nanomaterials Article
Two-Step Magnetization Reversal FORC Fingerprint of Coupled Bi-Segmented Ni/Co Magnetic Nanowire Arrays Javier García Fernández 1,2, *, Víctor Vega Martínez 2,3 , Andy Thomas 1 , Víctor Manuel de la Prida Pidal 2 and Kornelius Nielsch 1 1 2 3
*
Leibniz Institute for Solid State and Materials Research (IFW) Dresden, Helmholtzstraße 20, 01069 Dresden, Germany;
[email protected] (A.T.);
[email protected] (K.N.) Departamento de Física, Universidad de Oviedo, C/Federico Garcia Lorca 18, 33007 Oviedo, Asturias, Spain;
[email protected] (V.V.M.);
[email protected] (V.M.d.l.P.P.) Laboratorio Membranas Nanoporosas, Servicios Científico-Técnicos, Universidad de Oviedo, Campus El Cristo s/n, 33006 Oviedo, Asturias, Spain Correspondence:
[email protected]; Tel.: +34-985-103-347
Received: 11 June 2018; Accepted: 17 July 2018; Published: 19 July 2018
Abstract: First Order Reversal Curve (FORC) analysis has been established as an appropriate method to investigate the magnetic interactions among complex ferromagnetic nanostructures. In this work, the magnetization reversal mechanism of bi-segmented nanowires composed by long Co and Ni segments contacted at one side was investigated, as a model system to identify and understand the FORC fingerprint of a two-step magnetization reversal process. The resulting hysteresis loop of the bi-segmented nanowire array exhibits a completely different magnetic behavior than the one expected for the magnetization reversal process corresponding to each respective Co and Ni nanowire arrays, individually. Based on the FORC analysis, two possible magnetization reversal processes can be distinguished as a consequence of the ferromagnetic coupling at the interface between the Ni and Co segments. Depending on the relative difference between the magnetization switching fields of each segment, the softer magnetic phase induces the switching of the harder one through the injection and propagation of a magnetic domain wall when both switching fields are comparable. On the other hand, if the switching fields values differ enough, the antiparallel magnetic configuration of nanowires is also possible but energetically unfavorable, thus resulting in an unstable magnetic configuration. Making use of the different temperature dependence of the magnetic properties for each nanowire segment with different composition, one of the two types of magnetization reversal is favored, as demonstrated by FORC analyses. Keywords: nanowires; electrochemical deposition; First Order Reversal Curves; two-step magnetization reversal
1. Introduction The continuous growth of the nanotechnology makes the understanding of nature at such low scale necessary to develop more advanced devices. Generally, high yielding fabrication processes of nanomaterials do not allow to produce perfectly identical nanostructures, whose functional properties differ slightly one from another. Then, it is a challenge to measure and define the main parameters that represent the population of the fabricated nanostructures. In this context, self-ordered porous Anodic Aluminum Oxide membranes (AAOs) when used as templates, have been established as a large scale, suitable and low cost fabrication technique of a wide range of nanomaterials such as antidot thin films, nanotubes and nanowires, among others [1–4]. When the pores of the AAO membranes are filled with Nanomaterials 2018, 8, 548; doi:10.3390/nano8070548
www.mdpi.com/journal/nanomaterials
Nanomaterials 2018, 8, 548
2 of 15
a metallic material by means of electrodeposition techniques, the deposited material reproduces the pore shape, thus forming a self-ordered nanowire array [5,6]. Among the applications in other research fields such as photonics, electronics and sensing devices, ferromagnetic nanowires are also interesting for high density magnetic data storage and spintronics [7]. The latter application has been a subject of extensive study and requires the control of magnetic domain walls and magnetization reversal processes of the nanomaterial. With this purpose, heterogeneous ferromagnetic nanowires are being fabricated by modulating their shape, changing their diameter, and/or their composition, performing multi-segments and multi-layers [4,8–12]. Nevertheless, due to the low dimensions of the nanowires, their magnetic characterization is frequently performed when nanowires remain embedded into the pores of the alumina membrane, in order to have enough signal when many nanowires contribute to the total magnetization of the sample. However, when the nanowires are still arranged in the alumina template, the magnetostatic interactions among the nanowires become important due to the short inter-nanowire distances [13,14]. Consequently, it is still a challenge to extrapolate individual magnetic behaviors of complex nanowires systems from the macroscopic magnetic characterization of the whole nanowire array. The FORC analysis is in general a method to evaluate a hysteretic behavior whose detailed measuring principles and interpretations on different particle systems can be found in the literature [15]. When applied to magnetism, the main idea behind the method is to consider that the major hysteresis loops are a superposition of individual and, in principle, indivisible elements of hysteresis called as “hysterons”. Starting from positive saturation, the applied magnetic field is reduced to a certain value and increased again back to saturation recording the first order reversal curve itself. Then, this process is repeated for several reversal fields (Hr) to cover the whole descendant branch of the hysteresis loop. Finally, the FORC distribution is calculated from the mixed derivative of the magnetization with respect to the reversal (Hr) and applied (H) fields, as shown in Equation (1), thus evaluating the change on the magnetic susceptibility at different first order reversal curves. ρFORC (H, Hr ) = −
1 ∂2 m 2 ∂Hr ∂H
(1)
However, it is worth mentioning that, due to time considerations as well as technical limitations, in a FORC experiment, a reversal performed nanowire by nanowire (hysteron by hysteron) is rather difficult. In turn, the magnetic field step will reverse a group of hysterons, so the resulting apparent hysteron will be composed of a group of hysteretic elements. This fact will make the FORC diagram to be formed of “bigger” pixels, although, if the field step is fine enough, it can still reproduce a holistic view of the system. The aim of this work was to identify the FORC fingerprint of a two-step magnetization reversal process proposing a model to validate the experimental FORC analysis performed in a type system. For this purpose, a bi-segmented Ni/Co nanowire array composed by a 20 µm Ni long segment followed by 20 µm Co one was studied. The effect of the exchange coupling and/or magnetostatic interactions at the interface layer over the magnetic properties when both composition segments are in contact was investigated. The limitations of performing the magnetic characterization of a single ferromagnetic nanowire make the understanding of the magnetic behaviors of these complex systems difficult. However, First Order Reversal Curve (FORC) analysis provides valuable information concerning the intrinsic properties of the nanowires as well as the magnetic interactions among them [16–18]. 2. Materials and Methods 2.1. Samples Fabrication and Characterization AAO membranes were fabricated by means of the well-known two step anodization process, whose general details can be found elsewhere [19]. Particularly, for this study, the AAO membranes were anodized in 0.3 M oxalic acid at 3 ◦ C leading to a pore diameter of around 35 nm and distance
Nanomaterials 2018, 8, 548
3 of 15
among adjacent pores of 105 nm. The total anodization time was chosen so that the thickness of the AAO membranes would be 50 µm in all cases. Further, the remaining aluminum substrate was removed from the bottom side of the alumina membrane with a solution of 36 g/L of CuCl2 ·2H2 O + Nanomaterials 8, x FOR PEER REVIEW 3 ofpart 15 of the 500 mL/L of HCl 2018, at 37%. The barrier layer of the AAO membranes that occludes the bottom pores was removed by wet chemical etching in 5%wt phosphoric acid. After opening the barrier layer, among adjacent pores of 105 nm. The total anodization time was chosen so that the thickness of the the mean pore diameter was found to be approximately 42 nm, while the interpore distance of 105 nm AAO membranes would be 50 µm in all cases. Further, the remaining aluminum substrate was remained unaltered theside etching As the electrodeposition material removed from during the bottom of theprocess. alumina membrane with a solution of of 36 metallic g/L of CuCl 2·2H2O +requires an electrical contact, a thin AuThe film was layer sputtered one side of the alumina This 500 mL/L of HCl at 37%. barrier of the on AAO membranes that occludesmembrane. the bottom part of process the poresby was removed by wet chemical etching ainhomogeneous 5%wt phosphoriccontact acid. After opening the barrier was followed Au electroplating to perform occluding completely the layer, the mean diameter found to be approximately nm, whileof thethe interpore distance of channels at this side pore of the AAOwas membranes. To compare 42 the effect coupling between the 105 nm remained unaltered during the etching process. As the electrodeposition of metallic material two ferromagnetic segments with different composition, arrays of same dimensions but only filled requires an electrical contact, a thin Au film was sputtered on one side of the alumina membrane. with NiThis or Co segments were fabricated as a reference of each nanowires composition (Figure 1a). process was followed by Au electroplating to perform a homogeneous contact occluding The electrodeposition of Co and Ni nanowires was carried out by means of the Watts type electrolyte completely the channels at this side of the AAO membranes. To compare the effect of the coupling between two ferromagnetic segments arrays of g/L same H dimensions but300 g/L with fixed pH the of 4.3: 300 g/L NiSO O + different 45 g/Lcomposition, NiCl2 ·6H2 O + 45 4 ·6H2with 3 BO3 and only2 O filled or Co segments were fabricated as a reference of each nanowires composition · CoSO4 ·7H + 45with g/LNiCoCl 6H O + 45 g/L H BO , respectively. The temperature of the electrolytes 2 2 3 3 (Figure 1a). The electrodeposition of Co and Ni nanowires was carried out by means of the Watts was kept at 45 ◦ C during the electrodeposition to avoid boric acid precipitation. Co and Ni nanowires type electrolyte with fixed pH of 4.3: 300 g/L NiSO4·6H2O + 45 g/L NiCl2·6H2O + 45 g/L H3BO3 and were potentiostatically electrodeposited at fixed −1 V and −1.2 V vs. Ag/AgCl reference electrode, 300 g/L CoSO4·7H2O + 45 g/L CoCl2·6H2O + 45 g/L H3BO3, respectively. The temperature of the respectively, for 10was minkept in both Thethe bi-segmented Ni/Co nanowire array represented Figure 1b, electrolytes at 45cases. °C during electrodeposition to avoid boric acid precipitation. Coinand hereafter as coupled Ni/Coelectrodeposited sample, was prepared a first of pure Ni Nireferred nanowirestowere potentiostatically at fixed −1by V and −1.2electrodeposition V vs. Ag/AgCl reference electrode, respectively, for 10electrodeposition min in both cases. The Ni/Co nanowire represented long segment and a subsequent ofbi-segmented pure Co long segment byarray sequentially using the in Figure 1b, hereafter referred to as coupled Ni/Co sample, was prepared by a first electrodeposition two different electrolytes above mentioned. Although the single segment samples provide information of pure Ni long segment and a subsequent electrodeposition of pure Co long segment by sequentially regarding the intrinsic magnetization reversal behavior of the single element array, two pieces of Co using the two different electrolytes above mentioned. Although the single segment samples provide and Ni information nanowire regarding arrays were also stacked and measured together to reproduce a non-interacting the intrinsic magnetization reversal behavior of the single element array, two bimagnetic system (decoupled Ni/Co sample from nowand on),measured where the Ni and Co segments were not pieces of Co and Ni nanowire arrays were also stacked together to reproduce a nondirectlyinteracting in contactbimagnetic and far from each other (Figure Even if the of the system (decoupled Ni/Co 1c). sample from nowhysteresis on), whereloop the Ni and decoupled Co segmentswas wereexpected not directly contact far from each other (Figure Evenelement if the hysteresis Ni/Co sample toinshow a and trivial superposition of the 1c). single Co andloop Ni arrays, of the decoupled Ni/Co sample was expected to show a trivial superposition of the single element Co the FORC diagram of this system could show unexpected contributions due to the presence of the and Ni arrays, the FORC diagram of this system could show unexpected contributions due to the two different magnetic materials, which should be considered in the interpretation of the coupled presence of the two different magnetic materials, which should be considered in the interpretation of Ni/Co FORC results. the coupled Ni/Co FORC results.
Figure 1. Schematic drawings of the different samples: (a) single element Ni or Co nanowire arrays;
Figure 1. Schematic drawings of the different samples: (a) single element Ni or Co nanowire arrays; (b) coupled Ni/Co bi-segmented nanowire array; and (c) decoupled Ni/Co nanowire array where the (b) coupled bi-segmented array; andIn(c) Ni/Co nanowire where the arrowNi/Co represents the directionnanowire of the arrays stacking. alldecoupled figures, the Ni and Co segments array are green arrow represents the direction of the arrays stacking. In all figures, the Ni and Co segments are green and blue respectively, while the Au bottom contact is yellow. and blue respectively, while the Au bottom contact is yellow.
Nanomaterials 2018, 8, 548
4 of 15
Finally, morphological and compositional characterization have been performed by means of a Scanning Electron Microscope (SEM, JEOL 6100 LV, JEOL, Akishima, Tokyo, Japan) equipped with an Energy Dispersive X-ray microanalysis system (EDX, Inca Energy 200, Oxford Instruments, Abingdon, UK). The length of the Co and Ni segments after 10 min of electrodeposition is around 16 µm and 20 µm in each case, respectively, thus the high aspect ratio of the nanowires is assured. On the other hand, magnetic hysteresis loops carried out along parallel and perpendicular directions to nanowires long axis, together with FORC measurements along the parallel direction and at different temperatures, have been performed in a Vibrating Sample Magnetometer (VSM-Versalab, Quantum Design, San Diego, CA, USA) under a magnetic field up to ±3T and in the temperature range between 50 K and 400 K. 2.2. Statistical Approach of Modelling FORC Simulations The FORC method has been widely employed for the evaluation of individual properties of ferromagnetic nanoentities as extracted from macroscopic measurements such as in the case of magnetic nanowire arrays [17,18,20]. This method is based on the assumption that, in this kind of systems, the major magnetic hysteresis loop corresponds to the superposition of several magnetization reversal processes that may be attributed to the contribution of each nanowire into the array. Then, the reversal of the magnetization for each nanowire can be considered as an indivisible process of hysteresis called hysteron. This entity can be defined by its two intrinsic switching fields Hb and Hr . The first one corresponds to the applied magnetic field necessary to saturate the hysteron along the positive direction of magnetization, while Hr is the field necessary to saturate the hysteron back to negative saturation. Through the change of coordinates Hc = ( Hb − Hr )/2 and Hu = −( Hb + Hr )/2, it is possible to achieve valuable information concerning the intrinsic magnetic anisotropies and the magnetostatic interactions among the nanowires, respectively. Taking into account that, from the experimental point of view, not all the nanowires are identical in size and/or composition, the system can be described with a distribution of hysterons whose magnetic properties differ slightly from the mean ones. In this case, a Gaussian type distribution of hysterons, as in Equation (2), has been chosen through defining its mean switching field h HSW i and with its standard deviation (σ): ( H −h HSW i) 1 − 2σ2 G ( H, h HSW i) = √ e σ 2π
2
(2)
From previous equation, and modifying the expression to obtain normalized values, the magnetization can be obtained by a direct integration of the hysterons distribution: Hb − h HSW i √ m( Hb , Hr ) = m Hr + 1 + erf 2σ2
·θ (−Hr − Hb ) + θ ( Hb + Hr )
(3)
where m Hr denotes the value of the magnetization at a certain reversal field Hr . This value is obtained from the integration of the negative hysteron distribution centered at −h HSW i, i.e., − G ( H, −h HSW i). Furthermore, θ is the Heaviside function that takes the value 1 when its argument is positive and 0 otherwise. Then, the Heaviside function is important to take into account when the magnetization in Equation (3) reaches its saturation state. Finally, the switching process of one of the nanowires in the array is able to modify the overall magnetic field felt by its neighbors due to the spatial proximity among them. If the switching process is supposed to be homogeneous enough along the whole array, its effect is to produce an interaction magnetic field, called mean field interaction (MFI), pointing along the opposite direction with respect to the direction of the magnetization, which can be seen as a demagnetizing field [21–23]. In other words, the effective magnetic field that is felt by the nanowires in the array is reduced by a factor that is proportional to, among others, the distance among nanowires and their saturation magnetization [24]. This contribution can be added into the above equations
Nanomaterials 2018, 8, 548
5 of 15
through a renormalization of the magnetic field, distinguishing between both the applied and effective magnetic fields, Happ and He f f , respectively, [25–28] as Equation (4). He f f = Happ + αm
(4)
Nanomaterials 2018, 8, x FOR PEERof REVIEW 5 of 15 where the negative interaction magnetostatic origin among nanowires can be represented by α < 0, and denotes the strength of the interaction field at saturation (m = 1). where the negative interaction of magnetostatic origin among nanowires can be represented by 𝛼 Finally, to reproduce the magnetic behavior of the coupled Ni/Co bisegmented nanowire array, 0, and denotes the strength of the interaction field at saturation (𝑚 = 1). the contribution coupled segments be included. The schematic representation of Finally,oftoexchange reproduce the magnetic behavior must of the coupled Ni/Co bisegmented nanowire array, such magnetic behavior shown coupled in Figure 2. Let us consider the case a single representation bisegmented of nanowire the contribution of is exchange segments must be included. Theofschematic suchofmagnetic is shown in Figure 2. Let us consider case of a fields single H bisegmented composed a Ni andbehavior Co segments with their respective intrinsictheswitching Ni and HCo when nanowire composed of a Ni and Co segments with their respective intrinsic switching fields and the first HNi < HCo . Starting from positive saturation and then decreasing the applied magnetic𝐻 field, 𝐻 when 𝐻 𝐻 . Starting from positive saturation and then decreasing the applied magnetic switching event to occur would come from the Ni segment. However, as it is coupled with the Co field, the first switching event to occur would come from the Ni segment. However, as it is coupled segmentwith whose magnetization still pointingisalong the positive direction, effective switching field the Co segment whoseismagnetization still pointing along the positive the direction, the effective coupled of the coupled Ni segment is slightly increased H > H . Furthermore, it is reasonable 𝐻 . Furthermore, it isto think switching field of the coupled Ni segment is slightly increased Ni 𝐻 Ni reasonable to think that further decreasing the magnetic the switching of Ni segment that further decreasing the magnetic field, the switching offield, Ni segment will induce the Co will segment to coupled 𝐻 . In terms of FORC, induce the Co segment to switch earlier than its intrinsic value 𝐻 switch earlier than its intrinsic value HCo < HCo . In terms of FORC, starting the curve from the starting the curve from the antiparallel configuration (red line in Figure 2), the Ni segment is forced antiparallel configuration (red line in Figure 2), the Ni segment is forced to switch before its intrinsic to switch before its intrinsic switching field due to the magnetic interaction with the Co one so,coupled FORC,coupled , switching magnetic < HNi < HNi . 𝐻 the 𝐻 . interaction with the Co one so, HNi 𝐻 field due to
Figure 2. Schematic representation of the two-step magnetization reversal hysteresis loop of a
Figure 2. Schematic representation of the two-step magnetization reversal hysteresis loop of a bisegmented Ni(green)/Co(blue) nanowire. Dashed lines denote the intrinsic hysteresis loops for Ni bisegmented Ni(green)/Co(blue) nanowire. Dashed lines denote the intrinsic hysteresis loops for Ni and Co segments. The arrows indicate the change on the switching fields with respect their intrinsic and Covalues segments. arrows coupling indicatebetween the change the fields respect their intrinsic due to The the magnetic the Coon and Ni switching segments. Red line with indicates the expected values due to the between thefrom Co the andantiparallel Ni segments. Red line indicates the expected behavior of magnetic a first ordercoupling reversal curve starting configuration. behavior of a first order reversal curve starting from the antiparallel configuration. 3. Results and Discussion
3. Results Discussion 3.1. and Ni and Co Nanowire Arrays To understand the magnetic behavior of the coupled Ni/Co bisegmented nanowires sample, the 3.1. Ni and Co Nanowire Arrays
intrinsic properties of the single element arrays must be magnetically characterized in detail first. In
ToFigure understand theroom magnetic behavior of the loops coupled Ni/Coalong bisegmented 3a,b, the temperature hysteresis measured both the nanowires parallel and sample, perpendicular directions with respect the nanowire length axis are shown. First, in both cases, onein detail the intrinsic properties of the single element arrays must be magnetically characterized can observe a well-defined uniaxial magnetic anisotropy with its easy magnetization axis lying first. In Figure 3a,b, the room temperature hysteresis loops measured along both the parallel and parallel to the nanowires long axis. Moreover, by comparing each parallel hysteresis loops (Figure perpendicular directions with respect the nanowire length axis are shown. First, in both cases, 3c), Co segments exhibit larger coercive field (𝐻
= 1296 𝑂𝑒) than Ni ones (𝐻
= 884 𝑂𝑒). The fact
Nanomaterials 2018, 8, 548
6 of 15
one can observe a well-defined uniaxial magnetic anisotropy with its easy magnetization axis lying parallel to the nanowires long axis. Moreover, by comparing each parallel hysteresis loops (Figure 3c), Co segments exhibit field (HcCo = 1296 Oe) than Ni ones (HcNi = 884 Oe). The fact that Nanomaterials 2018,larger 8, x FORcoercive PEER REVIEW 6 of 15 Co hysteresis loop is more tilted along the magnetic field axis, while maintaining its width, evidences that Co hysteresis loop is more tilted along the magnetic field axis, while maintaining its width, that such segments are submitted to stronger magnetostatic interactions, associated to its stronger evidences that such segments are submitted to stronger magnetostatic interactions, associated to its saturation magnetization value. stronger saturation magnetization value.
Figure 3. Room temperature hysteresis loops measured along the parallel and perpendicular (black)
Figure 3. Room temperature hysteresis loops measured along the parallel and perpendicular (black) direction with respect to the nanowire long axis in (a) Co (blue) and (b) Ni (green) segment arrays. direction with respect to the nanowire long axis in (a) Co (blue) and (b) Ni (green) segment arrays. The comparison between the parallel hysteresis loops of Co and Ni segment arrays is shown in (c). The comparison between the parallel hysteresis loops of Co and Ni segment arrays is shown in (c). To extract the intrinsic switching field distribution (SFD) of the Co and Ni nanowire arrays and evaluate the magnetostatic interaction among them, FORC analysis was performed on these samples. To extract the intrinsic switching field distribution (SFD) of the Co and Ni nanowire arrays The parallel FORC diagrams of both samples are presented in Figure 4a,b. First, in the case of the Co and evaluate the magnetostatic interaction among them, FORC analysis was performed on these nanowire array, the FORC distribution shows two clear branches, which are typically present in samples.systems The parallel FORC diagrams both samples are presented in Figuremagnetostatic 4a,b. First, in the where the switching field of distribution of hysterons and antiparallel case of the Co nanowire the branches FORC distribution shows two branches, which are typically interactions coexist.array, These two of the FORC diagram were clear referred to as “wishbone” shape by systems Pike et al.,where who demonstrated experimentally and theoretically its origin settled a guide to present in the switching field distribution of hysterons and and antiparallel magnetostatic extractcoexist. the relevant information from this kind of FORC diagrams [29]. were However, it results interactions These two branches of the FORC diagram referred tonon-trivial as “wishbone” the bending of the distribution to lower HC values when Hu is negative. Such behavior is not ascribed shape by Pike et al., who demonstrated experimentally and theoretically its origin and settled a to an asymmetry of the SFD, but instead to an inhomogeneity of the mean field interaction. As guide todiscussed extract the relevant information from this kind of FORC diagrams [29]. However, it results in Section 2.2, under the MFI theory, the magnetostatic interaction among nanowires in the non-trivial the bending of to the to This lower HC that, values when Hu is negative. Such is array is proportional thedistribution magnetization. means at the coercive field value where thebehavior net magnetization is negligible, theSFD, strength of magnetostatic interaction should become not ascribed to an asymmetry of the but instead to an inhomogeneity of thealso mean fieldzero. interaction. However, the characteristic of nanowires in case of AAO templates can play an As discussed in Section 2.2, underarrangement the MFI theory, the magnetostatic interaction among nanowires in important role on how the nanowires interact near the coercive field of the system. This the array is proportional to the magnetization. This means that, at the coercive field value where the inhomogeneity can be taken into account by adding a second order term to the MFI as follows [24,30]:
net magnetization is negligible, the strength of magnetostatic interaction should also become zero. 𝐻 =of𝐻nanowires + 𝛼𝑚 + 𝛽(1 − 𝑚 of ) AAO templates can play an(5) However, the characteristic arrangement in case important role on howIn the nanowires interact near the coercive field of the system. This inhomogeneity can be Figure 4c,d, the effect of the β parameter on simulated FORC distributions can be observed. taken into account by adding a second order term to the MFI as follows [24,30]: The major effect is to bend the upper part of the vertical branch towards lower values of HC and to displace the main peak to negative values of the interaction field. The effect of β on 𝐻 at the coercive field (𝑚 = 0 in Equation out He f f(5)) = points Happ + αmthat + βthe1coercive − m2 field extracted from the major hysteresis loop does not correspond to the maximum of the SFD in the FORC distribution (see Table
(5)
In Figure 4c,d, the effect of the β parameter on simulated FORC distributions can be observed. The major effect is to bend the upper part of the vertical branch towards lower values of HC and to
Nanomaterials 2018, 8, 548
7 of 15
displace the main peak to negative values of the interaction field. The effect of β on He f f at the coercive 2018, 8, x FOR REVIEW 7 of 15 field (m Nanomaterials = 0 in Equation (5))PEER points out that the coercive field extracted from the major hysteresis loop does not correspond to the maximum of the SFD in the FORC distribution (see Table 1), showing the 1), showing the differences between the coercive field of the array and the switching field of the differences between the coercive field of the array and the switching field of the nanowires. nanowires.
Figure 4. Experimental FORC diagrams of Co (a) and Ni (b) segment arrays measured along the Figure 4. Experimental FORC diagrams of Co (a) and Ni (b) segment arrays measured along the nanowires axis. Simulated FORC diagrams for Co nanowire array with fixed experimental values of nanowires axis. Simulated FORC diagrams for Co nanowire array with fixed experimental values of SFD and α, for β = 0 Oe (c) and β = −200 Oe (d). SFD and α, for β = 0 Oe (c) and β = −200 Oe (d). Table 1. Values of the coercive field, HC, extracted from the major hysteresis loop and mean switching field, MFI at saturation (α) and standard deviation (σ) of the SFD extracted from FORC for Co and Ni nanowire segment arrays.
Table 1. Values of the coercive field, HC , extracted from the major hysteresis loop and mean switching 𝐅𝐎𝐑𝐂(σ) of the SFD extracted from FORC for Co and Ni field, MFI at Sample saturation (α) and standard deviation 𝛂(𝐎𝐞) 𝝈(𝑶𝒆) 𝐇𝐂𝐇𝐋 (𝐎𝐞) 𝐇𝐒𝐖 (𝐎𝐞) Co arrays. 1296 1175 −2350 185 nanowire segment Ni
884
Sample
HHL C (Oe)
850
HFORC SW (Oe)
−600
α(Oe)
135
σ(Oe)
Moreover, the experimental diagram shown by the Ni nanowire array corresponds to the soCo the Preisach 1296model, such1175 −2350 185 called “T-shape”. Under diagram has been wrongly proposed in the past to Ni 884 850 − 600 135which, in addition, be a consequence of a second magnetically harder phase present in the sample would not be interacting. However, the existence of such magnetic phase with different switching fields is hardly understood (up to 2000 Oe), while the main distribution is around 900 Oe. Moreover, the experimental diagram shown by the Ni nanowire array corresponds to the so-called Furthermore, both crystalline structure and morphology of the electrodeposited Ni nanowires are “T-shape”. Under the Preisach model, such diagram has been wrongly proposed in the past to be a very homogeneous, which would reduce the probability to find different magnetic phases or consequence of a second magnetically present theThat sample in addition, not nanowires having such a differenceharder in theirphase switching fieldsin [31]. shapewhich, of the FORC diagramwould of be interacting. However, theexplained existenceinofdetail suchbymagnetic with different fields is hardly Ni nanowire array was Dobrota etphase al. [32], concluding thatswitching it is a consequence of a nonlinear dependence of the field with is respect to the magnetization, understood (up to 2000 Oe), while theinteraction main distribution around 900normalized Oe. Furthermore, both i.e., crystalline the value of the interaction among wires depends on the magnetization state of the sample [33,34]. structure and morphology of the electrodeposited Ni nanowires are very homogeneous, which would Nevertheless, by comparing the FORC diagrams of Ni and Co nanowire arrays, the stronger reduce the probability to find different magnetic phases or nanowires having such a difference in magnetostatic interaction among Co nanowires is evidenced again, as pointed out by the higher value their switching fields [31]. That shape of the FORC diagram of Nielement nanowire arrayarrays was (Table explained in of α in Table 1. As a summary of the magnetic behavior of the single nanowire detail by et al. concluding is Ni a consequence of a magnetostatic nonlinear dependence 1),Dobrota Co segments are[32], magnetically harderthat thanitthe ones, with stronger interactions of the due field to thewith higherrespect saturation magnetization of Co (162.7 emu/g) with Ni (57.5 emu/g) [35]. interaction to the normalized magnetization, i.e., respect the value of the interaction among
wires depends on the magnetization state of the sample [33,34]. Nevertheless, by comparing the FORC diagrams of Ni and Co nanowire arrays, the stronger magnetostatic interaction among Co nanowires is evidenced again, as pointed out by the higher value of α in Table 1. As a summary of the magnetic
Nanomaterials 2018, 8, 548
8 of 15
behavior of the single element nanowire arrays (Table 1), Co segments are magnetically harder than the Ni ones, with stronger magnetostatic interactions due to the higher saturation magnetization of Co (162.7 emu/g) with respect Ni (57.5 emu/g) [35]. 3.2. Coupled and Decoupled Ni/Co Bi-Magnetic Systems Nanomaterials 2018, 8, x FOR PEER REVIEW
8 of 15
As presented in the previous sections, the easy magnetization axis of the nanowire arrays lies parallel3.2. to the nanowires longNi/Co axis.Bi-magnetic Therefore, the following studies focused on that direction along Coupled and Decoupled Systems which the hysteretic magnetization switching processes take place.axis In of Figure 5a, the arrays major lies hysteresis As presented in the previous sections, the easy magnetization the nanowire loops ofparallel Ni and Conanowires nanowire arrays are numerically superimposed inona that proportion 1:1 and are to the long axis. Therefore, the following studies focused direction along which thethe hysteretic magnetization switchingof processes take place.Ni/Co In Figure 5a, the major hysteresis compared with non-interacting segments the decoupled nanowire array. As expected, loops of Ni and Co nanowire are numerically in a proportion and are the resulting magnetic behavior of arrays the decoupled Ni/Cosuperimposed sample corresponds to the1:1 superposition of compared with the non-interacting segments of the decoupled Ni/Co nanowire array. As expected, the independent hysteresis loops of Ni and Co nanowires, reproducing a system where there is no the resulting magnetic behavior of the decoupled Ni/Co sample corresponds to the superposition of magnetic interaction between segments of each different element. Paying attention to the descendant the independent hysteresis loops of Ni and Co nanowires, reproducing a system where there is no branches of the hysteresis loops, itsegments is possible to different distinguish the Paying contribution type of segment. magnetic interaction between of each element. attentionoftoeach the descendant First, asbranches the magnetic field is reduced the decoupled sample, the magnetization down at the of the hysteresis loops, it isinpossible to distinguish the contribution of each typedrops of segment. First,as as the the magnetic field isarray reduced in the decoupled sample, decrease the magnetization drops down at the same value Co nanowire does it (Point I). Further of the magnetic field provokes same value as the Co nanowire array does it (Point I). Further decrease of the magnetic field provokes a second drop of the magnetization corresponding to the magnetization switching of Ni segments a second drop of the magnetization corresponding to the magnetization switching of Ni segments (Point II). The hysteresis loop of the coupled Ni/Co sample is compared with the hysteresis loop (Point II). The hysteresis loop of the coupled Ni/Co sample is compared with the hysteresis loop of of the decoupled sample ininthe thedifferences differences clearly evidenced. 5c is the decoupled sample theFigure Figure 5b, 5b, where where the are are clearly evidenced. Figure Figure 5c is anotheranother representation of the differences abovesamples samples emphasizing onhysteresis. their hysteresis. representation of the differencesamong among the the above emphasizing on their Such representation consists on the plot thehysteresis hysteresis loop at different values of the reduced Such representation consists on the plot ofofthe loopwidth width at different values of the reduced magnetization. theofcase the decoupled Ni/Co sample, hysteresisofofeach eachcomposition composition is magnetization. In the In case theofdecoupled Ni/Co sample, thethe hysteresis is clearly clearly evidenced, while, in the case of coupled Ni/Co nanowire array, such differences are evidenced, while, in the case of coupled Ni/Co nanowire array, such differences are suppressed and the suppressed and the overall magnetic hysteresis is close to that shown by the Ni nanowire array. As a overall magnetic hysteresis is close to that shown by the Ni nanowire array. As a first approximation, first approximation, the hysteresis and the tilt of the coupled Ni/Co hysteresis loop resembles to the the hysteresis and the tilt ofthe theswitching coupledfield Ni/Co hysteresis the equivalent equivalent system with of a Ni nanowire loop array resembles with the MFItostrength presented system with thebyswitching the Co one.field of a Ni nanowire array with the MFI strength presented by the Co one.
Figure 5. (a) Numerical superposition of the Co and Ni room temperature hysteresis loops (red open Figure 5. (a) Numerical superposition of the Co and Ni room temperature hysteresis loops (red open symbols) compared to the decoupled Ni/Co hysteresis loop (black). (b) Comparison between the symbols) compared to the decoupled Ni/Co hysteresis loop (black). (b) Comparison between the decoupled (black) and coupled (yellow) hysteresis loops. (c) Comparison of the hysteresis shape plots decoupled (black) and coupled hysteresis loops. (c) Comparison of the hysteresis shape plots (width of hysteresis loops at(yellow) different values of magnetization. (width of hysteresis loops at different values of magnetization.
Nanomaterials 2018, 8, 548
9 of 15
To obtain more detailed magnetic information of such magnetostatically interacting nanowire arrays, FORC analyses were also compared. The FORC diagram of a decoupled Ni/Co nanowire array Nanomaterials 2018, 8, x FOR PEER REVIEW 9 of 15 corresponds again to the superposition of the individual FORC diagrams of Ni and Co nanowire arrays, obtain6a. more detailed magnetic information of such magnetostatically interacting nanowireand the as shown inTo Figure However, the differences between the coupled Ni/Co FORC diagram FORCout analyses were also compared. The FORC ofinto a decoupled Ni/Co nanowire array previousarrays, one point that some additional effect mustdiagram be taken account. As the unique difference corresponds again to the superposition of the individual FORC diagrams of Ni and Co nanowire between both samples lies on the fact that the nanowires segments are in contact, it is reasonable to arrays, as shown in Figure 6a. However, the differences between the coupled Ni/Co FORC diagram think that coupling orout magnetostatic interactions at the interface between nanowire andexchange the previous one point that some additional effect must be taken into account. Asthe the two unique segments is the responsible such alies complex a first approximation, difference between bothofsamples on the FORC fact thatdistribution. the nanowiresAs segments are in contact, it isone can think that exchange coupling or magnetostatic interactions at one the interface between the considerreasonable two maintomechanisms for the magnetization reversal process of of the Ni/Co bi-magnetic twoOn nanowire is the responsible complex are FORC distribution. Asmagnetization a first nanowire. the one segments hand, if the switching fieldsofofsuch bothasegments close enough, the can consider two main mechanisms for the magnetization reversal process of one reversalapproximation, of the whole one nanowire could be carried out in a single step through injection and propagation of the Ni/Co bi-magnetic nanowire. On the one hand, if the switching fields of both segments are of a domain wall from the magnetically softer to the harder segment to avoid an unstable antiparallel close enough, the magnetization reversal of the whole nanowire could be carried out in a single step configuration of the magnetic moments. case would result in a one-step magnetization through injection and propagation of This a domain wall from the magnetically softer to the harderreversal of the Ni/Co nanowire with an effective switching field of dominated bymoments. the magnetically softer phase, segment to avoid an unstable antiparallel configuration the magnetic This case would result in a in one-step magnetization reversal the Ni/Co nanowirewould with anbe effective switching field i.e., Ni segment this case, whilst the meanof field interaction increased by the segment dominated by the magnetically softer phase, i.e., Ni segment in this case, whilst the mean field having higher saturation magnetization (Co segment in this case) [36]. On the other hand, if the interaction would be increased by the segment having higher saturation magnetization (Co segment switching fields of each kind of segments are different enough, the magnetization reversal would not in this case) [36]. On the other hand, if the switching fields of each kind of segments are different be carried out in step, which would reflected with distributions enough, thea single magnetization reversal wouldbe not be carried outtwo in adifferentiated single step, which would be in the FORC, as observed 6b. reflected with in twoFigure differentiated distributions in the FORC, as observed in Figure 6b.
Figure 6. Parallel FORC diagrams, measured along the easy magnetization axis, in decoupled (a) and
Figure 6. Parallel FORC diagrams, measured along the easy magnetization axis, in decoupled (a) and coupled (b) Ni/Co nanowire arrays. coupled (b) Ni/Co nanowire arrays. Since the difference between the switching fields of Ni and Co segments appears to be the main parameter that controls which kind of magnetization reversal process is promoted, the different Since the difference between the switching fields of Ni and Co segments appears to be the main temperature dependence of the switching fields in both cases can be used to tune the magnetization parameter that controls which kind of magnetization reversal process is promoted, the different reversal of the Co/Ni bi-segmented nanowire. In Figure 7a, the parallel hysteresis loops measured at temperature dependence ofpoint the switching fields both casesfield canand be used to tune thesegments. magnetization different temperatures out the increase of in both coercive remanence in Co reversalThis of the Co/Ni bi-segmented 7a,phenomena. the parallelFirst, hysteresis loops measured at effect has been associated in nanowire. the literatureIntoFigure two main the magnetostriction of temperatures Co and the different coefficients between the and metallic nanowiresinand different point thermal out the expansion increase of both coercive field remanence Co the segments. ceramic alumina membrane, could induce mechanical stresses on the nanowires their This effect has been associated in the literature to two main phenomena. First, theaffecting magnetostriction of effective magnetic anisotropy [37,38]. On the other hand, the magnetocrystalline anisotropy of the Co and the different thermal expansion coefficients between the metallic nanowires and the ceramic hcp crystalline phase of Co is reduced with increasing temperatures. The Co nanowires fabricated for aluminathis membrane, mechanical stresses affecting their effective work, whosecould detailsinduce are shown in Section 2.1, presenton thethe hcp nanowires crystalline structure which its cmagnetic [37,38]. On thenanowire other hand, the magnetocrystalline axisanisotropy points perpendicular to the long axis. When increasing the anisotropy temperature, of thethe hcp crystalline phase of Co isanisotropy reduced with increasing temperatures. The Co becomes nanowires fabricated magnetocrystalline of Co-hcp is reduced, the shape anisotropy more dominant for this and thus an increase of bothincoercive remanence along the parallel directionwhich to the nanowire work, whose details are shown Sectionfield 2.1,and present the hcp crystalline structure its c-axis points axis is expected [39]. In terms of switching field, Ni segments show almost negligible temperature perpendicular to the nanowire long axis. When increasing the temperature, the magnetocrystalline
anisotropy of Co-hcp is reduced, the shape anisotropy becomes more dominant and thus an increase of both coercive field and remanence along the parallel direction to the nanowire axis is expected [39]. In terms of switching field, Ni segments show almost negligible temperature dependence in comparison
Nanomaterials 2018, 8, 548 Nanomaterials 2018, 8, x FOR PEER REVIEW
10 of 15 10 of 15
with Co ones, resulting in an increase of the difference between the switching field of both segments as dependence in comparison with Co ones, in antemperatures, increase of thethe difference between the the temperature increases. Furthermore, withresulting increasing two-step magnetization switching field of both segments as the temperature increases. Furthermore, with increasing process should become more dominant while, for lower temperature, the magnetization reversal of temperatures, the two-step magnetization process should become more dominant while, for lower coupledtemperature, Ni/Co nanowires should be moreofhomogeneous since their respective fields are the magnetization reversal coupled Ni/Co nanowires should be moreswitching homogeneous closer tosince eachtheir other. respective switching fields are closer to each other.
Figure 7. Parallel hysteresis loops measured at different temperatures ranging from 50 K to 400 K for
Figure 7. Parallel hysteresis loops measured at different temperatures ranging from 50 K to 400 K for Co (a) and Ni (b) nanowire arrays. Co (a) and Ni (b) nanowire arrays. Figure 8 shows the proof of such switching field temperature dependence. At the highest temperature measured (400 of K),such a veryswitching complex FORC is observed where three clearhighest Figure 8 shows the proof field distribution temperature dependence. At the distributions can be seen (marked as I, II and III in Figure 8a), indicating a potential two-step temperature measured (400 K), a very complex FORC distribution is observed where three clear magnetization reversal process. While the temperature is decreasing, the FORC diagram evolves distributions can be seen (marked asthe I, coupled II and Ni/Co III innanowires Figure 8a), indicating a potential two-step towards a single distribution so that behave as a homogenous nanowire magnetization reversal While the temperature decreasing, the FORC diagram evolves and the reversal of process. the magnetization is carried out in onlyisone-step, i.e., a single and homogeneous distribution. However, of these two magnetization behaviors nanowire is towardsFORC a single distribution so thatthe thecoexistence coupled Ni/Co nanowires behave asreversal a homogenous over the wholeismeasured temperature range. Considering onlyand the homogeneous parallel and the expected reversal to of occur the magnetization carried out in only one-step, i.e., a single hysteresis loops of coupled Ni/Co bisegmented nanowire arrays shown in Figure 8b, this effect is FORC distribution. However, the coexistence of these two magnetization reversal behaviors is expected hardly appreciable. However, taking a closer look, one can observe a more pronounced twist of the to occur curve over close the whole measured temperature range. Considering only the parallel hysteresis loops of to remanence at 400 K, thus indicating a more evident presence of a biphase contribution. coupledInNi/Co arrays in Figure 8b, effect is appreciable. Figure bisegmented 8c, the coercivenanowire field of the three shown FORC distributions canthis be tracked ashardly a function of However, taking a closer look, one can observe a more switching pronounced curvetoclose to remanence temperature. First, Distribution I shows an intrinsic fieldtwist whichofisthe identical the coercive exhibited by the hysteresis loops presence of Ni segments. This distribution could reflect the switching at 400 K,field thus indicating a more evident of a biphase contribution. In Figure 8c, theofcoercive the Ni segment, as they would be isolated, however it would be unexpected that the Ni field of the three FORC distributions can be tracked as a function of temperature. First,segments Distribution I switching field is not modified with the coupling between Co segments. Moreover, as shown shows an intrinsic switching field which is identical to the coercive field exhibited by the hysteresis previously, strongly coupled segments would switch their magnetization in a single step at a loops ofswitching Ni segments. This distribution could reflectsofter the switching of the Ni segment, as theyI would field dominated by the magnetically segment, suggesting that Distribution be isolated, however it would be unexpected the Ni segmentsNi/Co switching field is not corresponds to the magnetization reversal of that the whole bi-segmented nanowire driven by modified Ni bysegments. decreasing Moreover, the reversalas field, the FORC diagram shows two different with thesegment. couplingFurthermore, between Co shown previously, strongly coupled segments distributions, i.e., a reversal magnetization carried in two steps (Distributions II and III).magnetically When would switch their magnetization in a single step at out a switching field dominated by the the reversal field is further decreased, some Ni segments reverse their magnetization staying in an softer segment, suggesting that Distribution I corresponds to the magnetization reversal of the whole antiparallel configuration compared to their respective Co segments. If at this point the magnetic field bi-segmented Ni/Co nanowire driven by Ni segment. Furthermore, by decreasing the reversal field, is increased back to saturation magnetization, the portion of Ni segments that are aligned antiparallel, the FORC shows different distributions, i.e., reversal magnetization carried willdiagram switch back beforetwo its intrinsic switching field would beareached, due to the coupling with the out in two steps and III). When theIII) reversal field further decreased, some Ni segments Co(Distributions segments at the II interface (Distribution that forces theisNi/Co segments to be magnetized in parallel. This effect willstaying show an switchingconfiguration field of such Ni segments to lower than their reverse their magnetization ineffective an antiparallel compared their respective Co intrinsic one due to that interface coupling. This behavior is also visible in the magnetization reversal segments. If at this point the magnetic field is increased back to saturation magnetization, the portion of Co segments, which are represented by the Distribution II in the FORC diagrams of Figure 8a. The of Ni segments that are aligned antiparallel, will switch back before its intrinsic switching field would switching field at high temperatures ascribed to the Distribution II is noticeably higher than the one be reached, due coupling with segments at the interface III)indicating that forces the shown byto Nithe or Distribution I, but the still Co lower than the coercive field of Co(Distribution nanowire arrays,
Ni/Co segments to be magnetized in parallel. This effect will show an effective switching field of such Ni segments lower than their intrinsic one due to that interface coupling. This behavior is also visible in the magnetization reversal of Co segments, which are represented by the Distribution II in the FORC diagrams of Figure 8a. The switching field at high temperatures ascribed to the Distribution II is noticeably higher than the one shown by Ni or Distribution I, but still lower than the coercive field of Co nanowire arrays, indicating that the Co segments (as the Ni ones) tend to leave the antiparallel
Nanomaterials 2018, 8, 548 Nanomaterials 2018, 8, x FOR PEER REVIEW
11 of 15 11 of 15
configuration at lower switching fields. As the temperature is decreased, Distributions I and II collapse that the Co segments (as the Ni ones) tend to leave the antiparallel configuration at lower switching at the same coercive field, although around room temperature both distributions can still be clearly fields. As the temperature is decreased, Distributions I and II collapse at the same coercive field, differentiated in the FORC diagram. Further decrease of the temperature results in the disappearing of although around room temperature both distributions can still be clearly differentiated in the FORC both previous I andofII,the being substituted for ainsingle one, indicating a single step reversal diagram.Distributions Further decrease temperature results the disappearing of both previous of the magnetization. In addition, at 400 K the relative intensity of Distributions II and Distributions I and II, being substituted for a single one, indicating a single step reversal ofIII theis higher magnetization. In addition, at 400 K the relative intensity of Distributions II and III is higher than the than the intensity of Distribution I, indicating that the magnetization reversal of the bi-segmented intensity of Distribution indicating that the process, magnetization reversal of the bi-segmented nanowires nanowires is mainly carried I,out in a two-step while decreasing the temperature the intensity is mainly carried out in a two-step process, while decreasing the temperature the intensity of the of the Distribution III decreases while the corresponding one of Distribution I increases accordingly. Distribution III decreases while the corresponding one of Distribution I increases accordingly. This This demonstrates once again againthat thatreducing reducing temperature, the amount of bi-segmented demonstrates once thethe temperature, the amount of bi-segmented Ni/Co Ni/Co nanowires that reverse their magnetization in a single step increase. nanowires that reverse their magnetization in a single step increase.
8. (a) Parallel FORC diagrams measured at different temperatures ranging from 400 K down Figure Figure 8. (a) Parallel FORC diagrams measured at different temperatures ranging from 400 K down to to 50 K. (b) Parallel hysteresis loops of coupled Ni/Co bisegmented sample measured at 300 K (black) 50 K. (b) Parallel hysteresis loops of coupled Ni/Co bisegmented sample measured at 300 K (black) and 400 K (red). (c) Summarizes the coercive field of the Ni and Co hysteresis loops, and the main and 400 K (red).field (c)ofSummarizes the coercive field of the Nimeasured and Co temperature hysteresis loops, and50the main switching the different Ni/Co FORC distributions in the range from switching field of the different Ni/Co FORC distributions in the measured temperature range from K to 400 K. 50 K to 400 K.
Finally, to validate the interpretation of the experimental results, the phenomenological contribution of a two-step magnetization process was implemented and simulation results are shown in Figure For that purpose, two main contributions were considered:results, a homogeneous nanowire Finally, to9.validate the interpretation of the experimental the phenomenological behavior a two-step magnetizationprocess reversalwas contribution. In the and first simulation case, a switching field contribution of and a two-step magnetization implemented results are shown distribution whose main switching field is close to the Ni one (𝐻 / = 850 𝑂𝑒) and an in Figure 9. For that purpose, two main contributions were considered: a homogeneous nanowire averaged mean field interaction value of α = −1400 Oe were fixed. In the case of the two-step behavior and a two-step magnetization reversal contribution. In the first case, a switching field = magnetization process, the switching field of Co coupled segments was chosen as 𝐻 , homogeneous distribution is closeA to Ni one (Hrespectively. =both 850cases, Oe) and an Ni/Co 1100 𝑂𝑒whose and 𝐻 main =switching 1000 𝑂𝑒 forfield Simulations andthe B (Figure 9a,b), In , averaged mean field interaction valuetoofswitch α = 100 −1400 Oe were In both the are case of the two-step the coupled Ni segment was imposed Oe before the Cofixed. one when magnetized coupled in parallel.process, In contrast, the firstfield orderof reversal curve starting from the magnetization thealong switching Co coupled segments wasantiparallel chosen asconfiguration, HCo,A = 1100 Oe , , coupled the Ni segment switches at 𝐻 = 200 𝑂𝑒 and 𝐻 = 300 𝑂𝑒 for Simulations A
and HCo,B = 1000 Oe for Simulations A and B (Figure 9a,b), respectively. In both cases, the coupled Ni segment was imposed to switch 100 Oe before the Co one when both are magnetized in parallel. In contrast, along the first order reversal curve starting from the antiparallel configuration, the Ni FORC,coupled FORC,coupled segment switches at HNi = 200 Oe and HNi = 300 Oe for Simulations A and B, respectively, leading in an effective switching field of 600 Oe in both cases. Although simulation
Nanomaterials 2018, 2018, 8, 8, 548 x FOR PEER REVIEW Nanomaterials
12 15 12 of of 15
and B, respectively, leading in an effective switching field of 600 Oe in both cases. Although results are in goodare agreement with the experimental data, providing the tools to simulation results in good agreement with the experimental data, providing thedetect tools two-step to detect magnetization processes, the assumption of a mean field interaction model is an important limitation two-step magnetization processes, the assumption of a mean field interaction model is an important to extract quantitative information from such diagrams. limitation to extract quantitative information from such diagrams.
Figure 9. 9. Simulated Simulated FORC FORC diagrams diagrams of of aacoupled coupledNi/Co Ni/Co bisegmented bisegmented nanowire nanowire array array for for the the cases cases Figure where the switching field of Co is much greater a) or slightly larger b) than the Ni switching field. where the switching field of Co is much greater a) or slightly larger b) than the Ni switching field.
4. Conclusions Conclusions 4. Ni/Co bi-segmented bi-segmented nanowire arrays were successfully and characterized. magnetically Ni/Co nanowire arrays were successfully fabricated fabricated and magnetically characterized. To extract the contribution of each nanowire composition to the magnetization To extract the contribution of each nanowire composition to the magnetization reversal processreversal as well process as well as the different magnetic interactions between them, Ni and Co nanowire arrays, and as the different magnetic interactions between them, Ni and Co nanowire arrays, and both stacked both stacked nanowire arrays (decoupled Ni/Co sample), were also investigated. From hysteresis nanowire arrays (decoupled Ni/Co sample), were also investigated. From hysteresis loops measured at loopstemperature, measured atboth room temperature, both Ni and Co nanowires a well-defined easy room Ni and Co nanowires present a well-defined easy present magnetization axis parallel to magnetization axis parallel to the nanowire. Furthermore, Ni nanowires have been found to be the nanowire. Furthermore, Ni nanowires have been found to be magnetically softer than the Co ones, magnetically softer than of theeach Co ferromagnetic ones, enabling the identification ferromagnetic segment enabling the identification segment contributionof to each the total magnetization when contribution to the total magnetization when both types of nanowire segments are measured both types of nanowire segments are measured together. As a first result, the magnetic behavior of the together. As a firstsample result, corresponds the magnetictobehavior of the decoupled Ni/Co sample corresponds to the decoupled Ni/Co the superposition of the respective Ni and Co independent superposition of the respective Ni and Co independent magnetization reversal behaviors. Moreover, magnetization reversal behaviors. Moreover, when the nanowires are in contact at one end, hysteresis when and the nanowires are indistributions contact at one end, hysteresis loops and thus the FORC are loops thus the FORC are modified drastically with respect to thedistributions not-interacting modified drastically with respect to the not-interacting case, which suggests that some additional case, which suggests that some additional effects are provided by the magnetic coupling at the effects are provided by the magnetic at the interface between Ni/Co InNi/Co a first interface between Ni/Co segments. In acoupling first approximation, the hysteresis loop ofsegments. the coupled approximation, the hysteresis loop of the coupled Ni/Co nanowire array suggests that the effective nanowire array suggests that the effective magnetization reversal process would be driven by the magnetization reversal process besegment, driven by softer ferromagnetic in this case Ni softer ferromagnetic phase, in thiswould case Ni as the far as the overall magneticphase, hysteresis is reduced segment, as far as the magnetic hysteresis is reduced values closer to thetwo Ni nanowire taking values closer tooverall the Ni nanowire array. However, eventaking from the hysteresis loop, different array. However, even from the hysteresis loop, two different magnetization reversal process can be magnetization reversal process can be distinguished. It results interesting the possibility to associate distinguished. It results interesting the possibility to associate one of these processes with one of these processes with a magnetization reversal taking place in two steps and what would bea magnetization reversal taking place in two steps and be the FORC signature this kind the FORC signature of this kind of phenomenon. The what FORCwould analysis made on this sampleofevidences of phenomenon. The FORC analysis made on this sample evidences the magnetic coupling between the magnetic coupling between the Ni and Co segments, whose diagram has been associated with Ni and segments, whose diagram has been associated withina each system in which thenanowire antiparallel athe system inCo which the antiparallel magnetization configuration bi-segmented is magnetization configuration in each bi-segmented nanowire is possible. Such antiparallel possible. Such antiparallel configuration as well as the magnetization reversal behavior associated configuration well as on the the magnetization reversal behavior associated depends strongly on the to it depends as strongly difference between the switching fieldtoofitboth nanowire segment difference between thetwo switching field of both nanowire segment compositions, finding different compositions, finding different FORC distributions with a critical switching field two difference of FORC distributions with a critical switching field difference of around 500 Oe. As this difference around 500 Oe. As this difference increases, the probability to achieve the antiparallel configuration increases, the probability the antiparallel configuration also increases, observed in the also increases, as observedtoinachieve the temperature dependence of the FORC analysis inas the bi-segmented temperature dependence of the FORC analysis in the bi-segmented array. The main Ni/Co nanowire array. The main result from this work lies in the Ni/Co FORC nanowire fingerprint identification result from this work lies in the FORC fingerprint identification characteristic of a two-step reversal characteristic of a two-step reversal magnetization of segmented nanowires. Such a result could shed magnetization of segmented nanowires. a result of could shed light on the magnetic light on the magnetic characterization and theSuch understanding vast nanostructure architectures that characterization and the understanding of vast nanostructure architectures that are, nowadays, under are, nowadays, under investigation. investigation.
Nanomaterials 2018, 8, 548
13 of 15
Author Contributions: Conceptualization: J.G.F. and V.M.d.l.P.; Funding acquisition: V.M.d.l.P. and K.N.; Investigation: J.G.F. and V.V.M.; Writing: J.G.F.; Writing—Review and Editing: A.T. and V.M.d.l.P.; and Supervision V.M.d.l.P. Funding: This work was supported by the Ministerio de Economia (Spain) under grants MINECO-17-MAT201676824-C3-3-R. Acknowledgments: The scientific support from the SCTs of the University of Oviedo is acknowledged. Conflicts of Interest: The authors declare no conflict of interest.
References 1.
2. 3.
4.
5. 6.
7.
8.
9.
10.
11. 12.
13. 14. 15.
Béron, F.; Pirota, K.R.; Vega, V.; Prida, V.M.; Fernández, A.; Hernando, B.; Knobel, M. An effective method to probe local magnetostatic properties in a nanometric FePd antidot array. New J. Phys. 2011, 13, 013035. [CrossRef] Sousa, C.T.; Leitao, D.C.; Proenca, M.P.; Ventura, J.; Pereira, A.M.; Araujo, J.P. Nanoporous alumina as templates for multifunctional applications. Appl. Phys. Rev. 2014, 1, 031102. [CrossRef] Prida, V.M.; Vega, V.; García, J.; Iglesias, L.; Hernando, B.; Mínguez Bacho, I. Electrochemical Methods for Template-Assisted Synthesis of Nanostructured Materials; Vázquez, M., Ed.; Woodhead Publishing: Cambridge, UK, 2015; ISBN 978-0-08-100164-6. Prida, V.M.; García, J.; Hernando, B.; Bran, C.; Vivas, L.G.; Vázquez, M. Electrochemical synthesis of magnetic nanowires with controlled geometry and magnetic anisotropy. In Woodhead Publishing Series in Electronic and Optical Materials; Vázquez, M., Ed.; Woodhead Publishing: Cambridge, UK, 2015; pp. 41–104, ISBN 978-0-08-100164-6. Nielsch, K.; Müller, F.; Li, A.-P.; Gösele, U. Uniform Nickel Deposition into Ordered Alumina Pores by Pulsed Electrodeposition. Adv. Mater. 2000, 12, 582–586. [CrossRef] Vázquez, M.; Hernández-Vélez, M.; Pirota, K.; Asenjo, A.; Navas, D.; Velázquez, J.; Vargas, P.; Ramos, C. Arrays of Ni nanowires in alumina membranes: magnetic properties and spatial ordering. Eur. Phys. J. B 2004, 40, 489–497. [CrossRef] Mourachkine, A.; Yazyev, O.V.; Ducati, C.; Ansermet, J.-P. Template Nanowires for Spintronics Applications: Nanomagnet Microwave Resonators Functioning in Zero Applied Magnetic Field. Nano Lett. 2008, 8, 3683–3687. [CrossRef] [PubMed] Salem, M.S.; Sergelius, P.; Corona, R.M.; Escrig, J.; Gorlitz, D.; Nielsch, K. Magnetic properties of cylindrical diameter modulated Ni80Fe20 nanowires: interaction and coercive fields. Nanoscale 2013, 5, 3941–3947. [CrossRef] [PubMed] Prida, V.M.; García, J.; Iglesias, L.; Vega, V.; Görlitz, D.; Nielsch, K.; Barriga-Castro, E.D.; Mendoza-Reséndez, R.; Ponce, A.; Luna, C. Electroplating and magnetostructural characterization of multisegmented Co54Ni46/Co85Ni15 nanowires from single electrochemical bath in anodic alumina templates. Nanoscale Res. Lett. 2013, 8, 263. [CrossRef] [PubMed] Böhnert, T.; Niemann, A.C.; Michel, A.-K.; Bäßler, S.; Gooth, J.; Tóth, B.G.; Neuróhr, K.; Péter, L.; Bakonyi, I.; Vega, V.; et al. Magnetothermopower and magnetoresistance of single Co-Ni/Cu multilayered nanowires. Phys. Rev. B 2014, 90, 165416. [CrossRef] Pereira, A.; Palma, J.L.; Vazquez, M.; Denardin, J.C.; Escrig, J. A soft/hard magnetic nanostructure based on multisegmented CoNi nanowires. Phys. Chem. Chem. Phys. 2015, 17, 5033–5038. [CrossRef] [PubMed] Fernandez-Roldan, J.A.; Perez del Real, R.; Bran, C.; Vazquez, M.; Chubykalo-Fesenko, O. Magnetization pinning in modulated nanowires: From topological protection to the “corkscrew” mechanism. Nanoscale 2018. [CrossRef] [PubMed] Clime, L.; Ciureanu, P.; Yelon, A. Magnetostatic interactions in dense nanowire arrays. J. Magn. Magn. Mater. 2006, 297, 60–70. [CrossRef] Raposo, V.; Garcia, J.M.; González, J.M.; Vázquez, M. Long-range magnetostatic interactions in arrays of nanowires. J. Magn. Magn. Mater. 2000, 222, 227–232. [CrossRef] Roberts, A.P.; Heslop, D.; Zhao, X.; Pike, C.R. Understanding fine magnetic particle systems through use of first-order reversal curve diagrams. Rev. Geophys. 2014, 52, 557–602. [CrossRef]
Nanomaterials 2018, 8, 548
16.
17. 18. 19. 20.
21. 22. 23. 24.
25. 26. 27. 28. 29. 30. 31. 32.
33. 34. 35.
36.
37.
14 of 15
Ruta, S.; Hovorka, O.; Huang, P.W.; Wang, K.; Ju, G.; Chantrell, R. First order reversal curves and intrinsic parameter determination for magnetic materials; Limitations of hysteron-based approaches in correlated systems. Sci. Rep. 2017. [CrossRef] [PubMed] Pike, C.R.; Roberts, A.P.; Verosub, K.L. Characterizing interactions in fine magnetic particle systems using first order reversal curves. J. Appl. Phys. 1999, 85, 6660–6667. [CrossRef] Stancu, A.; Pike, C.; Stoleriu, L.; Postolache, P.; Cimpoesu, D. Micromagnetic and Preisach analysis of the First Order Reversal Curves (FORC) diagram. J. Appl. Phys. 2003, 93, 6620–6622. [CrossRef] Lee, W.; Park, S.-J. Porous Anodic Aluminum Oxide: Anodization and Templated Synthesis of Functional Nanostructures. Chem. Rev. 2014, 114, 7487–7556. [CrossRef] [PubMed] Beron, F.; Carignan, L.-P.; Menard, D.; Yelon, A. Extracting Individual Properties from Global Behaviour: First-order Reversal Curve Method Applied to Magnetic Nanowire Arrays. In Electrodeposited Nanowires and their Applications; Lupu, N., Ed.; InTech: Rijeka, Croatia, 2010. Wohlfarth, E.P. Relations between Different Modes of Acquisition of the Remanent Magnetization of Ferromagnetic Particles. J. Appl. Phys. 1958, 29, 595–596. [CrossRef] Henkel, O. Remanenzverhalten und Wechselwirkungen in hartmagnetischen Teilchenkollektiven. Phys. Status Solidi 1964, 7, 919–929. [CrossRef] Kelly, P.E.; O’Grady, K.; Mayo, P.I.; Chantrell, R.W. Switching mechanisms in cobalt-phosphorus thin films. IEEE Trans. Magn. 1989, 25, 3881–3883. [CrossRef] Martínez Huerta, M.J.; De La Torre Medina, J.; Piraux, L.; Encinas, A. Self consistent measurement and removal of the dipolar interaction field in magnetic particle assemblies and the determination of their intrinsic switching field distribution. J. Appl. Phys. 2012, 083914. [CrossRef] Richter, H.J. The transition from longitudinal to perpendicular recording. J. Phys. D. Appl. Phys. 2007, 40, R149–R177. [CrossRef] Bertotti, G. Collections of Bistable Units. In Electromagnetism; Bertotti, G., Ed.; Academic Press: San Diego, CA, USA, 1998; pp. 433–477, ISBN 978-0-12-093270-2. Bertotti, G. Hysteresis in Preisach Systems. In Electromagnetism; Bertotti, G., Ed.; Academic Press: San Diego, CA, USA, 1998; pp. 479–506, ISBN 978-0-12-093270-2. Geshev, J.; Schmidt, J.E. Interaction fields evaluation in fine particle systems. IEEE Trans. Magn. 1997, 33, 2504–2508. [CrossRef] Pike, C.R.; Ross, C.A.; Scalettar, R.T.; Zimanyi, G. First-order reversal curve diagram analysis of a perpendicular nickel nanopillar array. Phys. Rev. B 2005, 71, 134407. [CrossRef] Che, X.; Neal Bertram, H. Phenomenology of δM curves and magnetic interactions. J. Magn. Magn. Mater. 1992, 116, 121–127. [CrossRef] Proenca, M.P.; Sousa, C.T.; Ventura, J.; Vazquez, M.; Araujo, J.P. Ni growth inside ordered arrays of alumina nanopores: Enhancing the deposition rate. Electrochim. Acta 2012, 72, 215–221. [CrossRef] Dobrotă, C.-I.; Stancu, A. Tracking the individual magnetic wires’ switchings in ferromagnetic nanowire arrays using the first-order reversal curves (FORC) diagram method. Phys. B Condens. Matter 2015, 457, 280–286. [CrossRef] Dobrotă, C.-I.; Stancu, A. Mean field model for ferromagnetic nanowire arrays based on a mechanical analogy. J. Phys. Condens. Matter 2013, 25, 035302. [CrossRef] [PubMed] Dobrotă, C.-I.; Stancu, A. What does a first-order reversal curve diagram really mean? A study case: Array of ferromagnetic nanowires. J. Appl. Phys. 2013, 113, 043928. [CrossRef] Schwerdt, J.I.; Goya, G.F.; Pilar Calatayud, M.; Herenu, C.B.; Reggiani, P.C.; Goya, R.G. Magnetic Field-Assisted Gene Delivery: Achievements and Therapeutic Potential. Curr. Gene Ther. 2012, 12, 116–126. [CrossRef] [PubMed] Navas, D.; Torrejon, J.; Béron, F.; Redondo, C.; Batallan, F.; Toperverg, B.P.; Devishvili, A.; Sierra, B.; Castaño, F.; Pirota, K.R.; et al. Magnetization reversal and exchange bias effects in hard/soft ferromagnetic bilayers with orthogonal anisotropies. New J. Phys. 2012, 14, 113001. [CrossRef] Navas, D.; Pirota, K.R.; Mendoza Zelis, P.; Velazquez, D.; Ross, C.A.; Vazquez, M. Effects of the magnetoelastic anisotropy in Ni nanowire arrays. J. Appl. Phys. 2008, 103, 07D523. [CrossRef]
Nanomaterials 2018, 8, 548
38.
39.
15 of 15
Vázquez, M.; Pirota, K.; Torrejón, J.; Navas, D.; Hernández-Vélez, M. Magnetic behaviour of densely packed hexagonal arrays of Ni nanowires: Influence of geometric characteristics. J. Magn. Magn. Mater. 2005, 294, 174–181. [CrossRef] Srivastav, A.K.; Shekhar, R. Crystal anisotropy induced temperature dependent magnetization in cobalt nanowires electrodeposited within alumina template. J. Magn. Magn. Mater. 2014, 349, 21–26. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
