A Review of Thin-Film Magnetoelastic Materials for Magnetoelectric Applications

Review A Review of Thin-Film Magnetoelastic Materials for Magnetoelectric Applications

1, 1, 1 1,2 1 Xianfeng Liang †, Cunzheng Dong † , Huaihao Chen , Jiawei Wang , Yuyi Wei , Mohsen Zaeimbashi 1, Yifan He 1, Alexei Matyushov 1, Changxing Sun 1 and Nianxiang Sun 1,*

1 Department of Electrical and Computer Engineering, Northeastern University, Boston, MA 02115, USA; [email protected] (X.L.); [email protected] (C.D.); [email protected] (H.C.); [email protected] (J.W.); [email protected] (Y.W.); [email protected] (M.Z.); [email protected] (Y.H.); [email protected] (A.M.); [email protected] (C.S.) 2 College of Science, Zhejiang University of Technology, Hangzhou 310023, China * Correspondence: [email protected]; Tel.: +1-(617)-373-3351 Xianfeng Liang and Cunzheng Dong contributed equally to this work. † Received: 26 January 2020; Accepted: 7 March 2020; Published: 10 March 2020

Abstract: Since the revival of multiferroic laminates with giant magnetoelectric (ME) coefficients, a variety of multifunctional ME devices, such as sensor, inductor, filter, antenna etc. have been developed. Magnetoelastic materials, which couple the magnetization and strain together, have recently attracted ever-increasing attention due to their key roles in ME applications. This review starts with a brief introduction to the early research efforts in the field of multiferroic materials and moves to the recent work on magnetoelectric coupling and their applications based on both bulk and thin-film materials. This is followed by sections summarizing historical works and solving the challenges specific to the fabrication and characterization of magnetoelastic materials with large magnetostriction constants. After presenting the magnetostrictive thin films and their static and dynamic properties, we review micro-electromechanical systems (MEMS) and bulk devices utilizing ME effect. Finally, some open questions and future application directions where the community could head for magnetoelastic materials will be discussed.

Keywords: magnetoelastic materials; magnetostriction; magnetoelectric devices; thin ﬁlms

1. Introduction

1.1. Multiferroic Materials Multiferroic materials are the materials that inherently exhibit two or more ferroic properties, such as ferroelectricity, ferromagnetism and ferroelasticity, etc. More recently, both single phase multiferroic materials and multiferroic composites have attracted intense interests due to the realization of strong magnetoelectric (ME) coupling, i.e., the control of electric polarization (P) by applying magnetic field (H) (direct ME effect), or the manipulation of magnetization (M) through electric field (E) (converse ME effect) [1–8]. Exciting progress has been made on novel multiferroic materials and multifunctional devices because of their high-performance ME coupling [9–14]. Based on the operating mechanisms for controlling different orders, multiferroic devices can be classified as the following groups listed in Table1. Applications, such as sensors, energy harvesters, etc. [ 15–25] have been developed according to the direct ME coupling, while converse ME coupling has been utilized to design voltage tunable devices including ME random access memory (MERAM), inductors, etc. [26–36]. Besides, another promising topic that has received the significant research interests during recent years is mechanically actuated ME antennas, which exploit both direct and converse ME couplings [37–46]. The combined high

Sensors 2020, 20, 1532; doi:10.3390/s20051532 www.mdpi.com/journal/sensors Sensors 2020, 20, 1532 2 of 27 permeability µ and permittivity ε offer great beneficial in compact RF/microwave devices, for example, miniaturized antennas, etc. [47,48]. Compared to conventional magnetic devices, these electric field tunable multiferroic devices have the advantages of lightweight, low power-consuming, compact, etc. For example, an integrated magnetic inductor based on solenoid structure with FeGaB/Al2O3 multilayer films was reported by Gao et al. [35]. A high quality factor and >100% inductance enhancement compared with that of the same size air core inductor across a wide frequency band of DC-2.5 GHz were achieved. Nan et al. [44] proposed the acoustically actuated ME antennas with a released ferromagnetic/piezoelectric thin-film heterostructures, which could miniaturize the antenna size in 1-2 orders without performance degradation over conventional compact antennas.

Table 1. Categories of various multiferroic devices. Adapted from [4].

ME Coupling Physical Mechanisms ME Devices References Magnetic/current sensors, Direct ME coupling H control of P energy harvesters, gyrators, [15–25] transformers E control of M switching MERAM [26–28] Converse ME coupling Voltage tunable inductors, E control of µ [29–36] ﬁlters, phase shifters Direct and converse Interaction between electric VLF mechanical antennas [37–42] ME coupling and magnetic phases Nanomechanical antennas [43–46] No ME coupling High µ and ε Compact antennas, etc. [47,48] Note: E/H, electric/magnetic ﬁeld; P/M, polarization/magnetization; µ/ε, permeability/permittivity; MERAM, ME random access memory; VLF, very low frequency.

In order to achieve large tunability in multiferroic devices, strong ME coupling is required in the heterostructures. The strength of ME coupling can be described by two coefficients: αDirect = ∂P/∂H (direct ME effect) and αConverse = ∂M/∂E (converse ME effect). Therefore, multiferroic materials are of great importance in determining the performance of these multiferroic devices listed in Table1. The ME effect was proposed by Curie in 1894 [49] and was proved by Landau and Lifshitz on the basis of the crystal symmetry more than 60 years later [50]. Dzyaloshinskii experimentally demonstrated the ME effect in single-phase multiferroic material Cr2O3 and intense research efforts have been made to explore the possibility of achieving ferroelectric and magnetic orders in a single-phase material since then. However, the single-phase materials, such as BiFeO3, BiMnO3, etc. are still suffering from the low Curie temperature and weak ME coupling coefficients [51]. By means of utilizing strain-mediated ME composites, which are composted of ferroelectric and magnetoelectric phases, a giant ME coupling even above room temperature can be obtained. The first work on combining piezoelectric and magnetostrictive effects was presented by Van Suchtelen in 1972 [52]. The reported ME voltage coefficient of BaTiO3/CoFe2O4 is 1-2 orders higher than single-phase multiferroic materials. Later then, a massive research work was carried out on ME composites including both bulk and thin-film materials due to their great potential for multifunctional RF/microwave devices. In comparison with bulk materials, thin-film ME composites have some distinctive advantages such as low interface losses, complementary metal–oxide–semiconductor (CMOS)-compatible fabrication process, etc. Therefore, they are more promising candidates for integrated RF/microwave ME devices.

1.2. Magnetoelastic Materials Owing to the renaissance of multiferroic materials, research works on magnetoelastic materials are increasing year by year as well. The year number of papers published on multiferroic, magnetoelectric and magnetoelastic is shown in Figure1, the data of which is recorded from GOOGLE SCHOLAR. It is clear that the number of papers on multiferroic and magnetoelectric keeps rising since 21st century, which means researchers become more interested in this topic. The number of papers published on magnetoelastic also increases slowly in recent years result from the revival interest in ME effect. The Sensors 2020, 20, x FOR PEER REVIEW 2 of 29 magnetoelastic is shown in Figure 1, the data of which is recorded from GOOGLE SCHOLAR. It is clear thatSensors the number2020, 20, 1532of papers on multiferroic and magnetoelectric keeps rising since 21st century, which3 ofmeans 27 researchers become more interested in this topic. The number of papers published on magnetoelastic also increases slowly in recent years result from the revival interest in ME effect. The selection of appropriate materialsselection plays of appropriate a key role in materials fabricating plays ME a keydevices role inwith fabricating good performance. ME devices The with properties good performance. of different typicalThe propertiespiezoelectric of and different magnetostrictive typical piezoelectric materials and used magnetostrictive for ME devices materials are listed usedin Table for ME2. Pb(Zr,Ti)O devices 3 (PZT)are-based listed in ceramics Table2. are Pb(Zr,Ti)O well-known3 (PZT)-based piezoelectric ceramics materials are well-known that have been piezoelectric widely employed materials that in ME deviceshave due been to widely their high employed piezoelectric in ME devicescoefficients due and to their low highcost. piezoelectricFor the magnetostrictive coefficients andphase, low Galfenol cost. (FeGa)For the and magnetostrictive FeCoSiB with large phase, magnetostriction Galfenol (FeGa) and constants FeCoSiB and with piezomagnetic large magnetostriction coefficients constants have been mostlyand piezomagneticused in thin-film coe MEfficients devices. have There been are mostly a variety used of in properties thin-film ME for devices.both piezoelectric There are aand variety magnetic of phasesproperties that need for both to be piezoelectric considered and when magnetic developing phases novel that need ME to applications, be considered such when as dielectric developing loss, polarization,novel ME applications,piezoelectric suchconstant, as dielectric electromechanical loss, polarization, coupling piezoelectric factor, magnetization, constant, electromechanical coercive magnetic field,coupling magnetostriction, factor, magnetization, etc. coercive magnetic field, magnetostriction, etc.

FigureFigure 1. The 1. The year year number number of ofpapers papers published published on on the the multiferroic, multiferroic, magnetoelectric magnetoelectric and and magnetoelastic.

Table 2. Parameters of typical piezoelectric and magnetostrictive materials. Table 2. Parameters of typical piezoelectric and magnetostrictive materials. Piezoelectric Phase Magnetostrictive Phase Parameters Piezoelectric phase Magnetostrictive phase PZT-5H PZT-8 PMN-0.33PT LiNbO3 AlN film Metaglas Terfenol-D FeGaC FeGaB FeCoSiB 1 Parametersd31,p(pC N− ) PZT265- PZT-37 PMN-1330 1 AlN 2 Terfenol- 1 − − − LiNbO3− − Metaglas FeGaC FeGaB FeCoSiB d33,p(pC N− ) 5H585 8 2250.33PT 2820 21film ~3.5–4 D εr 3400 1000 8200 30 ~10 d31,pQ(pCm 65 1000 100 100000 500 − −265 −37 −1330 −1 −2 158 * λNs(ppm)1) ~30 2000 81.2 70 annealed d d33,p(nm(pC A 1) ~3.5– 50.3 25 121.3 ~88 33,m − 585 225 2820 21 Nµ−1r) 4 45000 10 ~400 T ( C) 193 300 135 1200 >2000 395 650 C휺 ◦ 3400 1000 8200 30 ~10 References풓 [53][54][55][56][57–59][60][8][61][62][63] Qm 65 1000 100 100000 500 Note: d31,p/d33,p, transverse/longitudinal piezoelectric constant; d33,m, longitudinal piezomagentic constant; λs, saturated magnetostriction coefficient; εr, relative dielectric constant; µr, relative permeability; Tc, curie temperature;158* 흀풔(ppm) ~30 2000 81.2 70 Qm, mechanical quality factor; * The value is estimated using Young’s modulus and Poisson ratio of the filmannealed being 100 GPa and 0.3, respectively. d33,m(nm 50.3 25 121.3 ~88 A−1) Magnetoelastic effects refer to the couplings between magnetic and elastic properties of a 흁풓 45000 10 ~400 material, which can be divided into two categories: direct and inverse effects. The best-known TC (℃) 193 300 135 1200 >2000 395 650 direct magnetoelastic effects are magnetostriction and ∆E effects while some of the inverse effects are [57– describedReferences by special[53] terms[54] such[55] as Villari[56] effect and Matteuci[60] eff[8]ect in the[61] literature. [62] The[63] detailed 59] description of magnetoelastic effects are listed in Table3. Magnetostrictive materials have been playing Note: d31,p/d33,p, transverse/longitudinal piezoelectric constant; d33,m, longitudinal piezomagentic constant; an important role in applications ranging from actuators, sensors and energy harvesters. Here, we λs, saturated magnetostriction coefficient; Ɛr, relative dielectric constant; µ r, relative permeability; Tc, curie focus on the magnetostriction of magnetoelastic effects. By definition, magnetostriction means the change of shape or dimensions of a material during the process of magnetization. The magnetostriction was first identified by James Joule in 1842 when observing a sample of iron [64]. It can be categorized into spontaneous and forced magnetostriction, which are called volume and Joule magnetostriction. The illustration of volume and Joule magnetostriction of a spherical sample is shown in Figure2. The isotropic expansion of the sample near Curie temperature under zero magnetic field is due to the spontaneous magnetostriction, while the Joule magnetostriction results in the elongation of the sample Sensors 2020, 20, x FOR PEER REVIEW 3 of 29

temperature; Qm, mechanical quality factor; * The value is estimated using Young’s modulus and Poisson ratio of the film being 100 GPa and 0.3, respectively.

Magnetoelastic effects refer to the couplings between magnetic and elastic properties of a material, which can beSensors divided2020, 20, into 1532 two categories: direct and inverse effects. The best-known direct4 magnetoelastic of 27 effects are magnetostriction and ∆퐸 effects while some of the inverse effects are described by special terms such as Villariin the effect direction and ofMatteuci magnetic effect field. Thein the deformation literature. in theThe orthogonal detailed directiondescription to the of field magnetoelastic with the effects are listed inopposite Table sign 3. Magnetostrictiveand half the amplitude materials is called transverse have been magnetostriction. playing an important There are many role other in applications ranging fromfactors actuators, that can lead sensors to Joule and magnetostriction energy harvesters. such as the Here, stress, volume we focus change on at the high magnetostriction magnetic of magnetoelasticfields, effects. etc. which By are definition, out of the scope magnetostriction of this review. Figure means3 shows the the change magnetostriction of shape curves or dimensions for of a volume and Joule magnetostriction. As shown in Figure3a, a high volume magnetostriction of 0.82% material duringin FeRh the was process discovered of magnet above roomization. temperature The magnetostriction by Ibarra and Algarabel was [ 65first], which identified experimentally by James Joule in 1842 whendemonstrated observing the a predicted sample metastable of iron ferromagnetic [64]. It can high-volume be categorized state within into the antiferromagnetic spontaneous and forced magnetostriction,phase by which applying are magnetic called volume field. Dong and et al.Joule [66] magnetostriction. reported the Joule magnetostriction The illustration values of ofvolu someme and Joule magnetostrictiontypical magnetostrictiveof a spherical thinsample films includingis shown Ni, in Co, Figure FeGaB, 2. and The FeCoSiB isotropic as shown expansion in Figure of3b. the sample near

Curie temperature under zero magneticTable 3. fieldClassifications is due ofto magnetoelastic the spontaneous effects. magnetostriction, while the Joule magnetostriction results in the elongation of the sample in the direction of magnetic field. The deformation Magnetoelastic Effects in the orthogonal direction to the field with the opposite sign and half the amplitude is called transverse Direct Effects Inverse Effects magnetostriction. There are many other factors that can lead to Joule magnetostriction such as the stress, Joule magnetostriction Villari effect volume change Changeat high of dimensionsmagnetic in fields, the direction etc. of which applied are out of the scope of this review. Figure 3 shows the Change of magnetization due to applied stress magnetostriction curves formagnetic volume field and Joule magnetostriction. As shown in Figure 3a, a high volume Volume magnetostriction Nagaoka-Honda effect magnetostrictionChange of of0.82% volume in due FeRh to spontaneous was discovered magnetization aboveChange room of magnetizationtemperature due by to volume Ibarra change and Algarabel [65], ∆E effect which experimentallyDependence demonstrated of Young’s modulus the on predicted the state of metastableMagnetically ferromagnetic induced changes high in the-volume elasticity state within the antiferromagnetic phase by magnetizationapplying magnetic field. Dong et al. [66] reported the Joule magnetostriction Matteuci effect Wiedemann effect values of some typical magnetostrictive thin films includingHelical anisotropy Ni, Co, and electric FeGaB, and and magnetic FeCoSiB fields as shown in Torque induced by helical anisotropy Figure 3b. induced by a torque

Figure 2. Illustration of volume and Joule magnetostriction of a spherical sample. Figure 2. Illustration of volume and Joule magnetostriction of a spherical sample.

Table 3. Classifications of magnetoelastic effects.

Magnetoelastic effects Direct effects Inverse effects Joule magnetostriction Villari effect Change of dimensions in the direction of applied Change of magnetization due to applied stress magnetic field Nagaoka-Honda effect Volume magnetostriction Change of magnetization due to volume change

Sensors 2020, 20, x FOR PEER REVIEW 4 of 29

Change of volume due to spontaneous magnetization ∆푬 effect

Dependence of Young’s modulus on the state of Magnetically induced changes in the elasticity magnetization Matteuci effect Wiedemann effect Sensors 2020 20 Helical anisotropy and electric and magnetic fields Torque, induced, 1532 by helical anisotropy 5 of 27 induced by a torque

FigureFigure 3. (a 3.) Thermal(a) Thermal dependence dependence of the of volume the volume magnetostriction magnetostriction at 14.2 at 14.2T for T FeRh for FeRh (the (theline lineis a isvisual a guide).visual Reprinted guide). with Reprinted permission with from permission [65]. Copyright from [65 [1994],]. Copyright American [1994], Physical American Society. Physical (b) Measurement Society. results(b) Measurementof the magnetostriction results of constant the magnetostriction in the absolute constant value inof thedifferent absolute magnetic value ofthin diﬀ filmserent (the magnetic line is a visualthin guide). ﬁlms (the line is a visual guide).

1.3. 1.3.Magnetoelectric Magnetoelectric Applications Applications DueDue to the to theadvantages advantages of thin of thin-film-film devices devices such such as aslightweight, lightweight, low low cost, cost, high high spatial spatial resolution resolution and CMOSand-compatible CMOS-compatible fabrication fabrication process, thin process,-film thin-filmME heterostructures ME heterostructures are preferred are preferredin specific inapplications specific whereapplications miniaturization where miniaturization of the device is of crucial the device. Therefore, is crucial. we Therefore, focus on we the focus recent on progress the recent on progress thin-film magnetostrictiveon thin-film magnetostrictive materials in this materials article. Several in this article.types of Several integrated types multiferroic of integrated devices multiferroic based devices on thin- filmbased ME heterostructureson thin-film ME heterostructures including ME sensors including [18,63,67 ME sensors–71], inductors[18,63,67– 71 [35],], inductors filters [32] [35 ], and filters antennas [32] and antennas [43,72,73] are presented in this section. Based on the direct ME effect, a giant ME [43,72,73] are presented in this section. Based on the direct ME effect, a giant ME coefficient (αME) as high as coefficient (α ) as high as 5kV/cm Oe was obtained by the direct deposition of FeCoSiB on Si substrate 5kV/cm Oe was obtainedME by the direct deposition of FeCoSiB on Si substrate with a Si/SiO2/Pt/ AlN/FeCoSiB layerwith stack a Si [7/SiO1]. The2/Pt /highAlN /qualityFeCoSiB of layerAlN stackand FeCoSiB [71]. The films high allows quality the of enhancement AlN and FeCoSiB of limit films of detection allows 1/2 (LoD)the to enhancement be an extremely of limitlow value of detection of 400 푓푇 (LoD)/퐻푧1/ to2 at be the an mechanical extremely resonance low value frequency of 400 f T /ofHz 867 Hz.at theHere, mechanical resonance frequency of 867 Hz. Here, LoD is also frequently referred to as equivalent 푃 LoD is also frequently referred to as equivalent magnetic noiseq floor and given in the equation: LoD = √ , = P 푆2 magnetic noise floor and given in the equation: LoD S2 , where P is the power spectral density of 2 2 2 wherethe P arbitrary is the power quantity spectral in units density of au of2/ theHz, arbitrary such as V quantity2/Hz and in rad units2/Hz; of Sau is/Hz, the magnetic such as V field/Hz sensitivityand rad /Hz; S is ofthe the magnetic arbitrary field quantity sensitivity in units of the of auarbitrary/T, such quantity as V/T and in units rad/T. of Therefore, au/T, such the as V/T unit and for LoDrad/T. is Therefore,T/ √Hz. the Theunit singlefor LoD cantilevers is T/√퐻푧 were. The cutsingle from cantilevers wafers with were a widthcut from of 2.2wafers mm with and a alength width ofof 25.22.2 mm mm and as shown a length of 25.2in Figure mm as4a. shown These in cantilevers Figure 4a. were These mounted cantilevers on printedwere mounted circuit boardson printed (PCBs), circuit and boards the top-bottom (PCBs), and the electrodetop-bottom connections electrode wereconnections established were manually. established Operating manually. at theOperating working at point the working with a bias point dc fieldwith a −4 −7 biasof dc2.1 field10 of4 T2.1and× 10 a constant푇 and aca constant driving field ac driving of 1 10 field7T ,of the 1 ME× 10 coe푇ffi, cientsthe ME of coefficients the sensor with of the respect sensor × − × − withto respect frequency to frequency were measured were and measured pictured and in picturedFigure4b. in The Figure electromechanical 4b. The electromechanical resonance frequency resonance frequencyand quality and quality factor fittedfactor to fitted Lorentzian to Lorentzian equation equation were observed were observed as 867 as Hz 867 and Hz 310. and By 310. performing By performing 15 15 consecutiveconsecutive sensitivity sensitivity measurements andand averagingaveraging them, them, an an estimation estimation of of the the LoD LoD of of the the sensor sensor was was 400 ± 37 푓푇/퐻푧1/2 determineddetermined as as 400 37 f T/Hz1 and/2 and shown shown in inFigure Figure 4c.4 c.The The influences influences of of different di fferent noise noise regimes regimes on on the ± the LOD of a ME sensor were investigated by measuring LoD at various frequencies, which is depicted in Figure4d. Based on the converse ME e ffect, a voltage tunable inductor and bandpass filter were successfully demonstrated by Lin and Gao et al. [32,35] The schematics and SEM images are presented in Figure5. The solenoid structure using FeGaB /Al2O3 multilayer was bonded to a lead magnesium niobate-lead titanate (PMN-PT) piezoelectric slab to fabricate the tunable inductor. Figure5b shows the measured inductance under different E-field applied across the thickness of the PMN-PT slab, in which the inset shows the SEM picture of the inductor. A high tunable inductance of >100% is obtained over a large frequency range from 0 to 2 GHz and a peak inductance tunability of 191% at 1.5 GHz is obtained. The quality factor is also enhanced more than 100% over the frequency range from 0 to 1.5 GHz. Two coupled elliptic-shape nano-mechanical resonators consisting of FeGaB/Al2O3 multilayer Sensors 2020, 20, x FOR PEER REVIEW 5 of 29

LOD of a ME sensor were investigated by measuring LoD at various frequencies, which is depicted in Figure 4d. Based on the converse ME effect, a voltage tunable inductor and bandpass filter were successfully demonstrated by Lin and Gao et al. [32,35] The schematics and SEM images are presented in Figure 5. The solenoid structure using FeGaB/Al2O3 multilayer was bonded to a lead magnesium niobate- lead titanate (PMN-PT) piezoelectric slab to fabricate the tunable inductor. Figure 5b shows the measured inductance under different E-field applied across the thickness of the PMN-PT slab, in which the inset shows the SEM picture of the inductor. A high tunable inductance of >100% is obtained over a large frequencySensors 2020 range, 20, 1532from 0 to 2 GHz and a peak inductance tunability of 191% at 1.5 GHz is obtained.6 of 27The quality factor is also enhanced more than 100% over the frequency range from 0 to 1.5 GHz. Two coupled elliptic-shape nano-mechanical resonators consisting of FeGaB/Al2O3 multilayer and AlN film are excited in inand-plane AlN contourfilm are excited mode to in in-plane realize the contour electromagnetic mode to realize (EM) the transduct electromagneticion. The (EM) dependence transduction. of the measuredThe dependence resonant frequency of the measured on the applied resonant magnetic frequency field on was the appliedmeasured magnetic and shown field in was Figure measured 5d. Both E-fieldand and shown H-field in Figure tunability5d. Both were E-field also measured and H-field and tunability the frequency were dependence also measured on DC and voltage the frequency across the thicknessdependence of AlN on film DC voltage is shown across in ref.the thickness [13]. The of H AlN-field film frequency is shown tunability in ref. [13 ]. of The 5 kHz/Oe H-field frequency and E-field frequencytunability tunability of 5 kHz of/Oe 2.3 andkHz/Oe E-field were frequency observed tunability in same ofdevice. 2.3 kHz /Oe were observed in same device.

FigureFigure 4. (a 4.) Schematic(a) Schematic illustration illustration of the of cantilever the cantilever-based-based thin- thin-filmfilm magnetic magnetic sensor sensor using using inversed inversed bilayer −7 7 ME bilayerheterostructures. ME heterostructures. (b) ME coefficient (b) ME as coe a ffi functioncient as of a functionfrequency of (퐻 frequency푎푐 = 1 × 10 (Hac푇= and1 퐻10푏푖푎푠 =T ±and2.1 × × − 10−H4푇). The= electromechanical2.1 10 4T). The electromechanical resonance frequency resonance was measured frequency to was be measured867 Hz and to bethe 867 quality Hz and factor the is bias ± × − determinedquality factor as 310 is fromdetermined the applied as 310 Lorentzian from the applied fit. (c) Lorentzian Averaged LoDfit. (c plot) Averaged with error LoD bars plot with that indicate error Sensorsstandard 2020bars, 20 that , deviation.x FOR indicate PEER (REVIEW standardd) The measured deviation. ME (d) sensor’sThe measured LoD with ME sensor’s respect LoD to frequency. with respect Reproduced to frequency.6 of with 29 permissionReproduced from with[71]. permissionCopyright [2016], from [71 American]. Copyright Institute [2016], of AmericanPhysics. Institute of Physics.

FigureFigure 5. (a 5.) Structure(a) Structure model model of th ofe the voltage voltage tunable tunable inductor inductor based based on on FeGaB/PMN FeGaB/PMN-PT-PT heterostructures. heterostructures. (b) Measured(b) Measured inductance inductance under underdifferent diff erentE-field E-field applied applied across across the thickness the thickness of the of PMN the PMN-PT-PT slab slab(inset (inset shows theshows SEM picture the SEM of picturethe inductor). of theinductor). (c) Schematic (c) Schematic of the voltage of the tunable voltage filter tunable based filter on FeGaB/AlN based on FeGaB multilayers./AlN (d) multilayers.Dependence ( dof) Dependencethe measured of resonant the measured frequency resonant on the frequency applied on magnetic the applied field magnetic (inset shows field (insetthe SEM pictureshows of the SEMfilter). picture Reproduced of the filter).with permission Reproduced from with [32]. permission Copyright from [2016], [32]. IEEE. Copyright [2016], IEEE.

Utilizing both direct and converse ME effects, the unprecedented demonstration of ultra-compact ME NEMS antennas was proposed by us in 2017 [43] and was expected to have great impacts on our future antennas and communication systems. Two structures were designed to demonstrate the novel receiving and transmitting mechanism: nano-plate resonator (NPR) and thin-film bulk acoustic resonator (FBAR). The mechanism of the new antenna is explained as follow: from the receiving aspect, the magnetostrictive layer can detect the RF magnetic field component of the EM wave and induce an acoustic wave on the ferromagnetic layer. When this acoustic wave transfer to the piezoelectric thin-film, the dynamic voltage/charge would be generated due to the direct piezoelectric coupling; Reciprocally, from the transmitting aspect, the mechanical resonance, which is generated by applying RF electric field to the MEMS resonator, would induce acoustic wave that can be directly transfer to the upper ferromagnetic thin film. Then a dynamic change of the magnetization induced by acoustic wave due to the strong magnetostriction constant would generate magnetic current for radiation. As shown in Figure 6a,b, the antenna measurement setup is consisted of a horn antenna and our ME antenna whose radiative element is the suspended FeGaB/AlN circular disk. The electromechanical resonance frequency (fr,FBAR) of the ME 1 퐸 FBAR is defined by the thickness of the circular resonating disk and can be expressed by 푓 ~ √ , 푟,퐹퐵퐴푅 2푇 휌 where T is the total thickness, E is the effective Young’s modulus, and 휌 is the effective density of the ME disk. The performance of ME FBAR was measured and presented in Figure 6c,d,a resonance frequency of 2.53 GHz and calculated gain of -18 dBi were achieved from the results of return loss (S11), transmission and receiving (S12 and S21) curves. A non-magnetic aluminum (Al) FBAR structure was fabricated as a control device to demonstrate that the source of radiation is from the ME coupling. A similar electromechanical properties of the control device is shown in Figure 6e, however, the much lower gain in Figure 6f suggests that the ME coupling dominates the radiation of FBAR antennas.

Sensors 2020, 20, 1532 7 of 27

Utilizing both direct and converse ME effects, the unprecedented demonstration of ultra-compact ME NEMS antennas was proposed by us in 2017 [43] and was expected to have great impacts on our future antennas and communication systems. Two structures were designed to demonstrate the novel receiving and transmitting mechanism: nano-plate resonator (NPR) and thin-film bulk acoustic resonator (FBAR). The mechanism of the new antenna is explained as follow: from the receiving aspect, the magnetostrictive layer can detect the RF magnetic field component of the EM wave and induce an acoustic wave on the ferromagnetic layer. When this acoustic wave transfer to the piezoelectric thin-film, the dynamic voltage/charge would be generated due to the direct piezoelectric coupling; Reciprocally, from the transmitting aspect, the mechanical resonance, which is generated by applying RF electric field to the MEMS resonator, would induce acoustic wave that can be directly transfer to the upper ferromagnetic thin film. Then a dynamic change of the magnetization induced by acoustic wave due to the strong magnetostriction constant would generate magnetic current for radiation. As shown in Figure6a,b, the antenna measurement setup is consisted of a horn antenna and our ME antenna whose radiative element is the suspended FeGaB/AlN circular disk. The electromechanical resonance frequency (f ) of the ME FBAR is defined by the thickness of the circular resonating disk and can be r,FBAR q expressed by f 1 E , where T is the total thickness, E is the effective Young’s modulus, and r, FBAR ∼ 2T ρ ρ is the effective density of the ME disk. The performance of ME FBAR was measured and presented in Figure6c,d,a resonance frequency of 2.53 GHz and calculated gain of 18 dBi were achieved from the − results of return loss (S11), transmission and receiving (S12 and S21) curves. A non-magnetic aluminum (Al) FBAR structure was fabricated as a control device to demonstrate that the source of radiation is from the ME coupling. A similar electromechanical properties of the control device is shown in SensorsFigure 20206e,, 20 however,, x FOR PEER the REVIEW much lower gain in Figure6f suggests that the ME coupling dominates7 of 29 the radiation of FBAR antennas.

Figure 6. Thin-ﬁlm based ME FBAR antennas. (a) Illustration of the antenna measurement setup. Figure 6. Thin-film based ME FBAR antennas. (a) Illustration of the antenna measurement setup. (b) SEM (b) SEM image of the fabricated the ME FBAR. (c) Return loss curve (S22) of ME FBAR. The inset shows image of the fabricated the ME FBAR. (c) Return loss curve (S22) of ME FBAR. The inset shows the out-of- the out-of-plane displacement of the circular disk at resonance peak position. (d) Transmission and plane displacement of the circular disk at resonance peak position. (d) Transmission and receiving behavior receiving behavior (S12 and S21) of ME FBAR. (e) Return loss (S22) curve of the non-magnetic Al/AlN (S12 and S21) of ME FBAR. (e) Return loss (S22) curve of the non-magnetic Al/AlN control FBAR. (f) control FBAR. (f) Transmission and receiving behavior (S12 and S21) of the non-magnetic Al/AlN control Transmission and receiving behavior (S12 and S21) of the non-magnetic Al/AlN control FBAR. Reprinted with FBAR. Reprinted with permission from [43]. Copyright [2017], Springer Nature. permission from [43]. Copyright [2017], Springer Nature.

2. Fabrication and Characterization Methods

2.1. Fabrication Techniques of Thin Films Due to the advantages of thin films, there are a variety of thin-film preparation techniques that have been developed such as physical vapor deposition (PVD) [74,75], chemical vapor deposition (CVD) [76,77], atomic layer deposition (ALD) [78,79], and sol-gel [80] etc. Here, we focus on the description of PVD process which is the most commonly used method for thin film deposition. PVD is a thin-film deposition process in a vacuum and plasma environment in which films with thickness in the range of a few nanometers to a few micrometers are produced. The source materials are vaporized and nucleated onto a substrate surface in a vacuum environment to fabricate the thin films. The vapor phase is usually consisted of plasma or ions. There are three major PVD process: evaporation, ion plating and sputtering, where the mechanisms for creating vapor phase are thermal effect, ion bombardment and gaseous ions, respectively. Tungsten wire coils are normally utilized as the source for heating the target materials and the deposition rate of the thermal evaporation is high compared to other PVD processes. Different configurations of evaporation such as molecular beam epitaxy (MBE) and activated reactive evaporation (ARE) are exploited to improve the films’ qualities as well. Based on the sources of sputtering deposition, various sputtering processes were developed such as diode sputtering, reactive sputtering and magnetron sputtering etc. Our PVD system is magnetron sputtering which is able to do both DC and radio frequency (RF) sputtering. DC voltage is provided between the anode and cathode, which are used to place substrate and target material, to sustain the glow discharge. The target material for DC sputtering is usually metallic so that the glow discharge can

Sensors 2020, 20, 1532 8 of 27

2. Fabrication and Characterization Methods

2.1. Fabrication Techniques of Thin Films Due to the advantages of thin films, there are a variety of thin-film preparation techniques that have been developed such as physical vapor deposition (PVD) [74,75], chemical vapor deposition (CVD) [76,77], atomic layer deposition (ALD) [78,79], and sol-gel [80] etc. Here, we focus on the description of PVD process which is the most commonly used method for thin film deposition. PVD is a thin-film deposition process in a vacuum and plasma environment in which films with thickness in the range of a few nanometers to a few micrometers are produced. The source materials are vaporized and nucleated onto a substrate surface in a vacuum environment to fabricate the thin films. The vapor phase is usually consisted of plasma or ions. There are three major PVD process: evaporation, ion plating and sputtering, where the mechanisms for creating vapor phase are thermal effect, ion bombardment and gaseous ions, respectively. Tungsten wire coils are normally utilized as the source for heating the target materials and the deposition rate of the thermal evaporation is high compared to other PVD processes. Different configurations of evaporation such as molecular beam epitaxy (MBE) and activated reactive evaporation (ARE) are exploited to improve the films’ qualities as well. Based on the sources of sputtering deposition, various sputtering processes were developed such as diode sputtering, reactive sputtering and magnetron sputtering etc. Our PVD system is magnetron sputtering which is able to do both DC and radio frequency (RF) sputtering. DC voltage is provided between the anode and cathode, which are used to place substrate and target material, to sustain the glow discharge. The target material for DC sputtering is usually metallic so that the glow discharge can be maintained. RF sputtering is specifically designed to avoid charge building up on the surface of target materials, such as insulators. The magnetic field is exploited to restrict the movement of the secondary electrons emitted from the target surface and therefore efficiently increase the deposition rate.

2.2. Characterization Methods of Thin Films The properties of magnetic and piezoelectric phases are crucial on achieving large ME coupling strength, in which piezomagnetic and piezoelectric coefficients play significant roles among a variety of thin film properties. Given that ME heterostructures are widely applied in RF devices, extensive works have been carried out to develop the characterization methods for thin film materials. Some well-known techniques have been developed for decades, such as X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD), which are utilized to measure the compositions and crystal structures of thin films. Magnetic properties including hysteresis loops and ferromagnetic resonance (FMR) curves can be measured using vibrating sample magnetometers (VSM), superconducting quantum interference devices (SQUID), and FMR tester. A waveguide and vector network analyzer (VNA) are utilized to extract the permeability, loss tangent and FMR linewidth of magnetic thin films from the measured S parameters [81]. With regard to characterization of magnetostriction constants, both direct and indirect methods were proposed with different mechanisms [82–87]. Since the deflection of a cantilever structure with the magnetic film on the upper side is proportional to the magnetostriction, an instrument was devised to measure the magnetostriction as a function of applied field by measuring the change of mechanical resonance frequency [83]. By utilizing a continuous laser beam system, a non-contact method was developed to measure the saturation magnetostriction of a thin soft-magnetic film deposited on a thin glass substrate under a rotating magnetic field [84]. In addition to the increased interest in magnetostriction, ∆E effect that represents the change of Young’s modulus as a function of magnetic field has drawn a great amount of attention to build ultra-sensitive ME sensors. Based on the non-contact optical technique, a simple, compact, and sensitive system to measure the magnetostriction and ∆E effect of magnetic thin films was developed by us recently [66]. The schematic of the proposed magnetostriction tester is shown in Figure7a. A rotating adjustable AC magnetic field with a maximum value of 300 Oe is generated by two pairs of mutually perpendicular set Helmholtz coils with equivalent amplitude and 90◦ phase shift current. Sensors 2020, 20, 1532 9 of 27

Similar to the cantilever method for measuring piezoelectric coefficients, the tip displacement of the cantilever is detected by a MTI-2000 fiber-optic sensor due to the strain change in the magnetostriction film. Figure7b shows the schematic of the proposed delta-E tester, which is very similar to the magnetostriction tester. Second vibration mode of the cantilever is excited by applying a 0.1 Oe AC driving field through a solenoid with the frequency swept around 1555 Hz. A DC bias magnetic field is generated by the Helmholtz coil to induce magnetic domain change in the magnetostrictive films, and thus resulting in the change of Young’s modulus. According to the modified Euler-Bernoulli beam theorem, the change of Young’s modulus can be derived by measuring the shift in resonant frequency of the cantilever. Some typical magnetostrictive films such as Ni, Co, FeGa etc. with well-known magnetostriction values are used to calibrate the system. Figure7c shows the measured magnetostriction and piezomagnetic coefficient for as-deposited and annealed FeGaB thin films. The Sensorschange 2020, 20 in, x Young’s FOR PEER modulus REVIEW of FeGaB thin film is also measured and shown in Figure7d by9 of using 29 delta-E tester.

FigureFigure 7. (a 7.) Schematic(a) Schematic of the of proposed the proposed magnetostriction magnetostriction tester testerin three in-view. three-view. (b) Schematic (b) Schematic of the proposed of the deltaproposed-E tester delta-Ein three tester-view. in(c three-view.) Measured (magnetostrictionc) Measured magnetostriction and piezomagnetic and piezomagnetic coefficient for coeas-depositedfficient andfor annealed as-deposited FeGaB andthin annealedfilms. (d) Measured FeGaB thin delta films.-E effect (d) Measured for as-deposited delta-E and effect annealed for as-deposited FeGaB thin and films. Reprintedannealed with FeGaB permission thin films. from Reprinted [66]. Copyright with permission [2018], American from [66 ].Institute Copyright of Physics. [2018], American Institute of Physics. Both the fundamental studies of magnetism and boosting of magnetic industry are greatly influenced by the imagingBoth the of magnetic fundamental microstructures. studies of magnetism One of the most and boostingeffective ways of magnetic to characterize industry the are performance greatly of MEinfluenced-based devices by the imaging is the analyzation of magnetic of microstructures. magnetic domain One activity. of the most There eff ectiveare only ways a few to characterize methods for observingthe performance magnetic ofdomains ME-based down devices to device is the level analyzation including of magnetic magnetic force domain microscopy activity. (MFM)There are [88], scanningonly a electron few methods microscopy for observing with polarization magnetic domains analysis down (SEMPA) to device [89] and level soft including X-ray magnetic circular force dichroismmicroscopy–photoemission (MFM) [88], el scanningectron microscopy electron microscopy (XMCD-PEEM) with polarization [90]. Due to analysis magnetic (SEMPA) fields [emitting89] and soft from the X-raydomain magnetic in the magnetic circular dichroism–photoemission materials, the force gradient electron acting microscopy on a small (XMCD-PEEM)magnetic tip can [90 be]. sensed Due to by usingmagnetic ac detection fields emittingmethod [88]. from A the high domain contrast in the and magnetic spatial materials,resolution the of forcemagn gradientetic domain acting was on seen a small from the magneticmagnetization tip can patterns. be sensed SEMPA by using is acanother detection imaging method technique [88]. A highwith contrasthigh spatial and spatialresolution resolution for surface of magneticmagnetic microstructures domain was on seen the from 10-nm the-length magnetization scale [89]. patterns. By employing SEMPA a SEM is another electron imaging optical technique column, the secondarywith high electrons spatial w resolutionhose spin for polarization surface magnetic are directly microstructures related to the on local the 10-nm-lengthmagnetization scale orientation [89]. By are excitedemploying and emitted a SEM from electron the opticalferromagnetic column, sample the secondary surface. electrons Therefore, whose the surface spin polarization magnetization are directly map can be generated when the spin polarization of secondary electrons is analyzed across the sample surface. A very large average velocity of 600 m/s was detected for domain-wall motion in the NiFe layer by using XMCD-PEEM [90]. In particular, magnetic domain imaging by magneto-optical Kerr effect (MOKE) microscopy offers direct access to the behavior of local magnetization [91–94]. MOKE microscopy based on the Kerr and the Faraday effects is one of the most prominent techniques for observing magnetic domain and is able to visualize the magnetic dynamics on fast time-scales. Allowing for the direct imaging under a continuous field excitation, a picosecond wide-field MOKE microscopy for imaging magnetization dynamics is demonstrated in [92]. An example of nanosecond domain wall displacements and spin-wave generation are observed and displayed in Figure 8 [93]. Figure 8a shows the precessional domain wall motion with out-of-plane magnetization components of a (Fe90Co10)78Si12B10 sample. The spin waves, which origin from the domain wall precession within the central magnetic domains shown in Figure 8b, propagate

Sensors 2020, 20, 1532 10 of 27

related to the local magnetization orientation are excited and emitted from the ferromagnetic sample surface. Therefore, the surface magnetization map can be generated when the spin polarization of secondary electrons is analyzed across the sample surface. A very large average velocity of 600 m/s was detected for domain-wall motion in the NiFe layer by using XMCD-PEEM [90]. In particular, magnetic domain imaging by magneto-optical Kerr effect (MOKE) microscopy offers direct access to the behavior of local magnetization [91–94]. MOKE microscopy based on the Kerr and the Faraday effects is one of the most prominent techniques for observing magnetic domain and is able to visualize the magnetic dynamics on fast time-scales. Allowing for the direct imaging under a continuous field excitation, a picosecond wide-field MOKE microscopy for imaging magnetization dynamics is demonstrated in [92]. An example of nanosecond domain wall displacements and spin-wave generation are observed and displayed in Figure8[ 93]. Figure8a shows the precessional domain Sensors 2020, 20, x FOR PEER REVIEW 10 of 29 wall motion with out-of-plane magnetization components of a (Fe90Co10)78Si12B10 sample. The spin waves, which origin from the domain wall precession within the central magnetic domains shown in withFigure a velocity8b, propagate of 650 m/s. with Moreover, a velocity ofthe 650 direct m /s. imaging Moreover, of standing the direct spin imaging waves of at standing field excitations spin waves in a at field excitations in a Co Fe B /Ru/ Co Fe B anti-dot array at 2 GHz is exhibited in Figure8c. Co40Fe40B20/Ru/ Co40Fe40B20 anti40-dot40 array20 at 2 GHz40 40 is exhibited20 in Figure 8c.

FigureFigure 8. Domain 8. Domain images images of ofmagnetic magnetic dy dynamicsnamics for for a magnetic element.element. (a(a)) Domain Domain wall wall motion motion at wallat wall resonance.resonance. (b) (Domainb) Domain wall wall induced induced spin spin-wave-wave propagation propagation with timetime underunder a a 100MHz 100MHz ﬁeld field excitation. excitation.

[sample:[sample: (Fe (Fe90Co9010Co)7810Si)1278BSi1012(160B10 (160 nm)] nm)] (c) (c Imaging) Imaging of of standing standing spin-wave spin-wave modes modes in ina Co a 40 CoFe40FeB2040/BRu20/Ru// Co40CoFe4040BFe2040 antiB20 -anti-dotdot array array at 2at GHz.2 GHz. Reprinted Reprinted with with permission from from [93 [93].]. Copyright Copyright [2015], [2015], American American PhysicalPhysical Society. Society.

3. Magnetostrictive3. Magnetostrictive Thin Thin Films Films 3.1. Rare-Earth-Transition Metal Inter-Metallics 3.1. Rare-Earth-Transition Metal Inter-Metallics Magnetostriction was ﬁrst reported by James Joule in the early 1840s, who observed the change of Magnetostriction was first reported by James Joule in the early 1840s, who observed the change of length in iron particles when exposed to magnetic ﬁelds [64]. Villari discovered that applying stress to length in iron particles when exposed to magnetic fields [64]. Villari discovered that applying stress to magnetostrictive materials changes their magnetization [95]. However, magnetostrictive materials magnetostrictive materials changes their magnetization [95]. However, magnetostrictive materials have not have not been extensively used until World War II, when nickel-based alloys were employed in building been extensively used until World War II, when nickel-based alloys were employed in building sonar sonar applications. The magnetostriction of 3d transition metals Fe, Co and Ni can produce very high applications. The magnetostriction of 3d transition metals Fe, Co and Ni can produce very high forces, but forces, but the range of motion is small. The extraordinary magnetostrictive properties of the rare-earth the range of motion is small. The extraordinary magnetostrictive properties of the rare-earth elements such as terbium and dysprosium were first recognized in the early 1960’s, which can achieve magnetostriction of up to 10000 ppm at low temperatures [96]. However, they do not demonstrate significant magnetostriction at room temperature where practical devices must operate due to the low Curie temperature. The solution to this problem was pointed out by Callen in 1969, who suggested there was great promise that strong magnetism of the transition metals such as iron, which exhibit higher Curie temperatures, could increase the rare-earth magnetic order at high temperatures [97]. In attempt to produce large magnetostriction at room temperature, rare-earth elements were alloyed with 3d transition elements as rare-earth-transition metal inter-metallics.

3.1.1. TbDyFe

Sensors 2020, 20, 1532 11 of 27

elements such as terbium and dysprosium were ﬁrst recognized in the early 1960’s, which can achieve magnetostriction of up to 10000 ppm at low temperatures [96]. However, they do not demonstrate signiﬁcant magnetostriction at room temperature where practical devices must operate due to the low Curie temperature. The solution to this problem was pointed out by Callen in 1969, who suggested there was great promise that strong magnetism of the transition metals such as iron, which exhibit higher Curie temperatures, could increase the rare-earth magnetic order at high temperatures [97]. In attempt to produce large magnetostriction at room temperature, rare-earth elements were alloyed with 3d transition elements as rare-earth-transition metal inter-metallics.

Sensors3.1.1. 2020 TbDyFe, 20, x FOR PEER REVIEW 11 of 29 In 1979, the magnetic materials group at Carderock which was funded by the Office of Naval ResearchIn 1979, (ONR)the magnetic discovered materials that group the magnetocrystalline at Carderock which anisotropy was funded of by rare-earth-transition the Office of Naval Research metal (ONR)inter-metallics discovered can that be the significantly magnetocrystalline reduced by anisotropy adding Tb of and rare Dy-earth in certain-transition proportion metal inter into-meta iron.llics This can be ledsignificantly to the development reduced by of adding the Terfenol-D Tb and Dy alloy in certain (Tb0.3Dy proportion0.7Fe2), which into exhibitsiron. This giant led magnetostrictionto the development of theof about Terfenol 2000-D ppm alloy at (Tb room0.3Dy temperature0.7Fe2), which [ 98exhibits]. The largegiant strainmagnetostriction produced by of Terfenol-Dabout 2000 isppm useful at room in temperatureapplications [98]. as The production large strain of highproduced amplitude, by Terfenol low frequency-D is useful sound in applications wavs in water, as production certain types of high amplitude,of strain low gages, frequency vibration sound compensation wavs in water, and compensation certain types of for strain temperature gages, vibration induced compensation strains in large and compensationlaser mirrors. for Onetemperature of the draw induced backs strains of Terfenol-D in large laser is its mirrors. brittleness One in of withstanding the draw backs tensile of Terfenol stress,-D is itswhich brittleness limits itsin abilitywithstanding to withstand tensile shock stress, loads which or limits operate its in ability tension. to withstand shock loads or operate in tension.In terms of Terfenol-D thin films, several efforts on the fabrication techniques and characterization studiesIn terms have of been Terfenol conducted-D thin during films, theseveral last twoefforts decades on the [98 fabrication,99]. Terfenol-D techniques thin films and arecharacterization commonly studieprepareds have by been magnetron conducted sputtering during with the lasta single two alloy decades target [98,99 or by]. co-sputtering Terfenol-D thin with films multiple are commonly targets. preparedHowever, by magnetron the as-deposited sputtering films with are alwaysa single amorphous alloy target and or by their co- magnetoelasticsputtering with behaviorsmultiple targets. and However,magnetic the properties as-deposited are films inferior are thanalways their amorphous bulk counterpart. and their magnetoelastic To overcome these behaviors drawbacks and magnetic and propertiesenhance are their inferior mechanical than their and magnetic bulk counterpart. properties, To the overcome thin films these must drawbacks be crystallized and enhance either by their mechanicalpost-annealing and magnetic or by depositing properties, the filmthe thin on a films heating must substrate. be crystallized In 1994, Williamseither by P.post I. et.-annealing al. observed or by depositingin-plane t magnetostrictionhe film on a heating coe substrate.fficient of In 500 1994, ppm Williams in Tb0.3Dy P. 0.7I. et.Fe 2al.polycrystalline observed in-plane thin filmmagnetostriction as shown coefficientin Figure of9 [500100 ppm]. However, in Tb0.3Dy the0.7 saturationFe2 polycrystalline field achieved thin film at 4as kOe shown was in still Figure too high, 9 [100 which]. However, made it the saturationimpractical field for achieved the usage at in4 MEMSkOe was applications, still too high, as it which would made be very it impractical difficult to apply for the such usage a high in biasMEMS applications,magnetic fieldas it onwould chip. be very difficult to apply such a high bias magnetic field on chip.

FigureFigure 9. ( 9.a) (Ina)-plane, In-plane, longitudinal longitudinal magnetostriction magnetostriction λ‖ loopλ loop for fora polycrystalline a polycrystalline Tb0.3 TbDy0.30.7DyFe0.72 filmFe2 ﬁlmand (b) k λ‖ andand λ (⊥b )asλ a andfunctionλ as of a x function for TbxDy of1 x-xFe for2 films. TbxDy Reprinted1-xFe2 ﬁlms. with Reprinted permission with from permission [100]. Copyright from [100 [1994],]. k ⊥ AmericanCopyright Physical [1994], Society. American Physical Society.

3.1.2. SmFe and SmFeB 3.1.2. SmFe and SmFeB In addition to Terfenol-D alloys, samarium-based alloys have also been of great research In addition to Terfenol-D alloys, samarium-based alloys have also been of great research interests due interests due to their high Curie temperature. Unlike Terfenol-D alloys, which start to degrade to their high Curie temperature. Unlike Terfenol-D alloys, which start to degrade their magnetostrictive performance at 200 C°, Samarium-based alloys can be applied to much higher temperature environment without losing their magnetic properties [101]. One special character of Sm-Fe alloy is their negative sign of giant magnetostriction, which lies in the relationship between the magnetoelastic behaviors and the stress state of the magnetostrictive materials. Specifically, the compressive stress can increase the strain for magnetostrictive materials with positive sign, whereas decrease the strain for materials with negative sign. Therefore, in device applications using bulk magnetostrictive materials, usually they are intentionally applied with compressive stress rather than tensile stress for convenience to achieve the largest strain. However, in thin film applications, it is not necessary to apply any stress on the thin films, therefore, Sm-

Sensors 2020, 20, 1532 12 of 27

their magnetostrictive performance at 200 ◦C, Samarium-based alloys can be applied to much higher temperature environment without losing their magnetic properties [101]. One special character of Sm-Fe alloy is their negative sign of giant magnetostriction, which lies in the relationship between the magnetoelastic behaviors and the stress state of the magnetostrictive materials. Specifically, the compressive stress can increase the strain for magnetostrictive materials with positive sign, whereas decrease the strain for materials with negative sign. Therefore, in device applications using bulk magnetostrictive materials, usually they are intentionally applied with compressive stress rather than Sensors 2020, 20, x FOR PEER REVIEW 12 of 29 tensile stress for convenience to achieve the largest strain. However, in thin film applications, it is not necessary to apply any stress on the thin films, therefore, Sm-Fe based alloys do not suffer from Fe basedthe problem alloys do related not suffer to the from reduction the problem of strain related due to to the the applied reduction stress of onstrain negative due to magnetostrictive the applied stress on negativematerials. magnetostrictive Another advantage materials. is that Another the light advantage rare-earth is element that the samarium light rare is-earth more element abundant samarium than is morethe heavy abundant rare-earth than the element heavy terbium.rare-earth Therefore, element terbium. Sm-Fe based Therefore, alloys Sm have-Fe abased price alloys advantage have overa price advantageTerfenol-D over alloys. Terfenol These-D alloys. advantages These lead advantages to the investigation lead to the investigation and development and development of Sm-Fe based of thinSm-Fe basedfilms thin with films good with magnetostrictive good magnetostrictive characteristics. characteristics. In 1998,In 1998, Kim Kim et etal. al. conducted conducted a asystematic systematic investigation investigation on thethe magnetostrictivemagnetostrictive behavior behavio ofr of Sm-Fe Sm-Fe andand Sm Sm-Fe-B-Fe-B thin thin films films over over a awide wide composition composition range range from from 14.1 toto 71.771.7 at%at% Sm Sm [102 [102].]. As As displayed displayed in in FigureFigure 10, 10good, good magnetostrictive magnetostrictive characteristics characteristics has has been been found found at low at low magnetic magnetic fields fields in inboth both Sm Sm-Fe-Fe and and Sm-Fe-B thin films. At 100 Oe, magnetostriction of 350 ppm and 470 ppm have been achieved Sm-Fe-B thin films. At 100 Oe, magnetostriction of -350 ppm− and -470 ppm− have been achieved for Sm-Fe withfor 25% Sm-Fe Sm withand for 25% Sm Sm-Fe and-B with for Sm-Fe-B 27-31% Sm with respectively. 27–31% Sm The respectively. addition Theof B additionhas formed of B an has amorphous formed phasean amorphous in Sm-Fe-B phase thin films, in Sm-Fe-B which thin effectively films, which modified effectively the microstructure modified the microstructure and thereby andimprove therebyd the improved the magnetic properties. At the same time, a very high λs of 1200 ppm has been achieved magnetic properties. At the same time, a very high 휆푠 of -1200 ppm has been− achieved with an optimum Sm withcontent an optimumat 36.8% for Sm Sm content-Fe-B atthin 36.8% films, for which Sm-Fe-B is three thin films,times whichlarger isthan three Tb times-Fe based larger thin than films. Tb-Fe The excellentbased magnetostrictive thin films. The excellent characters magnetostrictive of Sm-Fe and Sm characters-Fe-B thin of films Sm-Fe made and them Sm-Fe-B good thincandidates films made for the applicationsthem good in candidatesMEMS device. for the applications in MEMS device.

FigureFigure 10. ( 10.a) The(a) Theλ-H λplots-H plots for some for some Sm-Fe Sm-Fe-B-B thin thinfilms. films. The numbers The numbers at the at curves the curves indicate indicate the Sm the content Sm in at%content (b) The in at%value (b of) The λ as value a function of λ as of a the function Sm conte of thent Smat fixed content magnetic at fixed fields magnetic of 5 kOe. fields The of results 5 kOe. of The Sm- Fe thinresults films of Sm-Feare indicated thin films by arefilled indicated circles while by filled those circles of Sm while-Fe- thoseB thin of films Sm-Fe-B are denoted thin films by are open denoted circles. Reprintedby open with circles. permission Reprinted from with [102]. permission Copyright from [1998], [102 ].North Copyright Holland [1998], (Elsevier). North Holland (Elsevier). 3.2. Rear-Earth-Free Alloys 3.2. Rear-Earth-Free Alloys The well-known problem of the magnetostrictive rear-earth-transition metal films is that a The well-known problem of the magnetostrictive rear-earth-transition metal films is that a large bias large bias magnetic field is usually required to achieve giant magnetostriction due to the large magnetic field is usually required to achieve giant magnetostriction due to the large magnetocrystalline magnetocrystalline anisotropy induced from the rear-earth elements, which is particularly impractical anisotropy induced from the rear-earth elements, which is particularly impractical for these for these magnetostrictive thin films to be applied on MEMS devices. Also, at RF and microwave magnetostrictive thin films to be applied on MEMS devices. Also, at RF and microwave frequencies, even frequencies, even small amount of rear-earth elements can result in a very lossy magnetic film. small amount of rear-earth elements can result in a very lossy magnetic film. Recently, strong Recently, strong magnetostrictive behaviors have been observed in Fe-Ga and Fe-Al alloys, which magnetostrictive behaviors have been observed in Fe•Ga and Fe•Al alloys, which exhibit large exhibit large magnetostriction of more than 400 ppm for Fe81.3Ga18.7 and 200 ppm for Fe83.4Al16.6 magnetostriction of more than 400 ppm for Fe81.3Ga18.7 and 200 ppm for Fe83.4Al16.6 respectively [103,104]. In these alloys, it has been found that the giant magnetostriction always occurred in the vicinity of ordered D03 and disordered A2 phase boundaries. In addition, since the first report on the magnetostrictive properties of Fe-Co alloy by Masiyama, much efforts have also been devoted in developing giant magnetostrictive Fe•Co alloys [105].

3.2.1. FeGa Based Alloy

Sensors 2020, 20, 1532 13 of 27 respectively [103,104]. In these alloys, it has been found that the giant magnetostriction always occurred in the vicinity of ordered D03 and disordered A2 phase boundaries. In addition, since the ﬁrst report on the magnetostrictive properties of Fe-Co alloy by Masiyama, much eﬀorts have also been devoted in developing giant magnetostrictive Fe-Co alloys [105].

3.2.1. FeGa Based Alloy To overcome the intrinsic drawbacks in RFe alloys, researchers began to search for other rear-earth-free magnetic materials with large magnetostriction, low saturation field, low cost and high mechanical durability. In 1999, the group at Carderock which was funded by the Office of Naval Research (ONR) invented a new exciting material, the Fe-Ga (Galfenol) alloy system. Galfenol alloys exhibit large saturation magnetostriction of 400 ppm in single crystal form and 300 ppm in polycrystalline form with a much lower saturation field of less than 200 Oe [106,107]. In addition, these alloys are stable over a wide temperature range from 265 C to 150 C and can sustain − ◦ ◦ approximately 350 MPa tensile stress. Galfenol alloys can be stress-annealed by applying a compressive force during the annealing process, which eliminates the need to apply external compressive stress for certain usage [108]. Although single crystal Fe-Ga alloys exhibit the advantage of large magnetostriction at relatively low saturation field, and can tolerate substantial compressive and tensile stress, however, they have large ferromagnetic resonance linewidth of 400–600 Oe at X band, which make them very lossy for integration into microwave devices [109]. Previous studies have found that the addition of trace amounts of metalloid atoms of C, B, and N into binary alloys has a beneficial effect on increasing magnetostriction. In 2007, Lou et al. reported magnetostrictive behavior of Fe-Ga-B thin films with different B contents [62]. As shown in Figure 11, through the incorporation of boron at 9–12 %, the gain size has been refined and the magnetocrystalline anisotropy has been effectively reduced, which results in the combination of excellent magnetic softness with a maximum saturation magnetostriction of 70 ppm at saturation field lower than 20 Oe. The reason for this, is that small amount of B atoms would stay in the interstitial sites of the bcc lattice and destabilize the D03 phase which is detrimental to magnetostriction, while higher amount of metalloid atoms would form clusters that decrease the magnetostriction constant [110]. It is also believed that a low concentration of metalloid atoms could lead to atomic pairs such as Ga–Ga or B–B, which enhance the magnetostrictive behavior. Beside boride Fe-Ga alloys, the magnetostrictive behavior of carbide Fe-Ga alloys have also been investigated. In 2018, Liang et al., reported the C-concentration dependence on magnetostriction of Fe-Ga-C films [61]. Figure 12 shows the saturation magnetostriction constants and maximum piezomagnetic coefficients of Fe-Ga-C films with different C content. The saturation magnetostriction λS reaches the maximum value of 81.2 ppm when C atoms incorporate into a Fe-Ga lattice with 11.1 at. % C content and continually reduces to 22 ppm with C composition of 14.2 at. %. The gradually increasing value of λS with the C content increasing from 0 at. % to 7.1 at. % is illustrated by the slowly transition from crystalline structures to amorphous state of Fe-Ga-C films. When the C content becomes more and more, the λS reaches its maximum value near the amorphous/crystalline phase boundary, where Fe-Ga-C films change from crystalline structure to amorphous phase. The maximum λS of Fe-Ga alloy systems occurs at the composition of Fe80Ga20, which is near the A2/D03 phase boundary and is consistent with Fe-Ga-C systems. The maximum λS value of Fe-Ga-C films is 81.2 ppm and is six times larger than that of the binary Fe-Ga film, while the maximum λS value of Fe-Ga-B systems only tripled, which means a better influence of C addition on enhancing magnetostriction constant of Fe-Ga films. Moreover, Fe-Ga-C films are more magnetic sensitive than that of Fe-Ga-B films due to their higher piezomagnetic coefficient of 9.71 ppm/Oe. Sensors 2020, 20, x FOR PEER REVIEW 13 of 29

To overcome the intrinsic drawbacks in RFe alloys, researchers began to search for other rear-earth- free magnetic materials with large magnetostriction, low saturation field, low cost and high mechanical durability. In 1999, the group at Carderock which was funded by the Office of Naval Research (ONR) invented a new exciting material, the Fe-Ga (Galfenol) alloy system. Galfenol alloys exhibit large saturation magnetostriction of 400 ppm in single crystal form and 300 ppm in polycrystalline form with a much lower saturation field of less than 200 Oe [106,107]. In addition, these alloys are stable over a wide temperature range from -265 °C to 150 °C and can sustain approximately 350 MPa tensile stress. Galfenol alloys can be stress-annealed by applying a compressive force during the annealing process, which eliminates the need to apply external compressive stress for certain usage [108]. Although single crystal Fe-Ga alloys exhibit the advantage of large magnetostriction at relatively low saturation field, and can tolerate substantial compressive and tensile stress, however, they have large ferromagnetic resonance linewidth of 400-6 00 Oe at X band, which make them very lossy for integration into microwave devices [109]. Previous studies have found that the addition of trace amounts of metalloid atoms of C, B, and N into binary alloys has a beneficial effect on increasing magnetostriction. In 2007, Lou et al. reported magnetostrictive behavior of Fe-Ga-B thin films with different B contents [62]. As shown in Figure 11, through the incorporation of boron at 9-12 %, the gain size has been refined and the magnetocrystalline anisotropy has been effectively reduced, which results in the combination of excellent magnetic softness with a maximum saturation magnetostriction of 70 ppm at saturation field lower than 20 Oe. The reason for this, is that small amount of B atoms would stay in the interstitial sites of the bcc lattice and destabilize the D03 phase which is detrimental to magnetostriction, while higher amount of metalloid atoms would form clusters that decrease the magnetostriction constant [110]. It is also believed thatSensors a low2020 concentration, 20, 1532 of metalloid atoms could lead to atomic pairs such as Ga–Ga or B–B,14 of which 27 enhance the magnetostrictive behavior.

SensorsFigure 2020Figure, 2011., x 11. (FORa) (Magnetostrictiona PEER) Magnetostriction REVIEW vs vs magnetic magnetic field. ﬁeld. ( (bb)) Saturation magnetostrictionmagnetostriction constant constant vs vs B content.B content.14 of 29 ReproducedReproduced with with permission permission from from [62]. [62 Copyright]. Copyright [2007], [2007], American American Institute Institute of Physics. of Physics.

Beside boride Fe-Ga alloys, the magnetostrictive behavior of carbide Fe-Ga alloys have also been investigated. In 2018, Liang et al., reported the C-concentration dependence on magnetostriction of Fe-Ga- C films [61]. Figure 12 shows the saturation magnetostriction constants and maximum piezomagnetic coefficients of Fe-Ga-C films with different C content. The saturation magnetostriction 휆푆 reaches the maximum value of 81.2 ppm when C atoms incorporate into a Fe-Ga lattice with 11.1 at. % C content and continually reduces to 22 ppm with C composition of 14.2 at. %. The gradually increasing value of 휆푆 with the C content increasing from 0 at. % to 7.1 at. % is illustrated by the slowly transition from crystalline structures to amorphous state of Fe-Ga-C films.

FiguFigurere 12. 12.(a) Magnetostriction(a) Magnetostriction constant constant of Fe- ofGa Fe-Ga-C.-C. (b) Saturation (b) Saturation magnetostriction magnetostriction constant constant and maximum and piezomagneticmaximum piezomagnetic coefficient of Fe coe-Gaffi-cientC. Reproduced of Fe-Ga-C. with Reproduced permission with from permission [61]. Copyright from [ 61[2019],]. Copyright IEEE. [2019], IEEE. When the C content becomes more and more, the 휆푆 reaches its maximum value near the amorphous/crystalline3.2.2. FeCo Based Alloys phase boundary, where Fe-Ga-C films change from crystalline structure to amorphousIt is well-knownphase. The maximum that Fe–Co 휆푆 alloyof Fe- systemsGa alloy havesystems very occurs soft magnetic at the composition properties. of Besides,Fe80Ga20, Fe-Co which is nearalloys the A2/D03 own the phase largest boundary saturation and magnetic is consistent flux density with Fe among-Ga-C thesystems. Fe, Co The and maximum Ni systems 휆푆 duevalue to of its Fe- Galargest-C films number is 81.2 ofppm Bohr and magnetons is six times according larger than to the that Slater–Pauling of the binary curve Fe-Ga [111 film,]. while the maximum 휆푆 value ofIt isFe reported-Ga-B systems that the only magnetostriction tripled, which of meansFe–Co binary a better alloys influence greatly of rely C onaddition their composition on enhancing magnetostrictionover a wide range constant in Masiyama’s of Fe-Ga films. researches. Moreover, There Fe-Ga were-C films two obviousare more peaks magnetic of magnetostriction sensitive than that of Fe-Ga-B films due to their higher piezomagnetic coefficient of 9.71 ppm/Oe. that were clearly measured, the 1st peak being around the composition of Fe30Co70 (90 ppm) and the other one being around the composition of Fe60Co40 (70 ppm) [105]. From the phase diagram of 3.2.2.the FeCo Fe-Co Bas binaryed Alloys systems, we can see a phase boundary between the bcc structure and its coexisting regionIt is well with-known fcc phase that atFe a– Co alloy content systems of ~75 have at. %.very It soft was magnetic found in properties. the previous Besides, report Fe that-Co thealloys ownmagnetostriction the largest saturation of Fe-Co magnetic systems flux could density be enhanced among bythe forming Fe, Co and a phase Ni systems boundary, due which to its islargest analogous number of Bohrto the magnetons magnetostriction according augment to the ofSlat Fe-Gaer–Pauling alloys curve near the [111 D03]. /A2 phase boundary. And they indeed achievedIt is reported the largest that the magnetostriction magnetostriction at a of Co Fe content–Co binary of ~70 alloys at. % greatly in that rely report. on their composition over a wide In range 2019, in Wang Masiyama’s investigated researches. the magnetostrictive There were two behavior obvious of peaks (Co0.5 ofFe magnetostriction0.5)xC1-x alloy thin that films were clearlywith measured, different carbon the 1st content peak being [112 ].around Figure the 13 compositionpresents the magnetostrictionof Fe30Co70 (90 ppm) results and ofthe these other Co-Fe-C one being aroundfilms the with composition increasing carbonof Fe60Co content40 (70 ppm) as a function [105]. From of the the in-plane phase diagram driving magnetic of the Fe field.-Co binary There systems, is no wemagnetostrictive can see a phase boundary response for between those samplesthe bcc structure with carbon and content its coexisting lower than regio 4%n with when fcc the phase magnetic at a Co contentfield of is ~ less 75 at. than %. 30It was Oe. found However, in the when previous the carbonreport that content the magnetostriction of Co-Fe-C films of is Fe higher-Co systems than 4%,could be theenhanced AC driving by forming field reduces a phase to boundary, 20 Oe and which gradually is analogous becomes smaller.to the magnetostriction This magnetostrictive augment behavior of Fe-Ga alloysindicates near the that D03/A2 the carbon phase content boundary. of 4% And is athey critical indeed value achieved for the the Co-Fe-C largest films’ magnetostriction phase boundary at a Co content of ~ 70 at. % in that report. In 2019, Wang investigated the magnetostrictive behavior of (Co0.5Fe0.5)xC1-x alloy thin films with different carbon content [112]. Figure 13 presents the magnetostriction results of these Co-Fe-C films with increasing carbon content as a function of the in-plane driving magnetic field. There is no magnetostrictive response for those samples with carbon content lower than 4% when the magnetic field is less than 30 Oe. However, when the carbon content of Co-Fe-C films is higher than 4%, the AC driving field reduces to 20 Oe and gradually becomes smaller. This magnetostrictive behavior indicates that the carbon content of 4% is a critical value for the Co-Fe-C films’ phase boundary between nanocrystalline and amorphous structures. As shown in Figure 13b, black squares denote the magnetostriction constants of Co-Fe-C films.

The saturation magnetostriction 휆푆 reaches a maximum value of 75ppm when carbon atoms are added to a Fe-Co lattice with 5.2 at. % carbon content, and then gradually reduces to 10 ppm at carbon content of 15.8 at. %.

Sensors 2020, 20, 1532 15 of 27

between nanocrystalline and amorphous structures. As shown in Figure 13b, black squares denote the magnetostriction constants of Co-Fe-C ﬁlms. The saturation magnetostriction λS reaches a maximum

Sensorsvalue 2020 of, 20 75ppm, x FOR whenPEER REVIEW carbon atoms are added to a Fe-Co lattice with 5.2 at. % carbon content,15 and of 29 then gradually reduces to 10 ppm at carbon content of 15.8 at. %.

FigureFigure 13. 13.(a) ( Magnetostrictiona) Magnetostriction constant constant of (Co of (Co0.5Fe0.50.5Fe)xC0.51−x)x filmsC1 x films with with x=0 to x = 15.8%.0 to 15.8%. (b) The( b saturated) The − magnesaturatedtostriction magnetostriction constant (black) constant and piezomagnetic (black) and piezomagnetic coefficient (red) coeffi withcient different (red) with carbon different content. carbon The structuralcontent. boundaries The structural of the boundaries alloy are marked of the alloy with are light marked orange with and light green orange dotted and lines. green (c) dottedPiezomagnetic lines.

coefficient(c) Piezomagnetic versus corresponding coefficient versusdriving corresponding magnetic field driving for (Co magnetic0.5Fe0.5)xC1−x field films for with (Co0.5 x=0Fe 0.5to) 13.2%.xC1 x films (d) The − Ms with(black) x = and0 to H 13.2%.c (red) ( dversus) The Mcarbons (black) content. and H cThe(red) blue versus open carbon triangles content. are the The fitted blue openMs from triangles broadband are FMRthe measurement. fitted Ms from Reproduced broadband FMRwith permission measurement. from Reproduced [112]. Copyright with permission [2019], American from [112 Physical]. Copyright Society. [2019], American Physical Society. The piezomagnetic coefficient of the Co-Fe-C films was shown by the red cycles in Figure 13b. The curve canThe be piezomagnetic divided into three coeffi regions:cient of When the Co-Fe-C the carbon films content was shown is less by than the red4 at. cycles %, the in Co Figure-Fe-C 13 filmsb. remainThe curveto be canbcc benanocrystalline divided into threephase; regions: as the Whencarbon the content carbon increases content isand less is thanless 4than at. %,6 at. the %, Co-Fe-C there is a coexistencefilms remain region to beof both bcc nanocrystalline nanocrystalline phase; and amorphous as the carbon phase content and the increases phase boundaries and is less than are labelled 6 at. %, as thethere dotted is a orange coexistence and region green lines;of both when nanocrystalline the carbon and content amorphous goes beyond phase and 6 at. the %, phase they boundaries transition to amorphousare labelled structure. as the dotted A clear orange peak and is obtained green lines; in thewhen coexistence the carbon region content and goes the beyond largest 6 at.value %, of piezomagneticthey transition coefficient to amorphous is achieved structure. of 10.3 A ppm/Oe clear peak at is4.8 obtained at. %. They in the also coexistence confirmed region the coexistence and the regionlargest by valueanalyzing of piezomagnetic FFT results of coe thefficient HRTE isM achieved images, of which 10.3 ppm are/ Oeshown at 4.8 in at. Figure %. They 14. alsoThe confirmedformation of atomthe pairs coexistence or clusters region during by analyzing the doping FFT of resultsmetalloid of the atoms HRTEM also destabilize images, which the D03 are shown phase. inTherefore, Figure 14 the. magnetostrictionThe formation constants of atom pairsand piezomagnetic or clusters during coefficients the doping of Co of-Fe metalloid-C films are atoms optimized also destabilize with the carbon the contentD03 phase.ranging Therefore, from 4% theto 6%. magnetostriction Compared to constantsthe reported and value piezomagnetic of magnetostriction coefficients constants of Co-Fe-C for films Ga-Fe- B andare the optimized widely withused theCo- carbonFe-B films, content the values ranging measured from 4% in to this 6%. paper Compared are much to the higher. reported Moreover, value of the maximummagnetostriction piezomagnetic constants coefficient for Ga-Fe-B dλ/dH and theof widelyCo-Fe-C used films Co-Fe-B is much films, larger the values than measured other well in-known this magnetostrictivepaper are much materials higher. Moreover,such as Metglas the maximum and Terfenol piezomagnetic-D. coefficient dλ/dH of Co-Fe-C films is much larger than other well-known magnetostrictive materials such as Metglas and Terfenol-D.

SensorsSensors 20202020, 20,, x20 FOR, 1532 PEER REVIEW 16 of16 29 of 27

FigureFigure 14. ( 14.a) A bright(a) A-field bright-ﬁeld STEM image STEM of image Ta/(Co,Fe)C/Ta/SiO2/Si of Ta/(Co,Fe)C/Ta (001)/SiO2 substrate./Si (001) ( substrate.b) The corresponding (b) The energycorresponding-dispersive energy-dispersivex-ray spectroscopy x-ray mappings spectroscopy for Fe, mappingsCo, and Ta for elements, Fe, Co, and respectively. Ta elements, (c)respectively. Cross sections of lowresolution(c) Cross sections TEM ofwith lowresolution white solid TEMrectangle with displaying white solid the rectangle corresponding displaying location the of corresponding HRTEM. (d)–(f) HRTEMlocation images of HRTEM. of Co-Fe (d-C–f )alloy HRTEM films images with carbon of Co-Fe-C content alloy 0%, ﬁlms 4.4%, with and carbon 13.2%, contentrespectively. 0%, 4.4%, The andFFT of selected13.2%, zones respectively. (yellow dotted The FFT squares) of selected are inserted zones (yellow at the dottedbottom squares) left corner. are Reproduced inserted at the with bottom permission left fromcorner. [112]. ReproducedCopyright [2019], with permissionAmerican Physical from [112 Society.]. Copyright [2019], American Physical Society.

3.3.3.3. Exchange Exchange Coupling Coupling in Multilayers in Multilayers In additionIn addition to the to state the state-of-the-art-of-the-art magnetostrictive magnetostrictive single single layer layermaterials, materials, the combination the combination of different of materialsdifferent in materialsmultilayer in structure multilayer has structure recently hasbeen recently proposed been as proposed an alternative as an approach alternative to approach develop giant to magnetostrictivedevelop giant magnetostrictivethin films with low thin loss films and with low low saturation loss and low field. saturation In multi field.layer In thin multilayer films, the thin magnetic films, propertiesthe magnetic in different properties layers in dicanfferent be synthesized layers can be through synthesized the process through of the exchange process of coupling, exchange which coupling, arises whenwhich the arisesthickness when in theeach thickness layer is smaller in each than layer the is smallerlength of than magnetic the length exchange of magnetic [113]. Through exchange exchange [113]. coupling,Through spin exchange information coupling, can be spin transmitted information between can be two transmitted different magnetic between twomaterials, different and magnetic the overall magneticmaterials, properties and the are overall determined magnetic by properties the average are valuedetermined of each by layer. the average By combining value of two each different layer. materialsBy combining with excellent two di ffsoftnesserent materials and giant with magnetostriction, excellent softness exchang and giante coupling magnetostriction, can be used to exchange reduce the saturationcoupling field can and be used improve to reduce the piezomagnetic the saturation coefficient. field and improve the piezomagnetic coefficient.

3.3.1. FeGa/NiFe Multilayers 3.3.1. FeGa/NiFe Multilayers

In 2019, Shi et al. studied the effect of exchange coupling between Fe83Ga17 and Fe50Ni50 multilayer In 2019, Shi et al. studied the effect of exchange coupling between Fe83Ga17 and Fe50Ni50 multilayer films films for enhancement of the ME effect [114]. As shown in Figure 15, all of the magnetic domain for enhancement of the ME effect [114]. As shown in Figure 15, all of the magnetic domain structures are structures are stripe domain. However, with the total film thickness fixed at 480 nm, as the thickness stripe domain. However, with the total film thickness fixed at 480 nm, as the thickness in each layer of Fe- in each layer of Fe-Ga and Fe-Ni become thinner, the stripe width become narrower. This means the Ga and Fe-Ni become thinner, the stripe width become narrower. This means the domain density has been domain density has been increased, which intensified the domain rotation and domain wall motion, increased, which intensified the domain rotation and domain wall motion, thereby enlarged the thereby enlarged the magnetostriction. It is believed that stripe domain structure indicates the films magnetostriction. It is believed that stripe domain structure indicates the films have weak out-of-plane have weak out-of-plane anisotropy [115], and the magnetic moments on the surface of the films rotate anisotropy [115], and the magnetic moments on the surface of the films rotate upward and downward upward and downward alternatively. With the Fe-Ni layers, the permeability increased and the total alternatively. With the Fe-Ni layers, the permeability increased and the total magnetocrystalline anisotropy magnetocrystalline anisotropy reduced, which in turn make the 90-degree domain wall motion easier. reduced, which in turn make the 90-degree domain wall motion easier. For the (8 nm Fe-Ga/8 nm Fe-Ni)30 For the (8 nm Fe-Ga/8 nm Fe-Ni) multilayers, the small magnetic anisotropy resulting from the multilayers, the small magnetic anisotropy30 resulting from the exchange coupling makes the domains rotate exchange coupling makes the domains rotate more easily under low magnetic field. The rotation more easily under low magnetic field. The rotation of the magnetic domains in the Fe-Ni layers also drive of the magnetic domains in the Fe-Ni layers also drive the rotation of the magnetic domains in the the rotation of the magnetic domains in the Fe-Ga layers, which makes the Fe-Ni layers act like a lubricant. Fe-Ga layers, which makes the Fe-Ni layers act like a lubricant. As shown in Figure 16b, the d33,m of As shown in Figure 16b, the d33,m of the (8 nm Fe-Ga/8 nm Fe-Ni)30 multilayers achieved to 8.1 ppm/Oe. the (8 nm Fe-Ga/8 nm Fe-Ni)30 multilayers achieved to 8.1 ppm/Oe. Compared to the Fe-Ga-B film, the piezomagnetic coefficient of the multilayer film has an increment of about 14%.

Sensors 2020, 20, x FOR PEER REVIEW 17 of 29

ComparedSensors 2020 to, 20the, 1532 Fe-Ga-B film, the piezomagnetic coefficient of the multilayer film has an increment17 of 27 of aboutCompared 14%. to the Fe-Ga-B film, the piezomagnetic coefficient of the multilayer film has an increment of about 14%.

FigureFigure 15. 15.( 15.a), ( a(a–db), ),(b ()),c are) ( andc) theand (d surface ()d are) are the magneticthe surface surface domainsmagnetic magnetic of domains domains (480 nm ofof Fe-Ga), (480 nmnm (30 FeFe nm--Ga),Ga), Fe-Ga (30 (30 /nm 30nm nmFe Fe-Ga/30 Fe-Ni)-Ga/30 nm8 ,nm (16 Fe Fe- -

Ni)8Ni)nm, (168, Fe-Ga (16nm nm Fe/16 -FeGa/16 nm-Ga/16 Fe-Ni) nm nm 15Fe Feand-Ni)-Ni)15 (8 15and nm and Fe-Ga(8 (8 nm nm/ 8Fe Fe nm-Ga/8-Ga/8 Fe-Ni) nmnm30 FeFeﬁlms.--Ni)Ni)30 Reprinted films. ReprintedReprinted with permission with with permission permission from [ 114from from]. [114].[114].Copyright Copyright Copyright [2019], [2019], [2019], Elsevier. Elsevier. Elsevier.

Figure 16. 16. (a(a,b) and) are (b)λ are-H andλ-H dandλ/dH-H dλ/dH curves-H curves of the of the samples, samples,λs ofλs theof the other other samples samples are are 137 137 ppm, ppm, 108 108 Figure 16. (a) and (b) are λ-H and dλ/dH-H curves of the samples, λs of the other samples are 137 ppm, 108 ppm,115 ppm and 97 ppm, d33,m of the (8 nm Fe-Ga/8 nm Fe-Ni)30 reaches 8.1ppm/Oe. Reprinted with ppm,115 ppm and 97 ppm, d33,m of the (8 nm Fe-Ga/8 nm Fe-Ni)30 reaches 8.1ppm/Oe. Reprinted with ppm,115 ppm and 97 ppm, d33,m of the (8 nm Fe-Ga/8 nm Fe-Ni)30 reaches 8.1ppm/Oe. Reprinted with permission from from [114 [114].]. Copyright Copyright [2019], [2019], Elsevier. Elsevier. permission from [114]. Copyright [2019], Elsevier. 3.3.2. TbFe/FeCo TbFe/FeCo Multilayers Multilayers 3.3.2. TbFe/FeCo Multilayers In 1999, Quandt et et at. reported reported exchange exchange coupled coupled Tb Tb4040FeFe60/Fe60/Fe50Co5050Co multilayers50 multilayers which which exhibits exhibits high highsaturationIn 1999, saturation magnetoelasticQuandt magnetoelastic et at. reportedcoupling coupling coefficexchangeients coe coupledinffi combinationcients Tb in40Fe combination with60/Fe 50lowCo saturation50 multilayers with low field saturation which of 20 mT exhibits [116,117 field ofhigh]. saturation20It has mT been [ 116magnetoelastic ,found117]. Itthat has the beencoupling saturation found coeffic that magnetostrictionients the saturation in combination was magnetostriction strongly with lowdepended saturation was stronglyon the field layer depended of 20thickness mT [116,117 on ratio the ]. It haslayerand been the thickness postfound annealing that ratio the and tre saturationatment, the post as annealingmagnetostriction shown in treatment,Figure 17wasc, asd. strongly shown independed Figure 17 onc,d. the layer thickness ratio and the post annealing treatment, as shown in Figure 17c,d. The optimum data of the TbFe/FeCo multilayers show that post annealing at 250◦C improves the magnetoelastic behavior, and the coercivity has been reduced from 10 mT to 2 mT, which reflects that post annealing can release almost all stress induced anisotropies, leading to a stress-free state of these multilayers. It is also believed that under these conditions any significant diffusion does not exist and nanocrystallization begin to appear in the TbFe layers, which leads to the enhancement of the saturation magnetoelastic coupling coefficient by 15%. At the same time, the overall temperature behavior is also compromised by the exchange coupled multilayers, and it is expected that the strong Fe50Co50 moments can help stabilized the Tb40Fe60 moments, resulting in an increment of the Curie temperature.

SensorsSensors 20202020, 20,, x20 FOR, 1532 PEER REVIEW 18 of18 29 of 27

40 60 FigureFigure 17. 17.(a) (Compaa) Comparisonrison of of the the magnetoelastic magnetoelastic coupling coecoefficientfficient of of 250 250◦C °C annealed annealed (7 nmTb(7 nmTb40Fe60Fe/9 /9 nmFenmFe50Co50Co) × 70) multilayer70 multilayer with with an an optimized optimized Tb Tb40FeFe60 andand a a SmFeB SmFeB film. film. (b (b)) Comparison Comparison of of the the 50 50 × 40 60 magnetoelasticmagnetoelastic coupling coupling coefficients coefficients of of (Tb (Tb4040FeFe60/Fe60/Fe50Co50Co50),50 (Tb), (Tb18Co1882Co /Fe82/75FeC75o25Co), 25(TbDyFe/), (TbDyFe FeSiCuNbB)/FeSiCuNbB) and (Tb33andFe67 (Tb /Fe33Fe80B6720)/ Fe multilayer80B20) multilayer films. films. (c) Saturation (c) Saturation magnetoelastic magnetoelastic coupling coupling coe coefficientfficient of of TbFe TbFe/FeCo/FeCo multilayersmultilayers as a as function a function of ofthe the soft soft magnetic magnetic layer layer thickness. thickness. (d ()d ∆) E∆ Eeffect effect of of as as-deposited-deposited and and at at 250 250 °C

annealed◦C annealed (7 nm (7 TbFe/8 nm TbFe nm/ 8 FeCo) nm FeCo) multilayers. multilayers. There There figures ﬁgures are arereproduced reproduced with with the the permission permission from [116,117from]. [ 116Copyright,117]. Copyright [1999] and [1999] [2000], and American [2000], American Physical PhysicalSociety. Society.

3.4.The Heat optimum Treatment data on Magnetostrictiveof the TbFe/FeCo Thin multilayers Films show that post annealing at 250°C improves the magnetoelastic behavior, and the coercivity has been reduced from 10 mT to 2 mT, which reflects that post 3.4.1. Stress Release by Post Annealing annealing can release almost all stress induced anisotropies, leading to a stress-free state of these multilayers.As a Ithighly is also magnetostrictive believed that under material, these FeCoSiB conditions is one any of the significant most commonly diffusion used does composition not exist and nanocrystallizationin Metglas foils. begin In 2002, to appear Quandt in the et al.TbFe investigated layers, which the leads optimization to the enhancement of the ∆E eofff ectthe andsaturation the magnetoelasticmagnetostriction coupling of amorphous coefficient (Fe by90 Co15%.10) 78AtSi 12theB10 samethin filmstime, bythe magnetic overall temperature field annealing behavior [118]. They is also compromiseddiscovered by a strong the exchange correlation coupled between multilayers, the magnetostrictive and it is expected susceptibility that the strong and the Fe stress50Co50 state, moments which can helpcan stabilized be controlled the Tb by40Fe the60 annealingmoments, temperature.resulting in an As increment shown in Figureof the Curie18c, the temperature. as-deposited FeCoSiB film was initially in compressive stress state before the annealing temperature reach to 170 ◦C. However, as 3.4.the Heat annealing Treatment temperature on Magnetostrictive continue Thin to increase,Films the stress state gradually switched to tensile stress. After annealing treatment, the compressive stress in the film has been effectively reduced, leading to the 3.4.1.reduction Stress Release of anisotropy by Post energy. Annealing The change of the Young’s modulus with applied magnetic field has beenAs a shown highly in magnetostrictive Figure 18d. After material, magnetic FeCoSiB annealing, is one a distinguished of the most Young’scommonly modulus used composition reduction of in Metglas50 GPa foils. has In been 2002, observed, Quandt which et al. investigated contributed tothe 30% optimization change of the of the Young’s ∆E ef modulusfect and the at saturation magnetostriction state, assuming the Young’s modulus of FeCoSiB thin film at saturation is 150 GPa. In Figure 18a,b, we can of amorphous (Fe90Co10)78Si12B10 thin films by magnetic field annealing [118]. They discovered a strong correlationsee the magnitude between the and magnetostrictive the field dependency susceptibility of the magnetostrictive and the stress state, susceptibility which can are be highly controlled depends by the annealingon the stresstemperature. state and As the shown corresponding in Figure 18 annealingc, the as- temperature.deposited FeCoSiB To achieve film was high initially magnetostriction in compressive at stresslow state magnetic before fields, the annealing a precise control temperature of film reach stress to has 170 to be °C. executed. However, For as positive the annealing magnetostrictive temperature continuethin films, to increase, usually tensilethe stress stress state state gradually is favorable switched to obtain to tensile the optimum stress. After performance. annealing treatment, the compressive stress in the film has been effectively reduced, leading to the reduction of anisotropy energy. The change of the Young’s modulus with applied magnetic field has been shown in Figure 18d. After magnetic annealing, a distinguished Young’s modulus reduction of 50 GPa has been observed, which

Sensors 2020, 20, x FOR PEER REVIEW 19 of 29

FeCoSiB thin film at saturation is 150 GPa. In Figure 18 a,b, we can see the magnitude and the field dependency of the magnetostrictive susceptibility are highly depends on the stress state and the corresponding annealing temperature. To achieve high magnetostriction at low magnetic fields, a precise controlSensors of2020 film, 20 stress, 1532 has to be executed. For positive magnetostrictive thin films, usually tensile stress19 of 27state is favorable to obtain the optimum performance.

FigureFigure 18. 18.(a) (Magnetoelastica) Magnetoelastic coupling coupling coefficient coefficient (b) (negativeb) negative absolute absolute value value of the of thefirst first derivative derivative of the magnetostrictiveof the magnetostrictive hysteresis hysteresis (c) stress (asc) stressa function as a functionof annealing of annealing temperature temperature (d) ∆E effect (d) ∆ Eof eanffect optimized of an

annealedoptimized (Fe90 annealedCo10)78Si12 (FeB1090 thinCo 10film.)78Si Reproduced12B10 thin film. with Reproduced permission with from permission [118]. Copyright from [118 [2002],]. Copyright IEEE. [2002], IEEE. 3.4.2. Phase Boundary Changes by Cooling Processes 3.4.2. Phase Boundary Changes by Cooling Processes In 2011,In 2011, Hunter Hunter et al. et investigated al. investigated the themagnetostrictive magnetostrictive behavior behavior and and microstructural microstructural properties properties of Co- Fe of alloy Co-Fe thin alloy films thin in films relation in relation to the to composition the composition and and thermal thermal process process [119 [119].]. Strong Strong dependence dependence of magnetostrictionof magnetostriction on the on cooling the cooling process process has been has been found. found. The analysis The analysis of the of microstructure the microstructure through synchrotronthrough synchrotronX-ray micro- X-raydiffraction micro-di showsffraction that the shows maximum that themagnetostriction maximum magnetostriction occurs at the (fcc+bcc)/bcc occurs phaseat theboundary. (fcc+bcc) As/bcc shown phase in Figure boundary. 19 a,b, As after shown annealing, in Figure the peak19a,b, of aftermagnetostriction annealing, theshifts peak to lower of Co magnetostrictioncomposition by the shifts same to amount lower as Co the composition (fcc + bcc)/bcc by phase the sameboundary amount shifts, as indicating the (fcc +thatbcc) the/bcc peak magnetostrictionphase boundary is linked shifts, to indicating this phase that interface. the peak At the magnetostriction temperature/composition is linked to close this to phase the (fcc interface. + bcc)/bcc phaseAt the boundary, temperature a remarkable/composition giant close magnetostriction to the (fcc + bcc) has/bcc been phase yield boundary, after annealing a remarkable followed giant by quenching,magnetostriction which is hasmore been than yield three after times annealing of its as•deposited followed by quenching,state. which is more than three times of its as-deposited state. The Co-Fe phase diagrams in Figure 19 b show that the bcc phase intersects with a mixed phase region of fcc Co and bcc Fe phases at temperatures lower than 912 ◦C and Co concentrations greater than 50%. The coexistence of the predominant bcc phase and a secondary fcc phase has also been confirmed by the TEM data shown in Figure 20c,d. It is believed that the precipitation of the fcc Co-rich grains into the bcc α-Fe matrix is the cause of the enhancement in magnetostriction, which function in the same way as the D03 precipitates in the Fe–Ga alloys.

Sensors 2020, 20, x FOR PEER REVIEW 20 of 29

SensorsSensors 20202020, 20,, x20 FOR, 1532 PEER REVIEW 20 of20 29 of 27

Figure 19. (a) Magnetostriction variation versus atomic percent cobalt for three differently prepared Co1−xFex composition spreads, as-deposited (black dots), slow-cooled (blue dots), quenched (red dots). (b) Co–Fe- phase diagram. The red curve highlights the approximate phase boundary between (fcc Co + bcc Fe) and bcc Fe. (c)-(e) Intensity plots of the synchrotron microdiffraction of Co1−xFex thin films. (c) as-deposited, (d) annealed and slow-cooled, (e) annealed and water-quenched composition spread samples. The black line

marked λmax in each spread indicates the approximate composition of the (fcc + bcc)/bcc phase boundary. FigureReprintedFigure 19. ( 19.witha) Magnetostriction(a permission) Magnetostriction from variation [119 variation]. Copyright versus versus atomic [2011], atomic percent Springer percent cobalt Nature. cobalt for three for threedifferently diﬀerently prepared prepared Co1−xFex

compositioCo1 xFenx compositionspreads, as-deposited spreads, as-deposited (black dots), (black slow- dots),cooled slow-cooled (blue dots), (blue quenched dots), quenched(red dots). (red (b) dots). Co–Fe- − phaseThe(b Co) diagram. Co–Fe-phase-Fe phase The diagrams diagram.red curve Thein highlights Figure red curve 19 the b highlights showapproximate that the the approximate phase bcc phase boundary phaseintersects between boundary with (fcc betweena Co mixed + bcc (fcc phase Fe) Co and region of fcc Co and bcc Fe phases at temperatures lower than 912 °C and Co1−x concentratix ons greater than 50%. The bcc+ Fe.bcc (c Fe))-(e) and Intensity bcc Fe. plots (c–e )of Intensity the synchrotron plots of themicrodiffr synchrotronaction microdiof Co ﬀFeraction thin films. of Co 1(c)x Feas-xdeposited,thin ﬁlms. (d) − coexistenceannealed(c) as-deposited, of and the slowpredominant-cooled, (d) annealed (e )bcc annealed and phase slow-cooled, and and water a secondary (e-quenched) annealed fcc composition and phase water-quenched has spreadalso been samples. composition confirmed The spreadblack by thlinee TEM data markedshownsamples. inλmax Figure Thein each black 20c, spread lined. It marked indicatesis believedλmax thein thatapproximate each the spread precipitation indicatescomposition the of approximateofthe the fcc (fcc Corich + bcc)/bcc composition grains phase into of boundary the the (fcc bcc. αFe matrixReprinted is+ bcc)the /bcccaus with phasee permission of boundary.the enhancement from Reprinted [119]. Copyrightin with magnetostriction, permission [2011], Springer from [which119 Nature.]. Copyright function [2011],in the Springer same way Nature. as the D03 precipitates in the Fe–Ga alloys. The Co-Fe phase diagrams in Figure 19 b show that the bcc phase intersects with a mixed phase region of fcc Co and bcc Fe phases at temperatures lower than 912 °C and Co concentrations greater than 50%. The coexistence of the predominant bcc phase and a secondary fcc phase has also been confirmed by the TEM data shown in Figure 20c,d. It is believed that the precipitation of the fcc Corich grains into the bcc αFe matrix is the cause of the enhancement in magnetostriction, which function in the same way as the D03 precipitates in the Fe–Ga alloys.

FigureFigure 20. 20.TEMTEM images images of Co0.73Fe0.27 of Co0.73Fe0.27 of ( ofa) (110) (a) (110) dark dark field field image image of as of-deposited as-deposited sample, sample, the thescale scale bar is 50 nm.bar is(b 50) corresponding nm. (b) corresponding SAED pattern SAED of pattern the as-deposited, of the as-deposited, (c) bright (fieldc) bright image field of imageannealed of annealed sample, the sample, the scale bar is 200 nm. (d) SAED pattern of the same sample as (c) using ~1.5-µm-diameter aperture showing the mixture structure of bcc and fcc phases, (e)[11] bcc diffraction pattern from grain marked ‘A’ in (c), and (f)[11] fcc diffraction pattern from grain marked ‘B’ in (c). Reprinted with

permission from [119]. Copyright [2011], Springer Nature. Figure 20. TEM images of Co0.73Fe0.27 of (a) (110) dark field image of as-deposited sample, the scale bar is 4.50 Summary nm. (b) corresponding and Conclusions SAED pattern of the as-deposited, (c) bright field image of annealed sample, the Although lots of studies have been done on the magnetostrictive thin films, there are still many challenges and opportunities need to be addressed in the coming years. The essential figure of merits in developing thin film magnetoelastic materials are the piezomagnetic coefficient, coercivity, and FMR Sensors 2020, 20, x FOR PEER REVIEW 21 of 29

scale bar is 200 nm. (d) SAED pattern of the same sample as (c) using ~1.5-µm-diameter aperture showing the mixture structure of bcc and fcc phases, (e) [011] bcc diffraction pattern from grain marked ‘A’ in (c), and (f) [011] fcc diffraction pattern from grain marked ‘B’ in (c). Reprinted with permission from [119]. Copyright [2011], Springer Nature.

4. Summary and Conclusions

SensorsAlthough2020, 20 , lots 1532 of studies have been done on the magnetostrictive thin films, there are still21 of many 27 challenges and opportunities need to be addressed in the coming years. The essential figure of merits in developing thin film magnetoelastic materials are the piezomagnetic coefficient, coercivity, and FMR linewidth.linewidth. Figure Figure 21 summarizes21 summarizes the thecurrent current status status of the of thedeveloped developed magnetostrictive magnetostrictive materials materials in terms in of terms these offigure these of figure merits. of merits. The direction The direction of developing of developing desired desired magnetostrictive magnetostrictive materials materials goes goes to to high saturationhigh saturation magnetostriction, magnetostriction, high high piezomagnetic piezomagnetic coef coeficientfficient with with low low coercivity coercivity and and narrow narrow FMR FMR linewidth.linewidth. The The light light rare rare-earth-based-earth-based alloys, alloys, such such as SmFe, as SmFe, tend tend to be to the be thefocus focus of researches of researches in rare in rare earth magnetostrictiveearth magnetostrictive materials materials due to due the to lack the lack of of heavy heavy rare rare earth earth sources. sources. As As for for rare-earth-free rare-earth-free magnetostmagnetostrictiverictive materials, materials, Fe- Fe-GaGa and and Co Co-Fe-Fe systems systems with with large large magnetostriction magnetostriction and and low low saturation saturation field arefield found are to found be the to most be the promising. most promising. New structures New structures such as such multilayers as multilayers that combine that combine properties properties of high magnetostrictionof high magnetostriction and low anisotropy and low anisotropy are promising are promising for developing for developing magnetoelastic magnetoelastic materials materials with both softnesswith bothand softnesslarge magnetostriction and large magnetostriction constants. The constants. stress state The and stress the state phase and boundary the phase can boundary also be cantuned to achievealso be tunedthe optimal to achieve magnetostrictive the optimal magnetostrictivethin films through thin heat films treatment through process. heat treatment process.

FigureFigure 21.21. MagnetostrictiveMagnetostrictive materials materials in in summary summary of of (a (a)) maximum maximum piezomagnetic piezomagnetic coe coefficientﬃcient d33,m d33,mvs vs coercivecoercive field ﬁeld Hc ( Hbc) (saturationb) saturation magnetostriction magnetostriction λs vsλs FMRvs FMR linewidth. linewidth.

In summary,In summary,during during last last decades, decades, extensive extensive progress hashas been been made made for for achieving achieving high-performance high-performance magnetostrictivemagnetostrictive materials materials and and different different kinds kinds of of thin filmfilm basedbased integrated integrated multiferroic multiferroic devices. devices. In In this this reviewreview paper, paper, we we primarily primarily devoted devoted our our efforts efforts on on the the thin thin film film types types of of magnetostrictive magnetostrictive materials materials and theirand applications their applications of RF MEMS of RF MEMS devices devices based basedon strain onstrain-mediatedmediated ME coupling, ME coupling, which which have havedrawn drawn a great deala greatof interests deal of in interests the past inyears. the pastThe performance years. The performance of these RF MEMS of these devices RF MEMS was devicesgreatly benefited was greatly from thebenefited large piezomagnetic from the large coefficients piezomagnetic of the magnetostrictive coefficients of the materials, magnetostrictive which leads materials, to a stron whichg ME leads coupling. to a Fromstrong the science ME coupling. point of From view, the understanding, science point modelling of view, understanding, and characterizing modelling of thin andfilm characterizingmagnetostrictive materialsof thin filmare of magnetostrictive great importance materials for the are multiferroic of great importance society. From for the the multiferroic materials society.point of From view, the new magnetostrictivematerials point materials of view, newwith magnetostrictivehigher piezomagnetic materials coefficients with higher are always piezomagnetic more than coe welcome.fficients areFrom thealways perspective more of than engineering welcome. applications, From the perspective integration of engineering and scale-up applications, with existing integration silicon processing and scale-up flows andwith electronics existing are silicon in particularly processing valuable. flows and Low electronics-temperature are in process particularly methods valuable. are desired Low-temperature for fabricating RF processmultiferroic methods devices are desiredwith low for loss fabricating as well. In RF addition, multiferroic magnetic devices sensors with lowhave loss been as introduced well. In addition, long ago magnetic sensors have been introduced long ago to electrical and biomedical research areas and various types of magnetic field sensors are proposed [120]. Recently, flexible magnetic sensors that based on various mechanisms, such as anisotropic magnetoresistance (AMR) [121], giant magnetoimpedance (GMI) [122], etc. have received significant research interests because of their applications in wearable and portable devices. Magnetic field sensors that are fabricated on flexible substrates, which can be attached or transferred into the human body, grant us more freedom to design different healthcare devices. Therefore, flexible thin film magnetoelastic materials that can be compatible with wearable electronics are also significant for developing functional device applications. Through the better understanding in the origination of magnetostriction and the loss mechanism of magnetostrictive thin films, there exist a bright future and great challenges for thin film magnetoelastic materials. These materials will very likely facilitate the development of novel integrated RF multiferroic devices, which are compact, lightweight, and power efficient and will have huge influences on our daily life. Sensors 2020, 20, 1532 22 of 27

Funding: This research was funded by the NSF TANMS ERC Award 1160504 and the material is based upon work supported by the U.S. Army Small Business Innovation Research Program Office and the Army Research Office under Contract No. W911NF-18-C-0095. Acknowledgments: Microfabrication was performed in the George J. Kostas Nanoscale Technology and Manufacturing Research Center. The authors would like to thank the staff of the George J. Kostas Nanoscale Technology and Manufacturing Research Center at Northeastern University. Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

References

1. Nan, C.-W.; Bichurin, M.; Dong, S.; Viehland, D.; Srinivasan, G. Multiferroic magnetoelectric composites: Historical perspective, status, and future directions. J. Appl. Phys. 2008, 103, 1. [CrossRef] 2. Ma, J.; Hu, J.; Li, Z.; Nan, C.W. Recent progress in multiferroic magnetoelectric composites: From bulk to thin films. Adv. Mater. 2011, 23, 1062–1087. [CrossRef][PubMed] 3. Srinivasan, G. Magnetoelectric composites. Annu. Rev. Mater. Res. 2010, 40, 153–178. [CrossRef] 4. Sun, N.X.; Srinivasan, G. Voltage Control of Magnetism in Multiferroic Heterostructures and Devices; World Scientific Publishing Company: Singapore, 2013; Volume 2, p. 1240004. [CrossRef] 5. Ramesh, R.; Spaldin, N.A. Multiferroics: Progress and prospects in thin films. Nat. Mater. 2007, 6, 21. [CrossRef][PubMed] 6. Spaldin, N.; Ramesh, R. Advances in magnetoelectric multiferroics. Nat. Mater. 2019, 18, 203. [CrossRef] 7. Palneedi, H.; Annapureddy, V.; Priya, S.; Ryu, J. Status and perspectives of multiferroic magnetoelectric composite materials and applications. Actuators 2016, 5, 9. [CrossRef] 8. Chu, Z.; PourhosseiniAsl, M.; Dong, S. Review of multi-layered magnetoelectric composite materials and devices applications. J. Phys. D: Appl. Phys. 2018, 51, 243001. [CrossRef] 9. Eerenstein, W.; Mathur, N.; Scott, J.F. Multiferroic and magnetoelectric materials. Nature 2006, 442, 759. [CrossRef] 10. Fiebig, M. Revival of the magnetoelectric effect. J. Phys. D: Appl. Phys. 2005, 38, R123. [CrossRef] 11. Spaldin, N.A.; Fiebig, M. The renaissance of magnetoelectric multiferroics. Science 2005, 309, 391–392. [CrossRef] 12. Martin, L.; Chu, Y.-H.; Ramesh, R. Advances in the growth and characterization of magnetic, ferroelectric, and multiferroic oxide thin films. Mater. Sci. Eng.: R: Rep. 2010, 68, 89–133. [CrossRef] 13. Tu, C.; Chu, Z.-Q.; Spetzler, B.; Hayes, P.; Dong, C.-Z.; Liang, X.-F.; Chen, H.-H.; He, Y.-F.; Wei, Y.-Y.; Lisenkov, I. Mechanical-Resonance-Enhanced Thin-Film Magnetoelectric Heterostructures for Magnetometers, Mechanical Antennas, Tunable RF Inductors, and Filters. Materials 2019, 12, 2259. [CrossRef] [PubMed] 14. Wang, Y.; Li, J.; Viehland, D. Magnetoelectrics for magnetic sensor applications: Status, challenges and perspectives. Mater. Today 2014, 17, 269–275. [CrossRef] 15. Zhai, J.; Xing, Z.; Dong, S.; Li, J.; Viehland, D. Detection of pico-Tesla magnetic fields using magneto-electric sensors at room temperature. Appl. Phys. Lett. 2006, 88, 062510. [CrossRef] 16. Dong, S.; Li, J.-F.; Viehland, D. Ultrahigh magnetic field sensitivity in laminates of TERFENOL-D and Pb (Mg 1/3 Nb 2/3) O 3–PbTiO 3 crystals. Appl. Phys. Lett. 2003, 83, 2265–2267. [CrossRef] 17. Sreenivasulu, G.; Laletin, U.; Petrov, V.; Petrov, V.; Srinivasan, G. A permendur-piezoelectric multiferroic composite for low-noise ultrasensitive magnetic field sensors. Appl. Phys. Lett. 2012, 100, 173506. [CrossRef] 18. Nan, T.; Hui, Y.; Rinaldi, M.; Sun, N.X. Self-biased 215MHz magnetoelectric NEMS resonator for ultra-sensitive DC magnetic field detection. Sci. Rep. 2013, 3. [CrossRef] 19. Li, M.; Dong, C.; Zhou, H.; Wang, Z.; Wang, X.; Liang, X.; Lin, Y.; Sun, N.X. Highly sensitive DC magnetic field sensor based on nonlinear ME effect. IEEE Sens. Lett. 2017, 1, 1–4. [CrossRef] 20. Dong, C.; He, Y.; Liang, X.; Chen, H.; Wei, Y.; Sun, N.X. Highly Sensitive Magnetic Sensor based on Magnetoelectric Effect and Cross-Modulation Technique. In Proceedings of the APS March Meeting 2019, Boston, MA, USA, 4–8 March 2019. 21. Tu, C.; Dong, C.; Chu, Z.; Chen, H.; Liang, X.; Sun, N.X. A passive isolator realized by magnetoelectric laminate composites. Appl. Phys. Lett. 2018, 113, 262904. [CrossRef] Sensors 2020, 20, 1532 23 of 27

22. Dong, S.; Zhai, J.; Li, J.; Viehland, D.; Priya, S. Multimodal system for harvesting magnetic and mechanical energy. Appl. Phys. Lett. 2008, 93, 103511. [CrossRef] 23. Onuta, T.-D.; Wang, Y.; Long, C.J.; Takeuchi, I. Energy harvesting properties of all-thin-film multiferroic cantilevers. Appl. Phys. Lett. 2011, 99, 203506. [CrossRef] 24. Zhai, J.; Gao, J.; De Vreugd, C.; Li, J.; Viehland, D.; Filippov, A.; Bichurin, M.; Drozdov, D.; Semenov, G.; Dong, S. Magnetoelectric gyrator. Eur. Phys. J. B 2009, 71, 383. [CrossRef] 25. Jia, Y.; Luo, H.; Zhao, X.; Wang, F. Giant magnetoelectric response from a piezoelectric/magnetostrictive laminated composite combined with a piezoelectric transformer. Adv. Mater. 2008, 20, 4776–4779. [CrossRef] 26. Kosub, T.; Kopte, M.; Hühne, R.; Appel, P.; Shields, B.; Maletinsky, P.; Hübner, R.; Liedke, M.O.; Fassbender, J.; Schmidt, O.G. Purely antiferromagnetic magnetoelectric random access memory. Nat. Commun. 2017, 8, 13985. [CrossRef] 27. Bibes, M.; Barthélémy, A. Multiferroics: Towards a magnetoelectric memory. Nat. Mater. 2008, 7, 425. [CrossRef] 28. Scott, J. Data storage: Multiferroic memories. Nat. Mater. 2007, 6, 256. [CrossRef] 29. Lou, J.; Reed, D.; Liu, M.; Sun, N. Electrostatically tunable magnetoelectric inductors with large inductance tunability. Appl. Phys. Lett. 2009, 94, 112508. [CrossRef] 30. Liu, G.; Cui, X.; Dong, S. A tunable ring-type magnetoelectric inductor. J. Appl. Phys. 2010, 108, 094106. [CrossRef] 31. Tatarenko, A.; Gheevarughese, V.; Srinivasan, G. Magnetoelectric microwave bandpass filter. Electron. Lett. 2006, 42, 540–541. [CrossRef] 32. Lin, H.; Nan, T.; Qian, Z.; Gao, Y.; Hui, Y.; Wang, X.; Guo, R.; Belkessam, A.; Shi, W.; Rinaldi, M. Tunable RF band-pass filters based on NEMS magnetoelectric resonators. Proceedings of 2016 IEEE MTT-S International Microwave Symposium (IMS), San Francisco, CA, USA, 22–27 May 2016; pp. 1–4. [CrossRef] 33. Tatarenko, A.; Srinivasan, G.; Bichurin, M. Magnetoelectric microwave phase shifter. Appl. Phys. Lett. 2006, 88, 183507. [CrossRef] 34. Ustinov, A.; Srinivasan, G.; Kalinikos, B. Ferrite-ferroelectric hybrid wave phase shifters. Appl. Phys. Lett. 2007, 90, 031913. [CrossRef] 35. Gao, Y.; Zare, S.; Onabajo, M.; Li, M.; Zhou, Z.; Nan, T.; Yang, X.; Liu, M.; Mahalingam, K.; Howe, B.M. Power-efficient voltage tunable RF integrated magnetoelectric inductors with FeGaB/Al 2 O 3 multilayer films. Proceedings of 2014 IEEE MTT-S International Microwave Symposium (IMS2014), Tampa, FL, USA, 1–6 June 2014; pp. 1–4. 36. Gao, Y.; Zardareh, S.Z.; Yang, X.; Nan, T.X.; Zhou, Z.Y.; Onabajo, M.; Liu, M.; Aronow, A.; Mahalingam, K.; Howe, B.M. Significantly Enhanced Inductance and Quality Factor of GHz Integrated Magnetic Solenoid

Inductors With FeGaB/Al2O3 Multilayer Films. IEEE Trans. Electron Devices 2014, 61, 1470–1476. [CrossRef] 37. Bickford, J.A.; McNabb, R.S.; Ward, P.A.; Freeman, D.K.; Weinberg, M.S. Low frequency mechanical antennas: Electrically short transmitters from mechanically-actuated dielectrics. Proceedings of 2017 IEEE International Symposium on Antennas and Propagation & USNC/URSI National Radio Science Meeting, San Diego, CA, USA, 9–14 July 2017; pp. 1475–1476. [CrossRef] 38. Bickford, J.A.; Duwel, A.E.; Weinberg, M.S.; McNabb, R.C.; Freeman, D.K.; Ward, P.A. Performance of Electrically Small Conventional and Mechanical Antennas. IEEE Trans. Antennas Propag. 2019.[CrossRef] 39. Fawole, O.C.; Tabib-Azar, M. An Electromechanically Modulated Permanent Magnet Antenna for Wireless Communication in Harsh Electromagnetic Environments. IEEE Trans. Antennas Propag. 2017, 65, 6927–6936. [CrossRef] 40. Kemp, M.A.; Franzi, M.; Haase, A.; Jongewaard, E.; Whittaker, M.T.; Kirkpatrick, M.; Sparr, R. A high Q piezoelectric resonator as a portable VLF transmitter. Nat. Commun. 2019, 10, 1715. [CrossRef] 41. Dong, C.; He, Y.; Li, M.; Tu, C.; Chu, Z.; Liang, X.; Chen, H.; Wei, Y.; Zaeimbashi, M.; Wang, X.; et al. A Portable Very Low Frequency (VLF) Communication System Based on Acoustically Actuated Magnetoelectric Antennas. IEEE Antennas Wirel. Propag. Lett. 2020.[CrossRef] 42. Chen, H.; Wang, X.; Gao, Y.; Shi, X.; Wang, Z.; Sun, N.; Zaeimbashi, M.; Liang, X.; He, Y.; Dong, C.; et al. Integrated Tunable Magnetoelectric RF Inductors. IEEE Trans. Microw. Theory Tech. 2020.[CrossRef] 43. Weldon, J.; Jensen, K.; Zettl, A. Nanomechanical radio transmitter. Physica Status Solidi b 2008, 245, 2323–2325. [CrossRef] Sensors 2020, 20, 1532 24 of 27

44. Nan, T.; Lin, H.; Gao, Y.; Matyushov, A.; Yu, G.; Chen, H.; Sun, N.; Wei, S.; Wang, Z.; Li, M.; et al. Acoustically actuated ultra-compact NEMS magnetoelectric antennas. Nat. Commun. 2017, 8, 296. [CrossRef] 45. Liang, X.; Chen, H.; Sun, N.; Lin, H.; Sun, N.X. Novel Acoustically Actuated Magnetoelectric Antennas. Proceedings of 2018 IEEE International Symposium on Antennas and Propagation & USNC/URSI National Radio Science Meeting, Boston, MA, USA, 8–13 July 2018; pp. 2189–2190. [CrossRef] 46. Lin, H.; Zaeimbashi, M.; Sun, N.; Liang, X.; Chen, H.; Dong, C.; Matyushov, A.; Wang, X.; Guo, Y.; Gao, Y. Future Antenna Miniaturization Mechanism: Magnetoelectric Antennas. Proceedings of 2018 IEEE/MTT-S International Microwave Symposium-IMS, Philadelphia, PA, USA, 10–15 June 2018; pp. 220–223. [CrossRef] 47. Yang, G.-M.; Xing, X.; Daigle, A.; Liu, M.; Obi, O.; Stoute, S.; Naishadham, K.; Sun, N.X. Tunable miniaturized patch antennas with self-biased multilayer magnetic ﬁlms. IEEE Trans. Antennas Propag. 2009, 57, 2190–2193. [CrossRef] 48. Yang, G.-M.; Xing, X.; Daigle, A.; Obi, O.; Liu, M.; Lou, J.; Stoute, S.; Naishadham, K.; Sun, N.X. Planar annular ring antennas with multilayer self-biased NiCo-ferrite ﬁlms loading. IEEE Trans. Antennas Propag. 2010, 58, 648–655. [CrossRef] 49. Curie, P. Sur la symétrie dans les phénomènes physiques, symétrie d’un champ électrique et d’un champ magnétique. J. Phys. Theor. Appl. 1894, 3, 393–415. [CrossRef] 50. Landau, L.D.; Bell, J.; Kearsley, M.; Pitaevskii, L.; Lifshitz, E.; Sykes, J. Electrodynamics of Continuous Media; Elsevier: Amsterdam, The Netherlands, 2013; Volume 8. 51. Catalan, G.; Scott, J.F. Physics and applications of bismuth ferrite. Adv. Mater. 2009, 21, 2463–2485. [CrossRef] 52. Van Suchtelen, J. Product properties: A new application of composite materials. Philips Res. Rep. 1972, 27, 28–37. 53. Hooker, M.W. Properties of PZT-based piezoelectric ceramics between-150 and 250 C; Technical Report, NASA/CR-1998-208708; NASA: Washington, DC, USA, 1998. 54. Berlincourt, D.; Krueger, H.; Near, C. Properties of Morgan Electro Ceramic Ceramics. Available online: http://www.ultrasonic-resonators.org/misc/references/articles/Berlincourt__\T1\textquoterightProperties_ of_Morgan_Electro_Ceramic_Ceramics\T1\textquoteright_(Morgan_Technical_Publication_TP-226).pdf (accessed on 10 March 2020). 55. Li, F.; Zhang, S.; Xu, Z.; Wei, X.; Luo, J.; Shrout, T.R. Composition and phase dependence of the intrinsic and extrinsic piezoelectric activity of domain engineered (1 x) Pb (Mg 1/3 Nb 2/3) O 3 x PbTiO 3 crystals. − − J. Appl. Phys. 2010, 108, 034106. [CrossRef] 56. Wu, J.; Gao, X.; Chen, J.; Wang, C.M.; Zhang, S.; Dong, S. Review of high temperature piezoelectric materials, devices, and applications. Acta Phys. Sinica 2018, 67, 207701.

57. Tsubouchi, K.; Sugai, K.; Mikoshiba, N. AlN material constants evaluation and SAW properties on AlN/Al2O3 and AlN/Si. In Proceedings of the 1981 Ultrasonics Symposium, Chicago, IL, USA, USA, 14–16 October 1981; pp. 375–380. [CrossRef] 58. Dubois, M.-A.; Muralt, P. Properties of aluminum nitride thin films for piezoelectric transducers and microwave filter applications. Appl. Phys. Lett. 1999, 74, 3032–3034. [CrossRef] 59. Sanchez-Rojas, J.-L.; Hernando, J.; Ababneh, A.; Schmid, U.; Olivares, J.; Clement, M.; Iborra, E. Advanced determination of piezoelectric properties of AlN thin films on silicon substrates. In Proceedings of the 2008 IEEE Ultrasonics Symposium, Beijing, China, 2–5 November 2008; pp. 903–906. [CrossRef] 60. Chu, Z.; Shi, H.; Shi, W.; Liu, G.; Wu, J.; Yang, J.; Dong, S. Enhanced resonance magnetoelectric coupling in (1-1) connectivity composites. Adv. Mater. 2017, 29, 1606022. [CrossRef] 61. Liang, X.; Dong, C.; Celestin, S.J.; Wang, X.; Chen, H.; Ziemer, K.S.; Page, M.; Mcconney, M.; Jones, J.G.; Howe, B.M. Soft Magnetism, Magnetostriction and Microwave Properties of Fe-Ga-C Alloy Films. IEEE Magn. Lett. 2018.[CrossRef] 62. Lou, J.; Insignares, R.; Cai, Z.; Ziemer, K.S.; Liu, M.; Sun, N.X. Soft magnetism, magnetostriction, and microwave properties of FeGaB thin films. Appl. Phys. Lett. 2007, 91, 182504. [CrossRef] 63. Greve, H.; Woltermann, E.; Quenzer, H.-J.; Wagner, B.; Quandt, E. Giant magnetoelectric coefficients in (Fe 90 Co 10) 78 Si 12 B 10-AlN thin film composites. Appl. Phys. Lett. 2010, 96, 182501. [CrossRef] 64. Joule, J.P., XVII. On the effects of magnetism upon the dimensions of iron and steel bars. Lond. Edinb. Dublin Philos. Mag. J. Sci. 1847, 30, 76–87. [CrossRef] 65. Ibarra, M.; Algarabel, P. Giant volume magnetostriction in the FeRh alloy. Phys. Rev. B 1994, 50, 4196. [CrossRef][PubMed] Sensors 2020, 20, 1532 25 of 27

66. Dong, C.; Li, M.; Liang, X.; Chen, H.; Zhou, H.; Wang, X.; Gao, Y.; McConney, M.E.; Jones, J.G.; Brown, G.J. Characterization of magnetomechanical properties in FeGaB thin films. Appl. Phys. Lett. 2018, 113, 262401. [CrossRef] 67. Zhao, P.; Zhao, Z.; Hunter, D.; Suchoski, R.; Gao, C.; Mathews, S.; Wuttig, M.; Takeuchi, I. Fabrication and characterization of all-thin-film magnetoelectric sensors. Appl. Phys. Lett. 2009, 94, 243507. [CrossRef] 68. Kirchhof, C.; Krantz, M.; Teliban, I.; Jahns, R.; Marauska, S.; Wagner, B.; Knöchel, R.; Gerken, M.; Meyners, D.; Quandt, E. Giant magnetoelectric effect in vacuum. Appl. Phys. Lett. 2013, 102, 232905. [CrossRef] 69. Piorra, A.; Jahns, R.; Teliban, I.; Gugat, J.; Gerken, M.; Knöchel, R.; Quandt, E. Magnetoelectric thin film composites with interdigital electrodes. Appl. Phys. Lett. 2013, 103, 032902. [CrossRef] 70. Salzer, S.; Jahns, R.; Piorra, A.; Teliban, I.; Reermann, J.; Höft, M.; Quandt, E.; Knöchel, R. Tuning fork for noise suppression in magnetoelectric sensors. Sens. Actuators A Phys. 2016, 237, 91–95. [CrossRef] 71. Yarar, E.; Salzer, S.; Hrkac, V.; Piorra, A.; Höft, M.; Knöchel, R.; Kienle, L.; Quandt, E. Inverse bilayer magnetoelectric thin film sensor. Appl. Phys. Lett. 2016, 109, 022901. [CrossRef] 72. Lin, H.; Zaeimbashi, M.; Sun, N.; Liang, X.; Chen, H.; Dong, C.; Matyushov, A.; Wang, X.; Guo, Y.; Gao, Y. NEMS Magnetoelectric Antennas for Biomedical Application. In Proceedings of the 2018 IEEE International Microwave Biomedical Conference (IMBioC), Philadelphia, PA, USA, 14–15 June 2018; pp. 13–15. [CrossRef] 73. Zaeimbashi, M.; Lin, H.; Dong, C.; Liang, X.; Nasrollahpour, M.; Chen, H.; Sun, N.; Matyushov, A.; He, Y.; Wang, X. NanoNeuroRFID: A Wireless Implantable Device Based on Magnetoelectric Antennas. IEEE J. Electromagn. RF Microw. Med. Biol. 2019.[CrossRef] 74. Sproul, W.D. Physical vapor deposition tool coatings. Surf. Coat. Technol. 1996, 81, 1–7. [CrossRef] 75. Mattox, D.M. Handbook of Physical Vapor Deposition (PVD) Processing; Elsevier: Oxford, UK, 2010. [CrossRef] 76. Wang, D.N.; White, J.M.; Law, K.S.; Leung, C.; Umotoy, S.P.; Collins, K.S.; Adamik, J.A.; Perlov, I.; Maydan, D. Thermal CVD/PECVD reactor and use for thermal chemical vapor deposition of silicon dioxide and in-situ multi-step planarized process. U.S. Patent No. 5,000,113, 19 March 1991. 77. Obraztsov, A.; Obraztsova, E.; Tyurnina, A.; Zolotukhin, A. Chemical vapor deposition of thin graphite films of nanometer thickness. Carbon 2007, 45, 2017–2021. [CrossRef] 78. Leskelä, M.; Ritala, M. Atomic layer deposition (ALD): From precursors to thin film structures. Thin Solid Film. 2002, 409, 138–146. [CrossRef] 79. Sneh, O.; Clark-Phelps, R.B.; Londergan, A.R.; Winkler, J.; Seidel, T.E. Thin film atomic layer deposition equipment for semiconductor processing. Thin Solid Film. 2002, 402, 248–261. [CrossRef] 80. Brinker, C.; Hurd, A.; Schunk, P.; Frye, G.; Ashley, C. Review of sol-gel thin film formation. J. Non Cryst. · Solids 1992, 147, 424–436. [CrossRef] 81. Kalarickal, S.S.; Krivosik, P.; Wu, M.; Patton, C.E.; Schneider, M.L.; Kabos, P.; Silva, T.J.; Nibarger, J.P. Ferromagnetic resonance linewidth in metallic thin films: Comparison of measurement methods. J. Appl. Phys. 2006, 99, 093909. [CrossRef] 82. Berry, B.; Pritchet, W. Vibrating reed internal friction apparatus for films and foils. IBM J. Res. Dev. 1975, 19, 334–343. [CrossRef] 83. Klokholm, E. The measurement of magnetostriction in ferromagnetic thin films. IEEE Trans. Magn. 1976, 12, 819–821. [CrossRef] 84. Tam, A.C.; Schroeder, H. A new high-precision optical technique to measure magnetostriction of a thin magnetic film deposited on a substrate. IEEE Trans. Magn. 1989, 25, 2629–2638. [CrossRef] 85. Shima, T.; Fujimori, H. An accurate measurement of magnetostriction of thin films by using nano-indentation system. IEEE Trans. Magn. 1999, 35, 3832–3834. [CrossRef] 86. Ekreem, N.; Olabi, A.; Prescott, T.; Rafferty, A.; Hashmi, M. An overview of magnetostriction, its use and methods to measure these properties. J. Mater. Process. Technol. 2007, 191, 96–101. [CrossRef] 87. Jahns, R.; Zabel, S.; Marauska, S.; Gojdka, B.; Wagner, B.; Knöchel, R.; Adelung, R.; Faupel, F. Microelectromechanical magnetic field sensor based on ∆E effect. Appl. Phys. Lett. 2014, 105, 052414. [CrossRef] 88. Rugar, D.; Mamin, H.; Guethner, P.; Lambert, S.; Stern, J.; McFadyen, I.; Yogi, T. Magnetic force microscopy: General principles and application to longitudinal recording media. J. Appl. Phys. 1990, 68, 1169–1183. [CrossRef] 89. Scheinfein, M.; Unguris, J.; Kelley, M.H.; Pierce, D.T.; Celotta, R. Scanning electron microscopy with polarization analysis (SEMPA). Rev. Sci. Instrum. 1990, 61, 2501–2527. [CrossRef] Sensors 2020, 20, 1532 26 of 27

90. Uhlíˇr,V.; Pizzini, S.; Rougemaille, N.; Novotný, J.; Cros, V.; Jiménez, E.; Faini, G.; Heyne, L.; Sirotti, F.; Tieg, C. Current-induced motion and pinning of domain walls in spin-valve nanowires studied by XMCD-PEEM. Phys. Rev. B 2010, 81, 224418. [CrossRef] 91. Hiebert, W.; Stankiewicz, A.; Freeman, M. Direct observation of magnetic relaxation in a small permalloy disk by time-resolved scanning Kerr microscopy. Phys. Rev. Lett. 1997, 79, 1134. [CrossRef] 92. Mozooni, B.; von Hofe, T.; McCord, J. Picosecond wide-field magneto-optical imaging of magnetization dynamics of amorphous film elements. Phys. Rev. B 2014, 90, 054410. [CrossRef] 93. McCord, J. Progress in magnetic domain observation by advanced magneto-optical microscopy. J. Phys. D: Appl. Phys. 2015, 48, 333001. [CrossRef] 94. Urs, N.O.; Mozooni, B.; Mazalski, P.; Kustov, M.; Hayes, P.; Deldar, S.; Quandt, E.; McCord, J. Advanced magneto-optical microscopy: Imaging from picoseconds to centimeters-imaging spin waves and temperature distributions. AIP Adv. 2016, 6, 055605. [CrossRef] 95. Villari, E. Ueber die Aenderungen des magnetischen Moments, welche der Zug und das Hindurchleiten eines galvanischen Stroms in einem Stabe von Stahl oder Eisen hervorbringen. Ann. Der Phys. 1865, 202, 87–122. [CrossRef] 96. Clark, A.E.; DeSavage, B.F.; Bozorth, R. Anomalous Thermal Expansion and Magnetostriction of Single-Crystal Dysprosium. Phys. Rev. 1965, 138, A216–A224. [CrossRef] 97. Koon, N.C.; Williams, C.M.; Das, B.N. Giant magnetostriction materials. J. Magn. Magn. Mater. 1991, 100, 173–185. [CrossRef] 98. Olabi, A.G.; Grunwald, A. Design and application of magnetostrictive materials. Mater. Des. 2008, 29, 469–483. [CrossRef] 99. Snodgrass, J.D.; McMasters, O. Optimized TERFENOL-D manufacturing processes. J. Alloy. Compd. 1997, 258, 24–29. [CrossRef] 100. Williams, P.I.; Lord, D.G.; Grundy, P.J. Magnetostriction in polycrystalline sputter-deposited TbDyFe films. J. Appl. Phys. 1994, 75, 5257–5261. [CrossRef] 101. Chen, C.H.; Walmer, M.S.; Walmer, M.H.; Liu, S.; Kuhl, E.; Simon, G. Sm2(Co,Fe,Cu,Zr)17 magnets for use at temperature 400 C. J. Appl. Phys. 1998, 83, 6706–6708. [CrossRef] ≥ ◦ 102. Lim, S.H.; Choi, Y.S.; Han, S.H.; Kim, H.J.; Shima, T.; Fujimori, H. Magnetostriction of Sm–Fe and Sm–Fe–B thin films fabricated by RF magnetron sputtering. J. Magn. Magn. Mater. 1998, 189, 1–7. [CrossRef] 103. Clark, A.E.; Hathaway, K.B.; Wun-Fogle, M.; Restorff, J.; Lograsso, T.A.; Keppens, V.; Petculescu, G.; Taylor, R. Extraordinary magnetoelasticity and lattice softening in bcc Fe-Ga alloys. J. Appl. Phys. 2003, 93, 8621–8623. [CrossRef] 104. Cullen, J.; Clark, A.; Wun-Fogle, M.; Restorff, J.; Lograsso, T.A. Magnetoelasticity of fe–ga and fe–al alloys. J. Magn. Magn. Mater. 2001, 226, 948–949. [CrossRef] 105. Masiyama, Y. Magnetostriction in CobaltIron alloys. Sci. Rep. Res. Tohoku A 1932, 21, 394. 106. Clark, A.E.; Wun-Fogle, M.; Restorff, J.B.; Lograsso, T.A. Magnetostrictive properties of Galfenol alloys under compressive stress. Mater. Trans. 2002, 43, 881–886. [CrossRef] 107. Srisukhumbowornchai, N.; Guruswamy, S. Large magnetostriction in directionally solidified FeGa and FeGaAl alloys. J. Appl. Phys. 2001, 90, 5680–5688. [CrossRef] 108. Clark, A.E.; Restorff, J.B.; Wun-Fogle, M.; Summers, E. Effect of stress annealing on Galfenol and Alfenol magnetostrictive alloys; SPIE: Bellingham, WA, USA; Volume 6170. [CrossRef] 109. Butera, A.; Gómez, J.; Weston, J.; Barnard, J. Growth and magnetic characterization of epitaxial Fe 81 Ga 19/ MgO (100) thin films. J. Appl. Phys. 2005, 98, 033901. [CrossRef] 110. Clark, A.; Restorff, J.; Wun-Fogle, M.; Hathaway, K.; Lograsso, T.A.; Huang, M.; Summers, E. Magnetostriction of ternary Fe–Ga–X (X= C, V, Cr, Mn, Co, Rh) alloys. J. Appl. Phys. 2007, 101, 09C507. [CrossRef] 111. Kakehashi, Y. Systematic Variations of Magnetic Properties in 3d Transition Metal Alloys. Prog. Theor. Phys. Suppl. 1990, 101, 105–117. [CrossRef] 112. Wang, J.; Dong, C.; Wei, Y.; Lin, X.; Athey, B.; Chen, Y.; Winter, A.; Stephen, G.M.; Heiman, D.; He, Y.; et al. Magnetostriction, Soft Magnetism, and Microwave Properties in Co-Fe-C Alloy Films. Phys. Rev. Appl. 2019, 12, 034011. [CrossRef] 113. Betz, J.; Mackay, K.; Peuzin, J.-C.; Givord, D.; Halstrup, B. Giant Magnetostrictive, Spring Magnet Type Multilayers and Torsion Based Microactuators. Mrs Online Proc. Libr. Arch. 1996, 459. [CrossRef] Sensors 2020, 20, 1532 27 of 27

114. Shi, J.; Wu, M.; Hu, W.; Lu, C.; Mu, X.; Zhu, J. A study of high piezomagnetic (Fe-Ga/Fe-Ni) multilayers for magnetoelectric device. J. Alloy. Compd. 2019, 806, 1465–1468. [CrossRef] 115. Hubert, A.; Schäfer, R. Magnetic Domains: The Analysis of Magnetic Microstructures; Springer Science & Business Media: Berlin/Heidelberg, Germany, 2008. 116. Quandt, E.; Ludwig, A. Giant magnetostrictive multilayers. J. Appl. Phys. 1999, 85, 6232–6237. [CrossRef] 117. Ludwig, A.; Quandt, E. Giant magnetostrictive thin films for applications in microelectromechanical systems. J. Appl. Phys. 2000, 87, 4691–4695. [CrossRef] 118. Ludwig, A.; Quandt, E. Optimization of the /spl Delta/E-effect in thin films and multilayers by magnetic field annealing. IEEE Trans. Magn. 2002, 38, 2829–2831. [CrossRef] 119. Hunter, D.; Osborn, W.; Wang, K.; Kazantseva, N.; Hattrick-Simpers, J.; Suchoski, R.; Takahashi, R.; Young, M.L.; Mehta, A.; Bendersky, L.A.; et al. Giant magnetostriction in annealed Co1 xFex thin-films. Nat. − Commun. 2011, 2, 518. [CrossRef][PubMed] 120. García-Arribas, A.; Gutiérrez, J.; Kurlyandskaya, G.V.; Barandiarán, J.M.; Svalov, A.; Fernández, E.; Lasheras, A.; De Cos, D.; Bravo-Imaz, I. Sensor applications of soft magnetic materials based on magneto-impedance, magneto-elastic resonance and magneto-electricity. Sensors 2014, 14, 7602–7624. [CrossRef][PubMed] 121. Wang, Z.; Wang, X.; Li, M.; Gao, Y.; Hu, Z.; Nan, T.; Liang, X.; Chen, H.; Yang, J.; Cash, S. Highly sensitive flexible magnetic sensor based on anisotropic magnetoresistance effect. Adv. Mater. 2016, 28, 9370–9377. [CrossRef][PubMed] 122. García-Arribas, A.; Combarro, L.; Goiriena-Goikoetxea, M.; Kurlyandskaya, G.V.; Svalov, A.V.; Fernández, E.; Orue, I.; Feuchtwanger, J. Thin-film magnetoimpedance structures onto flexible substrates as deformation sensors. IEEE Trans. Magn. 2016, 53, 1–5. [CrossRef]

© 2020 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/).