;11:54065. [40] Pertsev NA, Kohlstedt H. Resistive switching via the [59] Sakamaki M, Amemiya K. Observation of an electric converse magnetoelectric effect in ferromagnetic multi­ field-induced interface redox reaction and magnetic layers on ferroelectric substrates. Nanotechnology. modification in GdOx/Co thin film by means of 2010;21:475202. depth-resolved X-ray absorption spectroscopy. Phys [41] Manchon A, Pizzini S, Vogel J, et al. X-ray analysis of Chem Chem Phys. 2018;20:20004–20009. the magnetic influence of oxygen in Pt/Co/AlOx [60] Bauer U, Yao L, Tan AJ, et al. Magneto-ionic control of trilayers. J Appl Phys. 2008;103:07A912. interfacial magnetism. Nat Mater. 2015;14:174–181. [42] Manchon A, Ducruet C, Lombard L, et al. Analysis of [61] Zhou X, Yan Y, Jiang M, et al. Role of oxygen ion oxygen induced anisotropy crossover in Pt/Co/MOx migration in the electrical control of magnetism in trilayers. J Appl Phys. 2008;104:43914. Pt/Co/Ni/HfO2 films. J Phys Chem C. [43] Ghidini M, Mansell R, Pellicelli R, et al. Voltage- 2016;120:1633–1639. driven annihilation and creation of magnetic vortices [62] Herrera Diez L, Liu YT, Gilbert DA, et al. Nonvolatile in Ni discs. Nanoscale. 2020;12:5652–5657. ionic modification of the Dzyaloshinskii-Moriya [44] Gilbert I, Chavez AC, Pierce DT, et al. Magnetic interaction. Phys Rev Appl. 2019;12:34005. microscopy and simulation of strain-mediated con­ [63] Robbennolt S, Yu P, Nicolenco A, et al. Magneto- trol of magnetization in PMN-PT/Ni nanostructures. ionic control of magnetism in two-oxide nanocom­ Appl Phys Lett. 2016;109:162404. posite thin films comprising mesoporous cobalt fer­ [45] Li Q, Tan A, Scholl A, et al. Electrical switching of the rite conformally nanocoated with HfO2. Nanoscale. magnetic vortex circulation in artificial multiferroic 2020;12:5987–5994. structure of Co/Cu/PMN-PT(011). Appl Phys Lett. [64] Van Den Brink A, Van Der Heijden MAJ, 2017;110:262405. Swagten HJM, et al. Large time-dependent coercivity [46] Sun E, Cao W. Relaxor-based ferroelectric single crys­ and resistivity modification under sustained voltage tals: growth, domain engineering, characterization application in a Pt/Co/AlOx/Pt junction. J Appl Phys. and applications. Prog Mater Sci. 2014;65:124–210. 2015;117:17C717. Sci. Technol. Adv. Mater. 22 (2021) 984 Y. CHEN et al. [65] Navarro-Senent C, Quintana A, Isarain-Chávez E, [71] Robbennolt S, Nicolenco A, Mercier Fernandez P, et al. Enhancing magneto-ionic effects in mag­ et al. Electric field control of magnetism in iron netic nanostructured films via conformal deposi­ oxide nanoporous thin films. ACS Appl Mater tion of nanolayers with oxygen acceptor/donor Interfaces. 2019;11:37338–37346. capabilities. ACS Appl Mater Interfaces. [72] De Rojas J, Quintana A, Lopeandía A, et al. Boosting 2020;12:14484–14495. room-temperature magneto-ionics in a non-magnetic [66] Matricardi C, Garcia-Pomar JL, Molet P, et al. High- oxide semiconductor. Adv Funct Mater. throughput nanofabrication of metasurfaces with 2020;30:2003704. polarization-dependent response. Adv Opt Mater. [73] Donahue MJ, Porter G. OOMMF user’s guide, 2020;82:000786. version 1.0 OOMMF user’s guide, version 1.0. [67] Gibert-Roca M, Molet P, Mihi A, et al. Near infrared 1999. organic photodetectors based on enhanced charge [74] Haratani S, Aoyama T, Sato I. Micromagnetic transfer state absorption by photonic architectures. simulations of polycrystalline cobalt islands for J Mater Chem C. 2020;8:9688–9696. patterned media. IEEE Trans Magn. 2000;36: [68] Molet P, Gil-Herrera LK, Garcia-Pomar JL, et al. 3164–3166. Large area metasurfaces made with spherical silicon [75] Arscott S. SU-8 as a material for lab-on-a-chip-based resonators. Nanophotonics. 2020;9:943–951. mass spectrometry. Lab Chip. 2014;14:3668. [69] Vernickaite E, Tsyntsaru N, Cesiulis H. [76] Qiao R, Zhang XL, Qiu R, et al. Fabrication of super­ Electrodeposition and corrosion behaviour of nanos­ paramagnetic cobalt nanoparticles-embedded block tructured cobalt–tungsten alloys coatings. Trans Inst copolymer microcapsules. J Phys Chem C. Met Finish. 2016;94:313–321. 2007;111:2426–2429. [70] Schneider CA, Rasband WS, Eliceiri KW. NIH image [77] Xiao S, Kolb JF, Malik MA, et al. Electrical breakdown to ImageJ: 25 years of image analysis. Nat Methods. and dielectric recovery of propylene carbonate. IEEE 2012;9:671–675. Trans Plasma Sci. 2006;34:1653–1661.