Introduction

During the past decade, data storage has become one of the most pressing issues that has confronted our society and the information age. The increasing use of the cloud for the storage of digital information involves the use of hard disk drives which use magnetism for the storage and retrieval of digital information1. In 2014, about 0.56 billion hard disk drives that provided more than 540 exabytes of storage capacity were produced1,2. However, due to problems associated with increasing the storage capacity of hard disk drives, the data storage community faces the necessity of exploring and developing new concepts in attempts to increase areal density1,2,3. In recent years, extensive efforts have been devoted to exploring alternate approaches for improving the storage density of magnetic recording devices, and include two-dimensional magnetic recording (TDMR)4,5, heat-assisted magnetic recording (HAMR)6,7,8,9, microwave-assisted magnetic recording (MAMR)10,11, bit-patterned magnetic recording (BPMR)12,13,14,15, and optical switching of magnetic domains16,17,18. For instance, a broadband near-field thermal extraction device based on hyperbolic metamaterials can significantly enhance near-field energy transfer by extracting evanescent waves with arbitrarily large lateral wave vectors6. Near-field thermal extraction has important practical implications in HAMR6,8. Free-floating magnetic structures composed of in situ cross-linked magnetically assembled nanoparticles have been reported to retain their patterned shape during manipulation with external magnetic fields12 while a magnetic pattern can serve as an elementary base in magnetic recording media such as BPMR12,14. By exploiting the field-confining capability as plasmonic nanoantennas, ferrimagnetic TbFeCo thin films have the potential for use in the optical switching of magnetic domains, while the importance of a chemically homogeneous sample structure is highlighted for optical switching-based recording technologies16. The diazotization of Ru(bpy)2(phen-NH2)2+ in the presence of carbon nanotubes (CNT) results in the formation of nanodots on CNTs, which are capable of transducing photo stimuli into electricity and magnetism at ambient conditions, and the resultant photomagnetic CNTs are a multifunctional material that have the potential for use in in optical switching-based recording technologies17.

Because of the higher storage density19,20,21 and lower power consumption22,23,24,25, the electric potential controlled magnetic device has attracted interest in the area of magnetic recording research. Exploring electric potential tuned magnetic devices in electrolytic conditions has also attracted great interest24,25,26,27,28 because the solution process is a low cost, rapid method that is amenable for use in commercial production. In the first report of electric potential induced magnetism using FePt and FePd films in NaOH as an electrolyte, the coercive force (HC) could be enhanced by about 4.5% by varying the electric potential between −1000 and −400 mV26. This pioneering work has motivated researchers to further examine electric potential tuned magnetism in electrolytes24,25,27,28. For example, MgO(2.0 nm)/Co (0.4 nm)/Pt (1.2 nm)/Ta (3.0 nm) films show magnetic switching near the coercive field when the electric potential is tuned from +2000 to −2000 mV in an ionic liquid electrolyte24. The ferromagnetic proximity effect shows that an electric field can be used to control the magnetic moments in films made of Pd, originally a nonmagnetic element25. An electrochemical etching process can be used to reduce the thickness of Co films with less surface defects while the electric field induced depression of the HC was observed to be reproducible for the case of an electrochemically etched specimen28. The focus of this research was on electric potential tuned magnetic phenomena that occur via the solution process with the characteristics of highly efficient controlling and ease of production.

Exploring the possible mechanisms responsible for developing electric potential tuned magnetic phenomena is an important issue in terms of developing more efficient magnetic recording devices. From the literature, the mechanisms of the electric potential tuned magnetic phenomena are mainly related to the Curie temperature24,27,29 and magnetic anisotropy energy (MAE)30,31,32,33. For example, the Curie temperature of Ga0.93Mn0.07As increases as the gate voltage sweeps from positive to negative values29. Based on first principle calculations of the electric field effects on magnetic anisotropy for ferromagnetic thin films on Pt(111), the MAE increases with increasing electric field and decreases with decreasing electric field30. Giant electric-field-assisted modifications of the interfacial magnetic anisotropy have been reported for the case of a half-metallic Heusler compound/MgO interface31. However, the mechanism responsible for how different thicknesses of the ferromagnetic layer (tFM) affect electric potential tuned magnetic phenomena is still unclear.

In this letter, tuning the magnetic properties by an electric potential for Co/Pt(111) of nanometer thickness in the Bloch-wall regime was investigated for a solution process. For Co/Pt(111) in the Bloch-wall regime, the depression of the HC is detected by applying a negative electric potential in an electrolytic condition. The electric potential tuned HC and tFM shows an inverse square relation that can be successfully explained using an electric potential tuning-magnetic anisotropy energy (EPT-MAE) model. For a smaller tFM, the electric potential induced change of the HC is larger, while the electric potential induced change of the MAE is again magnified. A higher controlling efficiency of the tunable HC in thinner ferromagnetic layers was demonstrated using our apparatus. Based on the characteristics of a tunable HC produced by changing the electric potential using a solution process, we propose a novel technique for achieving electric-potential-tuned magnetic recording (EPTMR). A schematic plot of the EPTMR with storage and writing processes is shown in Fig. 1. For a storage pillar, the HC is tunable by a negative electric potential, and ranged from a storage area of high HC to a writing area of relatively small HC within an available head field range. In additional, the patterned pillars are isolated in write or read process because of the parallel potential channel design. The parallel potential channel is designed to switch the write and read process by changing the supplied electric potential for each pillar that can avoid the exchange coupling between pillars. In these design, the potential controllable magnetic pillars can be structured as three-dimensional stacking like many electronic devices34 which improve the amount of pack clusters and enhance the areal density in a disc. Combining the solution process and controlling the electric potential for the operation of a magnetic device is a viable technique for achieving a highly controlled efficiency and a lower power consumption with a reduced cost of production. The developed EPTMR shows promise for use in applications in which the solution process is combined with electric potential control for magnetic data storage.

Figure 1: A schematic plot showing the storage and writing processes of EPTMR.
figure 1

The inset shows the designs of a magnetic recording disc for EPTMR.

Full size image

After waiting for a certain time at a suitable potential for Co deposition, a linear potential scan in the positive potential direction is performed to strip off the Co from the film. Anodic peaks from cyclic voltammograms (CV) for strip off measurements for Co/Pt(111) thinner than 18 nm are shown in Fig. 2a. The thickness of the cobalt can be obtained from the strip off method by measuring the quantity of electric charge, by integrating the charge under the anodic peak28,35. Figure 2b shows Kerr signals versus the magnetic field measured at electric potential V = −500 mV for a Co/Pt(111) layer thinner than 18 nm. No hysteresis is detected in the polar configuration. Hysteresis occurs only in the longitudinal configuration. This shows that the easy axis of magnetization lies on the surface plane. Shape anisotropy is dominant for the Co/Pt(111) in this thickness range36,37.

Figure 2
figure 2

(a) The CV of strip off measurements for Co/Pt(111). (b) Kerr signals versus the magnetic field measured at V = −500 mV for Co/Pt(111) in both the longitudinal (solid lines) and polar (dashed lines) configurations.

Full size image

Figure 3a shows the saturated Kerr intensities in both the longitudinal and polar configurations versus the Co thickness (tCo). As the tCo increases, the saturated Kerr intensity in the longitudinal configuration increases between 4.5 and 18 nm, due to the relation between the magnetization and the thickness of the ferromagnetic layers38. This phenomenon has also been reported for Co/Cu(001) films in electrolytic conditions39 and for Mn-doped ZnO films40. The relation between the saturation magnetization (MS) and tFM can be expressed by

Figure 3
figure 3

(a) Saturated Kerr intensities versus the thickness of the Co layer in both the longitudinal and polar configurations. (b) Coercive force versus Co thickness in the longitudinal configuration.

Full size image

by considering the amount of magnetic moments in a ferromagnetic layer38,39,40,41,42. A plot of the coercive force of Co/Pt(111) versus tCo is shown in Fig. 3b. The value for HC increases rapidly with increasing tCo below 6.8 nm. From the literature, Bloch wall energy density decreases with increasing film thickness because of the increased magnetostatic energy due to the appearance of charged surface above and below the wall, while the Néel wall energy shows opposite trend because of the proportional relation with the area of the charged surface inside the film43. The analysis of the demagnetization energy for Néel-wall regime shows that the increase of the HC by increasing the film thickness44,45. As the thicknesses of magnetic layers increase, the change from of domain-wall types from Néel wall to Bloch wall has been reported for different magnetic materials in nanometer thicknesses43,44,45,46,47. For Co/Pt(111) films thicker than 6.8 nm, HC decreases monotonously which follows Bloch wall behavior35,44,48. By increasing tFM, the increase in HC to a maximum at around several nanometers followed by a monotonous decrease for thicker films has been previously reported and can be described by changing the domain wall type from a Néel wall to a Bloch wall43.

In order to elucidate the role of the electric potential on the magnetic properties of Co/Pt(111) in Néel-wall and in Bloch-wall regimes, Fig. 4a and b show Kerr signals versus the magnetic field for 4.5 and 6.8 nmCo/Pt(111), respectively. For Co/Pt(111) in the Néel-wall regime (tCo = 4.5 nm), no significant change in HC is detected at different electric potentials. The Néel-wall is related to the film thickness in order to minimize the magnetostatic energy. At a sufficiently small film thickness, the magnetostatic energy is no longer significant and the Néel-wall is stabilized. In addition, the energy density of the Néel-wall is related to its magnetization and film thickness43. Therefore, the electric potential might not easily induce a change in the Néel-wall in few hundred millivolts. For Co/Pt(111) in the Bloch-wall regime (tCo = 6.8 nm), an electric potential induced change in HC occurs. For a larger negative electric potential (V = −500 mV), the value for HC is smaller. This result provides strong evidence for the negative electric potential induced depression in HC for Co/Pt(111) in the Bloch-wall regime. Variations in HC are shown in Fig. 4c for Co/Pt(111) in the Bloch-wall regime under conditions where the electric potential is repeatedly changed between V = −500 and −400 mV. The jump in HC between 0.42 and 0.48 kOe is highly reproducible within hours of the start of the experimental tests. At this electric potential range, the typical current is around few μA/cm2. The heating power density by the electric current is less than a few μW/cm2and therefore the effects of thermal annealing can be neglected. In additional, in the Bloch-wall regime, the films are robust against variations in the electric potential for magnetic switching44. We therefore conclude that controlling HC by means of the electric potential for magnetic films in the Bloch-wall regime represents a possible route for preparing spintronic switching devices. In the following discussions for the electric potential induced phenomena, we focus on Co/Pt(111) in the Bloch-wall regime.

Figure 4
figure 4

Kerr signals versus the magnetic field measured at V = −500 and −400 mV for (a) 4.5 and (b) 6.8 nm Co/Pt(111) in the longitudinal configuration. (c) Variations of the HC for 6.8 nm Co/Pt(111) under conditions of continuously changing electric potential V = −500 and −400 mV.

Full size image

As shown in Fig. 5a for Co/Pt(111) with tCo between 7 and 18 nm, the HC measured at both V = −500 and−400 mV decrease monotonously with increasing tCo. At V = −500 mV for an individual tCo, the reduction in HC is more pronounced. By tuning the electric potential V, the electric field E applied to the specimen can be adjusted. The change in the coercive force (ΔHC) at electric fields E1 and E2 can be defined as

Figure 5
figure 5

(a) HC versus Co thickness for Co/Pt(111) with Co thickness thicker than 6.8 nm measured at V = −500 and −400 mV in the longitudinal configuration. (b) The experimental data and simulation results for ΔHC versus the thickness of the Co layer. (c) The experimental data and simulation results for the ln(ΔHC) versus ln(tCo).

Full size image

Figure 5b shows plots of ΔHC versus tCo where the electric fields E1 and E2 correspond to the electric potentials V1 = −500 mV and V2 = −400 mV, respectively. The value for ΔHC decreases with increasing tCo. In a further analysis of the data in Fig. 5b by taking the logarithms of both the ΔHC and tCo, a well linear relation between ln(ΔHC) and ln(tCo) is observed, as shown in Fig. 5c. A slope of −2.00 ± 0.06 was obtained from the linear fitting of the data. This shows that the ΔHC is proportional to the inverse of the square of tCo. Based on the above experimental evidence, we proposed an EPT-MAE model to explain the inverse square relation between the electric potential tuned HC and the thickness of the ferromagnetic layer. Detailed discussions are below.

Both the magnetic anisotropy and the imperfections in ferromagnetic materials give rise to hysteresis loss and an increase in the coercive force44. Being an upper limit of the HC, an easy-axis magnetization process results in a square M-H loop, and is characterized in the single domain behavior or pinned wall limit by rotational hysteresis, i.e.,

where Ku is the MAE of the ferromagnetic layer43. By considering the hysteresis loss energy for the pinning of the domain wall motion, the lower limit of the coercive force can be expressed by

where n is the amount of pinning sites and ε is the pinning energy for a domain wall motion44. Combining Eqs (3) and (4), the coercive force can be expressed as

where p (0 ≤ p ≤ 1) is the percentage of the contribution of the magnetic anisotropy energy. By tuning the electric potential applied to a ferromagnetic film, previous investigations indicate that the corresponding magnetoresistance49,50, Hall effects51,52 and magnetization switching10,23,28 are highly reproducible. By choosing suitable electric potential ranges, the changing of pinning sites such as impurities, grain boundary, dislocations, and voids in the ferromagnetic films can be avoided10,23,28,49,50,51,52. Assuming that pinning sites are independent of the applied electric potential, the pinning of domain wall motion related to hysteresis loss energy can be neglected for an analysis of ΔHC. Therefore, from Eq. (5), ΔHC can be expressed by

Because the MS is proportional to tFM as shown in Eq. (1), the relation between the tFM and the ΔHC can be expressed by

In previous reports, the first principle calculations for Fe and Co films on Pt(111) show that the MAE is enhanced when the electric field E is increased30. The electric voltage effect on MAE of magnetic layers is expected via charge redistribution among different electron d-orbitals (some or all of dxy, dyz, dxz, dz2, and dx2 − y2) in the presence of electric field53. For ultrathin films, it has been reported that the absorption atoms such as H+, OH− and CO on the top of Co can influence the interface MAE via controllable electric field effect54,55. In this report, we start from Co/Pt(111) thicker than 4.5 nm where the MAE contains both the bulk and interfacial contributions. The adsorbed species are H+ and K+ in the electrolyte while the large electric field created by the electrical double layer near the electrode surface is responsible to the change of the MAE of the magnetic system. The electric field applied on the specimen can be adjusted by tuning the electric potential V. The change in the electric field (ΔE) can be defined as

Under different electric fields, the MAEs are different. In this study, the magneto-optical Kerr effect (MOKE) measurements were carried out at electric potentials V1 = −500 mV and V2 = −400 mV. By further considering the thickness of the ferromagnetic layer, the relation between the change in magnetic anisotropy energy ΔKu at the electric fields E1 and E2 (corresponding to V1 = −500 mV and V2 = −400 mV, respectively) and the change in the electric field ΔE can be expressed as

where β is the interfacial magnetic anisotropy energy coefficient31. By substituting Eq. (9) into Eq. (7), the dependence of ΔHC on layer thickness can be expressed as

From the discussions of the electric field induced change of MAE, the ΔHC versus tFM shows an inverse square relation, as shown in Eq. (10). This relation is in good agreement with the experimental data in Fig. 5c where the slope of a plot of ln(ΔHC) versus ln(tCo) is −2.00 + 0.06. This result shows that combining MAE and electric field effects can successfully explain the inverse square relation between the ΔHC and the tFM.

Based on Eqs (7, 9, and 10), an EPT-MAE model that explains the relation between the electric potential tuned HC and tFM is proposed and is schematically illustrated in Fig. 6. For a smaller tFM, the electric potential induced change in HC is larger, as indicated in Eq. (7). The electric potential induced change of MAE again magnifies the change in HC as suggested in Eq. (9). The resultant effects are responsible for the inverse square relation between ΔHC and tFM in Eq. (10). In the left panel of Fig. 6, a more negative electric potential shows the effect of reduction of HC. Compared to the right panel of Fig. 6 for a half tFM, the electric potential induced ΔHC is enhanced by four times. This shows that the electric potential induced ΔHC is much more pronounced for thinner ferromagnetic layers. From a technological point of view, by changing the electric potential in ferromagnetic layers with smaller thicknesses, a higher controlling efficiency of the tunable HC is demonstrated in our apparatus. The establishment of the EPT-MAE model provides insights into our knowledge of the relation between electric potential tuning HC, the ferromagnetic layer thickness, and associated mechanisms.

Figure 6
figure 6

A schematic plot shows the electric potential induced change of the coercive force for ferromagnetic layers with different thicknesses in the Bloch-wall regime.

Full size image

For magnetic recording applications, superparamagnetic behavior is the limitation for further reducing the volume of a magnetic bit, but EPTMR is a possibility, as shown in Fig. 1. From the literature, in order for magnetic grains in a typical size distribution to remain stable for ~10 years, the ambient temperature stability factor KuV/kBT needs to be larger than 70 for a commercial device9; where V is the volume of a nanoparticle; kB is the Boltzmann constant; and T is the sample temperature. To enhance the areal density of a magnetic recording disc, increasing the MAE is the key point. When using a material with a higher MAE, a higher writing field is necessary, but this is a disadvantage for real applications. By employing an electric potential tuned mechanism, the findings reported here show that 100 mV can reduce the coercive force by 15% for thin magnetic films (6.8 nm Co/Pt(111) as an example) in the Bloch wall regime. In practice, by using EPTMR, it is possible to create stable recording media with a high writing ability.

In conclusion, for Co/Pt(111) of nanometer thickness in the Bloch-wall regime, a depression in HC is detected when a negative electric potential is applied in an electrolytic condition. For Co/Pt(111) under conditions of a continuously changing electric potential, the increase in HC between two states is highly reproducible. As the value for tFM increases, both the HC measured at V = −500 and −400 mV decrease monotonously. The reduction of HC for an individual tCo is more pronounced at a more negative potential. An electric potential tuned HC and tFM have an inverse square relationship with one another. By changing the electric potential in ferromagnetic layers with smaller thicknesses, a higher controlling efficiency of tunable HC is achieved. Assuming that the pinning domains are independent of the applied electric potential, an EPT-MAE model in which the MAE and the electric field effects are combined is established. This provides insights into our knowledge of the relation between electric potential tuning HC and the thickness of the ferromagnetic layer. The implementing of EPTMR is a potentially promising route for applications combining a solution process and electric potential control in magnetic data storage.

Methods

A platinum crystal was oriented within 0.5° of the [111] direction as checked by x-ray diffraction. The diameter of the crystal is 8 mm while the thickness is 2 mm. This specimen size is suitable for the designed apparatus for the MOKE measurements in electrochemical environments28,35,56. Before loading into the electrochemical cell, the surface of the crystal was mechanically polished with alumina powder with a diameter down to 0.05 μm and was then annealed using a hydrogen torch. The details of the process for preparing Pt electrode can be found in the literature28,57,58,59. All solutions were prepared with high purity chemicals (HCl and CoCl2) and ultra-pure water(>18 MΩ cm). To avoid contamination by oxygen, the solutions were purged with high purity argon for 1 hr before the experiments. The electrochemical cell used for the CV and MOKE measurements was developed and built in our laboratory34,56. Pt and Ag wires were used as the counter electrode and the reference electrode, respectively. The electricity of the standard three-electrode setup was controlled by a potentiostat that was constructed in Bonn34,56,60,61. The CV of the Pt(111) electrode in a pure supporting electrolyte containing HCl and KCl, and with a CoCl2 additive are shown in the Supplementary Information 1. For the condition of the film deposition, by comparing the CV for specimens in electrolytes with and without CoCl2, the electric potential range that is suitable for the electrodeposition of Co/Pt(111) was determined. The Co layer was electrodeposited at an electric potential more negative than −700 mV for a pre-determined time, depending on the thickness of the layer. The potential range for MOKE measurements was chosen to be between −400 and −600 mV to avoid the possible deposition/dissolution of the Co deposits. A He–Ne laser with a wavelength of 632.8 nm was used as the light source for the MOKE measurements. The Kerr signals as a function of magnetic field were measured by a photodiode and recorded by a computer-controlled multimeter to generate the hysteresis loops.

By applying an electric potential, an electrical double layer (EDL) near the electrode surface was established62. The thickness of the EDL was in the nanometer scale, thus permitting hundreds of millivolts to create an electric field as large as 108V/m25,63. A schematic plot in Supplementary Information 2 elucidates the specimen structure where the electric field influences the ferromagnetic layer supported on the Pt electrode through the establishment of the EDL. The electric field is responsible for the change in MAE and the related change in the HC. Surface species are different in different electrolytes, and the effects of point of a zero charge (PZC) need to be taken into consideration. In this report, we used the same electrolyte and the same electrochemical conditions, the PZC was not altered. Discussions of the EDL as well as the PZC can be found in the Supplementary Information 2 and may be of importance for further applications in the case for solution processes.

Additional Information

How to cite this article: Chang, C.-H.-T. et al. Tuning coercive force by adjusting electric potential in solution processed Co/Pt(111) and the mechanism involved. Sci. Rep. 7, 43700; doi: 10.1038/srep43700 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.