# Fileset

[Advanced Physics Research - 2024 - Modak - High‐Throughput Optimization of Magnetoresistance Materials Based on Lock‐In.pdf](https://mdr.nims.go.jp/filesets/69f07311-4f55-4695-80c9-f4196e389c38/download)

## Creator

[Rajkumar Modak](https://orcid.org/0000-0001-7939-3289), [Takamasa Hirai](https://orcid.org/0000-0002-5577-8018), [Yuya Sakuraba](https://orcid.org/0000-0003-4618-9550), [Seiji Mitani](https://orcid.org/0000-0002-1348-0774), [Koichi Oyanagi](https://orcid.org/0000-0001-8784-078X), [Takumi Yamazaki](https://orcid.org/0000-0001-5289-8615), [Takeshi Seki](https://orcid.org/0000-0003-3195-7051), [Ken‐ichi Uchida](https://orcid.org/0000-0001-7680-3051)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[High‐Throughput Optimization of Magnetoresistance Materials Based on Lock‐In Thermography](https://mdr.nims.go.jp/datasets/6b23b47b-8ab8-444f-93a9-fa255eb5a203)

## Fulltext

High‐Throughput Optimization of Magnetoresistance Materials Based on Lock‐In ThermographyRESEARCH ARTICLEwww.advphysicsres.comHigh-Throughput Optimization of MagnetoresistanceMaterials Based on Lock-In ThermographyRajkumar Modak,* Takamasa Hirai, Yuya Sakuraba, Seiji Mitani, Koichi Oyanagi,Takumi Yamazaki, Takeshi Seki, and Ken-ichi Uchida*With the giant magnetoresistance (GMR) effect serving as a vital componentin modern spintronic technologies, researchers are dedicating significantefforts to improve the performance of GMR devices through materialexploration and design optimization. However, traditional GMR measurementapproaches are inefficient for comprehensive material and deviceoptimization. This study proposes a high-throughput current-in-plane GMRmeasurement technique based on thermal imaging of Joule heating utilizinglock-in thermography (LIT). This LIT-based technique is advantageous forefficiently evaluating films with varying compositions and thickness gradients,which is crucial for ongoing material exploration and design optimization toenhance the GMR ratio. First, it is demonstrated that using CoFe/Cumultilayers, the simple Joule heating measurement based on LIT enablesquantitative estimation of the GMR ratio. Then, to confirm the usefulness ofthe proposed method in high-throughput material screening, a case study isshown to investigate the GMR of CoCu-based granular films with acomposition gradient. These techniques allow to determine the optimumcomposition with maximum GMR ratio using the single composition-gradientfilm and reveal Co22Cu78 as the optimal composition, yielding the largestGMR ratio among the reported polycrystalline CoCu-based granular films.This demonstration accelerates the material and structural optimization ofGMR devices.1. IntroductionThe giant magnetoresistance (GMR) effect, discovered in Fe/Crsuperlattices in 1988,[1,2] is a phenomenon where the resistanceof a metallic multilayer, consisting of ferromagnetic (FM) andR. Modak, T. Hirai, Y. Sakuraba, S. Mitani, K. Oyanagi, T. Seki, K. UchidaNational Institute for Materials ScienceTsukuba 305-0047, JapanE-mail: modak.rajkumar@nims.go.jp; uchida.kenichi@nims.go.jpThe ORCID identification number(s) for the author(s) of this articlecan be found under https://doi.org/10.1002/apxr.202400021© 2024 The Author(s). Advanced Physics Research published byWiley-VCH GmbH. This is an open access article under the terms of theCreative Commons Attribution License, which permits use, distributionand reproduction in any medium, provided the original work is properlycited.DOI: 10.1002/apxr.202400021nonmagnetic (NM) layers, changes de-pending on the magnetization direc-tion of each FM layer. Later, GMR wasfound also in several bi-metallic or multi-metallic layers containing FM and anti-ferromagnetic or NM metals, as well assingle-layer granular materials.[3–7] Sinceits discovery, GMR has been the sub-ject of extensive research to understandthe underlying physics and exploit itsvast technical applications. Over pastdecades, the applications of GMR con-cepts have extended across diverse do-mains, from revolutionizing read/writeheads in hard disk drives[8] to automotivesensors,[9] biomedical sensors,[10] flexiblestrain sensors,[11] non-destructive materi-als testing,[12] and beyond.In recent years, current-in-plane(CIP) GMR devices have attracted re-newed interest due to their structuralsimplicity and ease of fabrication,making them promising candidatesfor high-sensitivity magnetic fieldsensors,[13,14] biosensors,[15,16] emergingspin-orbit torque devices,[17] and flexibleelectronics.[18,19] However, the CIP-GMRratios in conventional materials are oftenlower compared to tunneling magnetoresistance[20–24] andcurrent-perpendicular-to-plane GMR,[25–28] limiting their sensi-tivity in these emerging applications. Therefore, research isongoing to optimize material and device design for improvedCIP-GMR performance. Researchers actively explore materialK. OyanagiFaculty of Science and EngineeringIwate UniversityMorioka 020-8551, JapanT. Yamazaki, T. Seki, K. UchidaInstitute for Materials ResearchTohoku UniversitySendai 980-8577, JapanT. SekiCenter for Science and Innovation in SpintronicsTohoku UniversitySendai 980-8577, JapanAdv. Physics Res. 2024, 3, 2400021 2400021 (1 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbHhttp://www.advphysicsres.commailto:modak.rajkumar@nims.go.jpmailto:uchida.kenichi@nims.go.jphttps://doi.org/10.1002/apxr.202400021http://creativecommons.org/licenses/by/4.0/http://crossmark.crossref.org/dialog/?doi=10.1002%2Fapxr.202400021&domain=pdf&date_stamp=2024-05-24www.advancedsciencenews.com www.advphysicsres.comFigure 1. Joule heating in materials with GMR. Schematic of the change in Joule heating due to current-in-plane GMR in a ferromagnetic metal(FM)/nonmagnetic metal (NM) multilayer. With a parallel (antiparallel) magnetization configuration of the FM layers, the multilayer has a low (high)electrical resistance, which results in small (large) Joule heating when a charge current Jc is passed. The same situation occurs in a magnetic granularmaterial when it exhibits GMR.compositions, layer structures, and fabrication techniques, seek-ing higher GMR ratios and improved stability. Various ap-proaches have been adopted for this purpose. The systematic tun-ing of FM layer composition and thickness,[29–31] NM spacer layerthickness,[3,32] and seed layer composition and thickness[33–35] area few popular methods for improving the GMR response.Despite these efforts, the complexity and time-consumingnature of systematic GMR measurements for uniform filmswith various compositions/thicknesses remain significant bar-riers to high-throughput material screening, which is crucialfor effectively identifying promising materials with desirableGMR properties. In recent years, researchers have used wedge-shaped films to determine the optimal thickness for spacer layerseffectively.[36,37] However, this method requires the formation ofmicrostructured devices and subsequent GMR measurements ineach device. The limited number of devices achievable along thecomposition gradient during microfabrication restricts the min-imum difference in composition that can be investigated. This,coupled with the time-consuming process of measuring GMRfor each device individually, significantly limits the utility of thisapproach for high-throughput material screening. Also, the dis-crete data points from individual microfabricated devices may re-sult in overlooking the optimum condition for the best GMR ra-tio. Therefore, it is necessary to develop an alternative measure-ment approach that overcomes these limitations and expeditesthe screening and optimization of GMR materials.This study addresses this need by introducing and validatinga simple and effective CIP-GMR measurement method based onthermal imaging of Joule heating, utilizing the lock-in thermog-raphy (LIT) technique.[38–41] The resistance R of a material ex-hibiting the GMR effect experiences significant variations con-current with the changes in its magnetization configuration dueto the influence of the external magnetic field H. The GMR ra-tio is determined as ∆R/R, where ∆R is the change in resistancebetween parallel and antiparallel magnetization states (betweenhigh-H and zero-H conditions) in multilayer (granular) film.When a charge current Jc is applied to the material to measurethe GMR effect, Joule heating Q also appears, where Q ∝ J2c R.Given the direct proportionality between Joule heating and R, therelationship to quantify GMR can also be expressed as ∆Q/Q,where ∆Q is the change in the heat release due to Joule heatingbetween parallel and antiparallel magnetization state (betweenhigh-H and zero-H conditions) in a multilayer (granular) film(Figure 1), suggesting a possibility of quantifying the GMR ra-tio through the Joule-heating measurement.Building upon this foundational understanding, we employedthe LIT technique to measure the temperature modulation in-duced by Joule heating in a conductor for GMR estimation.The LIT method allows for highly sensitive Joule-heating mea-surements and also suppression of parasitic effects due to ther-mal diffusion by measuring the temperature modulation at ahigh lock-in frequency.[39,40] This study shows the precise es-timation of the CIP-GMR ratio from thermal images of Jouleheating for a CoFe/Cu multilayer film, demonstrating the ac-curacy of this method. Then, a comprehensive case study onCoCu-based granular films is presented. This study showcasesthe ability of the proposed techniques to simultaneously estimatecomposition-dependent GMR, thereby highlighting its utility forhigh-throughput GMR material investigation. The experimentsalso unveil the optimal composition of the granular film to beCo22Cu78, achieving a large room-temperature GMR ratio exceed-ing 8% at μ0H= 0.9 T, where μ0 denotes the vacuum permeability,and a saturation GMR ratio of ≈28% at 4 K, record-high values forpolycrystalline CoCu-based single-layer granular materials.[42–45]These results emphasize the accuracy and versatility of the LIT-based method and demonstrate its potential to drive advance-ments in material exploration and device optimization in spin-tronics.2. Results and Discussion2.1. Validation of LIT-Based GMR Measurement MethodTo validate the reliability of the proposed LIT-based GMR mea-surement method, we conducted experiments using a uniformreference material: the epitaxial CoFe/Cu multilayer film. Thisfilm is known to exhibit a very large CIP-GMR ratio at roomtemperature.[36,46] The film was structured into a strip geometrywith a width of 0.2 mm and a length of 6.0 mm for the mea-surement. The temperature modulation due to Joule heating wasrecorded using the LIT technique while an ON/OFF-modulatedAC charge current with an ON-state current magnitude of Jc andfrequency f was applied to the sample along the length direc-tion. The experimental configuration for the LIT measurementsis shown in Figure 2a. Using an infrared camera, we recordedthermal images of the surface of the CoFe/Cu multilayer film andAdv. Physics Res. 2024, 3, 2400021 2400021 (2 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.comFigure 2. Thermal imaging of Joule heating using LIT. a) Schematics of the CoFe/Cu multilayer film structure used in the present study and the exper-imental setup for measuring the change in Joule-heating-induced temperature modulation due to GMR when Jc is applied to the film. During the LITmeasurements, an ON/OFF-modulated AC charge current with an ON-state current magnitude Jc and frequency f was applied to the sample. Underthese conditions, the output Joule heating oscillates with the same f as the applied current, as illustrated in the input and output time charts. b) Lock-inamplitude (A) and phase (ϕ) signals produced by Joule heating in the CoFe/Cu multilayer film, acquired at Jc = 60 mA and f = 25.0 Hz, and applied mag-netic fields of μ0H = +0.3 and 0.0 T. μ0 denotes the vacuum permeability. AAEE and ϕAEE represent the signals produced by the anomalous Ettingshauseneffect (AEE) for the film at |μ0H| = 0.3 T. During the AEE measurement, Jc = 60 mA, f = 25 Hz, and zero DC offset was applied to the sample.[48–52]c) H dependence of the A signal obtained from the Joule heating measurements on the CoFe/Cu multilayer film, recorded at discrete magnetic fieldvalues during the incremental and decremental sweeps of H. The A values were obtained by averaging the raw signals in an area of 0.1 × 0.9 mm2on the sample, as represented by the dashed rectangular box in b). The error bars represent the standard deviations of the data in the correspondingrectangular box. The solid curves represent the H dependence of the electrical resistance R of the film measured by the four-probe method by applyinga 60 mA direct current.extracted the charge-current-induced temperature change oscil-lating at the same f as the Jc through Fourier analysis. The ob-tained thermal images were transformed into lock-in amplitudeA and phase ϕ images, where A represents the magnitude of thecharge-current-induced temperature change and ϕ indicate thesign and time delay of the temperature change. We coated thesample surface with a thin layer of insulating black ink. Thisink has a very high infrared emissivity of > 0.94, guaranteeinga strong and consistent infrared emission signal regardless ofthe inherent emissivity of the material and improving the signal-to-noise ratio in the measured data. Consequently, the black inkcoating in this measurement broadens the scope of materials thatcan be studied. All measurements were performed at room tem-perature and atmospheric pressure.To evaluate the GMR-induced change in Joule heating, werecorded the μ0H dependence of A and ϕ images of the CoFe/Cumultilayer film at Jc = 60 mA and f = 25.0 Hz, with H appliedalong the width direction. The measurements were performedat the maximum available lock-in frequency of the system todisregard the effects of heat convection on the environment.[47]Figure 2b displays the representative A and ϕ images recordedat μ0H = +0.3 and 0.0 T. It is important to note that the LIT im-ages obtained at a finite H may encompass signals arising fromthe anomalous Ettingshausen effect (AEE).[48–52] However, for thepresent CoFe/Cu multilayer film, the AEE signals are smallerthan the noise level of the Joule-heating contribution (Figure 2b).This implies that the observed LIT images predominantly reflectJoule heating effects, allowing us to disregard AEE contributionin A for our GMR estimation. The μ0H dependence of A, obtainedby averaging the region indicated by the dashed rectangular boxin Figure 2b covering an area of 0.1 × 0.9 mm2, is presentedin Figure 2c as discrete data points. In this figure, we comparethe μ0H dependence of A with that of R of the sample, repre-sented by solid lines. The μ0H dependence of R was measuredby continuously sweeping μ0H from +0.3 to −0.3 T in forwardand reverse directions and recording R using the conventionalfour-probe method. The results indicate that the magnitude of Afollows the same H dependence as R, suggesting that the rela-tive variations in R can be determined from the LIT images, pro-viding a means of calculating the GMR ratio based on the rela-tion ∆Q/Q = ∆R/R. For the LIT images of the present CoFe/Cumultilayer film, ∆Q/Q can be rewritten as ∆A/A = (AH−A0)/A0.Here, AH (A0) represents the Joule-heating-induced A signal atthe parallel (antiparallel) magnetization state at high (zero) H.[36]Adv. Physics Res. 2024, 3, 2400021 2400021 (3 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.comFigure 3. Estimation of GMR ratio from thermal imaging of Joule heat-ing. a) Jc dependence of the magnetic-field-induced change ratio of theJoule heating amplitude ∆A/A = (A+0.3 T−A0.0 T)/A0.0 T (black squares) ac-quired at f = 25.0 Hz. Here, A+0.3 T and A0.0 T represent the A values atμ0H = +0.3 and 0.0 T, respectively. Gray circles represent the GMR ra-tio ∆R/R = (R+0.3 T−R0.0 T)/R0.0 T, where R+0.3 T and R0.0 T are the resis-tances of the film at μ0H = +0.3 and 0.0 T, respectively, estimated by theconventional four-probe method. b) f dependence of ∆A/A, acquired atJc = 60 mA. The gray line represents the ∆R/R value estimated by theconventional four-probe method.It is noteworthy that if the AEE contribution is substantial andthe samples exhibit small or no hysteresis, the AEE contribu-tion can be eliminated for the GMR estimation by calculatingthe H-even-dependent component of A (i.e., Aeven) using LIT im-ages recorded at ±H and replacing AH with Aeven because theAEE signal exhibits the H-odd dependence.[48–52] However, forthe present CoFe/Cu multilayer film with tiny AEE and substan-tial hysteresis, we focus on raw LIT images.Now, we present the quantitative determination of the CIP-GMR ratio for the CoFe/Cu multilayer film utilizing the LITimages. Figure 3a represents ∆A/A = (A+0.3 T−A0.0 T)/A0.0 T as afunction of Jc with discrete black squares. Here, A+0.3 T and A0.0 Tdenote the averaged A values acquired at μ0H=+0.3 and 0.0 T, re-spectively. The data show the Jc independence of ∆A/A, which isconsistent with the nature of GMR. To compare ∆A/A with theconventional GMR ratio ∆R/R of the film, we further assessedthe Jc-dependent ∆R/R through conventional four-probe GMRmeasurements, as represented by gray circles in Figure 3a. Ad-ditionally, we compared ∆A/A obtained using f-dependent LITmeasurements with ∆R/R obtained using the four-probe methodin Figure 3b. The results confirm that the ∆A/A values estimatedusing the LIT technique agree with the GMR ratio ∆R/R obtainedvia the conventional four-probe measurements, thereby validat-ing the proposed LIT-based GMR measurement method. Thesefindings confirm that the GMR ratio can be accurately estimatedusing the LIT-based thermal images of Joule heating. It is impor-tant to note that the observed A due to Joule heating strongly de-pends on the thermal conductivity of the substrate through heatconduction. As the thermal conductivity increases, heat transferfrom the film to the substrate increases, leading to a decrease inmeasured A.[39,51] However, because the rate of heat transfer tothe substrate is independent of the positions, the relative changeof the A signal at each H is not affected by the substrate species.Thus, the GMR can be estimated from the Joule heating imagesregardless of the substrate choice. Nevertheless, films on sub-strates with lower thermal conductivities will provide larger A sig-nals, resulting in a better signal-to-noise ratio compared to thoseon substrates with high thermal conductivities substrate for thesame charge current value.2.2. Demonstration of High-Throughput MagnetoresistanceMaterial ScreeningTo demonstrate the utility of the proposed LIT-based GMR mea-surement method for high-throughput material screening, wepresent a case study involving CoCu-based granular single-layerfilms. The observation of the GMR effect using the LIT methodis possible not only for the FM/NM multilayers but also for vari-ous phase-separated granular films (e.g., Co–Cu, Co–Ag).[5,6,53,54]The GMR properties of Co100-qCuq granular films have been ex-tensively investigated across various composition regions by di-verse research groups over several decades.[5,6,42–44,54–56] Withinthe scope of these studies, the composition range of 70 ≤ q ≤ 90at.% is expected to exhibit large GMR ratios. However, exist-ing studies have primarily focused on selecting film composi-tions, resulting in a lack of systematic information for compo-sitional variations. The present study addresses this research gapby utilizing the proposed technique to systematically explore thecomposition-dependent GMR behavior of the Co100-qCuq granu-lar films.Employing a combinatorial deposition technique,[57–59] wefabricated two 100 nm-thick Co100-qCuq films with varyingcompositions from q = 65.0 to 100.0 at.% on single quartz glasssubstrates at room temperature. One of the films was used in theas-deposited state, while the other film was annealed at 573 Kfor 10 min after the deposition to boost the formation of thegranular structure. These films have linear composition gradi-ents spanning a length of 7.0 mm on a single substrate of a10.0 × 10.0 mm2 dimension (Figure 4a). The distribution ofthe film composition was confirmed by position-dependent com-position measurements using an electron probe microanalyzerwith a spot size of 5 μm in diameter, as shown in Figure 4b.Figure 4c,d shows the 2D image profiles of out-of-plane X-raydiffraction (XRD) measurements for the as-deposited and 573K-annealed Co100-qCuq composition-gradient films, respectively,measured at different positions along the composition gradient.In all composition regions for both films, the 2D XRD scans re-vealed an fcc diffraction pattern, similar to that of a pure Cu film,indicating that the Co is dispersed in the fcc-Cu matrix. However,as the Co concentration increased, the diffraction peak positionAdv. Physics Res. 2024, 3, 2400021 2400021 (4 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.comFigure 4. Fabrication of CoCu-based composition-gradient granular films. a) Schematic of the Co100-qCuq composition-gradient film structure, where auniform Cu layer and wedge-shaped Co and Cu layers were alternately deposited on a quartz glass substrate using a linear moving shutter and a sampleholder rotator. b) Position y dependence of the film composition measured by thermal emission electron probe microanalyzer. c,d) 2D XRD profilesfor the as-deposited c) and 573 K-annealed d) Co100-qCuq composition-gradient films measured at 1 mm intervals along the composition gradient. Thecompositions written on the 2D profile were estimated from the y-dependent composition data.shifted to a higher angle of 2𝜃 due to a decrease in lattice pa-rameters. The ring-shaped diffraction pattern observed in the 2Dimages indicates the polycrystalline nature of the films.We estimated the composition dependence of GMR for theas-deposited and 573 K-annealed Co100-qCuq films utilizing LITimages. We fabricated wire-shaped structures from both filmshaving a width of 1.0 mm and a length of 8.0 mm, with thecomposition gradient along the length direction. During theLIT measurements, the wires from the as-deposited and 573 K-annealed films were electrically connected, forming a U-shapedconfiguration with charge currents flowing in opposite direc-tions along the length (composition gradient) direction, as il-lustrated in Figure 5a. For the GMR estimation, we collected Aand ϕ images induced by an ON/OFF-modulated charge currentwith Jc = 60 mA and f = 25.0 Hz at μ0H = +0.9 and 0.0 T. Itshould be noted that the LIT measurements at a high lock-in fre-quency are important to suppress the effect of thermal diffusionfor composition-gradient film.[39,40] H was applied in a directionperpendicular to the composition gradient. Following the previ-ously described procedure, we extracted ∆A ( = A+0.9 T−A0.0 T) uti-lizing the A images acquired at μ0H = +0.9 and 0.0 T, as shownin Figure 5b. Here, we neglect the AEE contribution to the mea-sured A images at μ0H = +0.9 T due to its negligibly small valueas compared to the Joule heating contribution. Subsequently, weestimated composition-dependent ∆A/A, utilizing the line pro-files for the ∆A image and the A image at μ0H = 0.0 T obtainedby averaging data points over 0.6 mm along the width directionrepresented by the white dashed region in Figure 5b. Figure 5cshows the q dependence of ∆A/A for the as-deposited and 573 K-annealed Co100-qCuq films. The data reveal that the ∆A/A, repre-senting the GMR ratio for the film, gradually increases beyond q> 65 at.%, reaching a peak near q= 78 at.% for the 573 K-annealedfilm and q = 80 at.% for the as-deposited film, and then graduallydecreases to zero for pure Cu (q = 100 at.%). This observed trendin the q-dependence of the GMR ratio for the Co100-qCuq filmsaligns well with previous findings reported in the literature.[42]The outcomes distinctly identify the optimal compositions forlarge room-temperature GMR for both films simultaneously, allobtained through the single LIT measurement. Additionally, thecomposition resolution achieved with the LIT-based GMR mea-surement is significantly higher compared to conventional GMRmeasurement methods. For instance, each data point in Figure 5c(corresponding to a pixel in the LIT image of Figure 5b) repre-sents a composition change of ≈0.075 at.%, which can be fur-ther improved by decreasing the composition gradient. Althoughthe actual special resolution is worse than the pixel size due tothe temperature broadening in the black ink layer, the composi-tion resolution is still sufficiently excellent to accelerate the GMRstudies. Interestingly, the data in Figure 5c reveal that the opti-mal compositions for maximum GMR at room temperature dif-fer between as-deposited and 573 K-annealed Co100-qCuq films.This difference can be attributed to the formation of Co precip-Adv. Physics Res. 2024, 3, 2400021 2400021 (5 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.comFigure 5. LIT-based high-throughput measurement of composition-dependent GMR for CoCu-based granular films. a) Schematic of the experimentalsetup designed for the concurrent measurement of GMR in the as-deposited and 573 K-annealed Co100-qCuq composition-gradient films. b) A, ϕ,∆A, and ϕ∆A images resulting from Joule heating of the films, acquired at Jc = 60 mA, f = 25.0 Hz, and μ0H of +0.9 and 0.0 T and AAEE and ϕAEEimages resulting from AEE of the films, acquired at |μ0H| = 0.9 T. Here, ∆A represents A+0.9 T−A0.0 T with A+0.9 T and A0.0 T respectively being the Avalues at μ0H = +0.9 and 0.0 T and ϕ∆A represents the lock-in phase of ∆A. In the inset graph, ϕ∆A ≈200° signifies A0.0 T > A+0.9 T. During the AEEmeasurement, Jc = 60 mA, f = 25 Hz, and zero DC offset was applied to the sample. c) q dependence of ∆A/A for the as-deposited and 573 K-annealedCo100-qCuq composition-gradient films, simultaneously estimated using the proposed LIT-based GMR measurement technique. The black and red stripsrepresent the q values exhibiting the largest −∆A/A value for as-deposited and 573 K-annealed Co100-qCuq composition-gradient films, respectively. d)H dependence of R for the 573 K-annealed Co100-qCuq composition gradient film near the optimum composition (Co22Cu78) at 300 K. The maximumGMR ratio ∆R/R = (R+0.9 T−R0.0 T)/R0.0 T for the optimum composition at room temperature was determined to be ≈8.1% for μ0H = +0.9 T, which isconsistent with the GMR ratio obtained by the LIT-based GMR measurement technique. R+0.9 T and R0.0 T represent the resistances at μ0H = +0.9 and0.0 T, respectively. e) H dependence of R at 4 K for the optimum composition. The saturation GMR ratio for the film was determined using the relation∆R/R = (RH−Rmax)/Rmax, where Rmax and RH represent the maximum and minimum resistances of the film near zero and at μ0H = +5.0 T, respectively.itates due to phase decomposition during the annealing treat-ment at 573 K, as reported in the literature.[60] While a detailedmicrostructural analysis of the films is beyond the scope of thiswork, the present study successfully demonstrates the effective-ness of the proposed LIT-based GMR measurement method forinvestigating composition-dependent GMR behaviors.This high-throughput and systematic GMR measurement re-veals that Co22Cu78 in the 573 K-annealed film exhibits a largeGMR ratio (−∆R/R ≡ −∆A/A) of ≈8.8 ± 0.9% at μ0H = +0.9 Tand room temperature. This GMR ratio is the highest recordedamong polycrystalline CoCu-based single-layer granular films ofany composition.[42–45] Notably, a larger saturated GMR ratio of20% has been reported for (111)-oriented epitaxial Co–Cu filmsin a previous study at room temperature.[54] However, in the sameresearch, a much smaller saturated GMR ratio of 7% was ob-served at room temperature for (001)-oriented epitaxial Co–Cufilms,[54] reflecting the high dependency of GMR on crystal orien-tation for epitaxial Co–Cu films. We note that the previous studydefined the saturated GMR ratio differently, using the maximumchange in R over the measurement H range divided by the R athigh H. Following the same approach for our current film yieldsa GMR ratio of ≈9.6 ± 0.8% at μ0H = +0.9 T and room tempera-ture. While larger GMR ratios can be achieved in epitaxial filmson single-crystalline substrates, their use in commercial applica-tions is limited due to the complexity of film fabrication and thehigh cost of single-crystalline substrates. Therefore, the observa-tion of a large GMR ratio in polycrystalline CoCu-based single-layer granular films at room temperature is crucial.To confirm the best GMR value obtained from the LIT-basedhigh-throughput screening, we conducted conventional four-probe GMR measurements for the same film near the optimalcomposition region. For the measurement, we used the 573 K-annealed Co100-qCuq composition-gradient film patterned intoparallel wire shapes perpendicular to the direction of the compo-sition gradient. Each wire had a width of 0.4 mm and a spacingof 0.4 mm. It is important to note that we can neglect the compo-sitional variation along the width direction for these wires, whichis ≈2.0%. Figure 5d shows the H dependence of R for the wirenear the optimum composition of Co22Cu78 recorded at 300 K.Utilizing the relation ∆R/R = (R+0.9 T−R0.0 T)/R0.0 T, where R+0.9 Tand R0.0 T respectively represent the resistances at μ0H=+0.9 and0.0 T, we determined the GMR ratio for the film composition toAdv. Physics Res. 2024, 3, 2400021 2400021 (6 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.combe 8.1% which is consistent with the results obtained from theLIT-based measurements. In Figure 5e, we present the H depen-dence of R for the same film at 4 K. The data reflect the saturationbehavior of the R of the film with minimum R at high H and amaximum R at finite ±H, which can be correlated with the coer-civity field of the film.[6] Since the maximum R occurs at finite±H, not at zero, we use (RH−Rmax)/Rmax instead of (RH−R0)/R0to determine the saturation GMR ratio for the film at 4 K, whereRmax and RH represent the maximum and minimum resistancevalues of the film, respectively (Figure 5e). The saturation GMRratio obtained using the above relation was observed to be 28.0%.In comparison, for a previously reported Co20Cu80 single-layergranular film, fabricated by magnetron sputtering onto a quartzglass substrate, a maximum GMR ratio of 16.5% was achieved at5 K at μ0H = 5.0 T.[6]This demonstration clearly illustrates the effectiveness of theLIT-based GMR measurement method in accelerating the opti-mization of materials and devices, presenting clear advantagesover traditional GMR measurements. This method eliminatesthe need to fabricate many films or numerous Hall bars, whichare often required for conventional GMR measurements, mak-ing the process time-consuming and laborious. Furthermore,the LIT-based method utilizing composition-gradient and wedge-shaped films significantly enhances composition and thicknessresolution, respectively, overcoming inherent limitations in con-ventional GMR measurement techniques that might lead tooverlooking optimal compositions and layer thicknesses. No-tably, for multilayers consisting of two-dimensional composi-tion/thickness gradients, the method demonstrated here is ef-fective for determining not only the optimum composition butalso the optimal thickness of each layer. The utilization of theproposed LIT-based GMR measurement method with such two-dimensional multilayers to simultaneously assess multiple filmparameters will further boost the optimization process of GMRmaterials and devices.3. ConclusionWe proposed and demonstrated a high-throughput screeningmethod for assessing the GMR properties of materials by utiliz-ing thermal imaging of Joule heating measured by the LIT tech-nique. By utilizing the CoFe/Cu multilayer film, we confirmedthe accurate and reliable estimation of the GMR ratio using thisproposed method. Furthermore, the usefulness of this approachwas demonstrated through its application to the composition-gradient CoCu-based magnetic granular films. The demonstra-tion utilizing multiple composition-gradient films simultane-ously highlighted the practicality of the proposed method forhigh-throughput optimization of film composition with preci-sion. Furthermore, the high-throughput investigation revealed apreviously unexplored CoCu-based granular composition with aGMR ratio exceeding 8% at μ0H= 0.9 T at room temperature, anda saturation GMR ratio of ≈28% at 4 K, setting new record highvalues for polycrystalline CoCu-based single-layer granular ma-terials. This advancement holds great promise for significantlyimproving the efficiency of material screening and design opti-mization processes for spintronics technology.4. Experimental SectionFabrication of CoFe/Cu Multilayer Film: The CoFe/Cu multi-layer film used in this study possessed the following structure:Co50Fe50(3.0 nm)/[Cu(1.6 nm)/Co50Fe50(3.0 nm)]33, where the numbersin the parentheses denote the thicknesses of the layers and the subscript33 denotes the number of the repeated depositions. This film was de-posited on a single-crystalline MgO (001) substrate at room temperatureusing magnetron sputtering. A 5 nm-thick Al layer was deposited on themultilayer without breaking the vacuum to prevent sample oxidation. Theas-deposited film was post-annealed at 523 K in a vacuum while applyingan in-plane magnetic field of 0.3 T for 1 h. This film is identical to thatutilized in the previous study,[46] where comprehensive details regardingits fabrication conditions are available.Fabrication of Co100-qCuq Composition-Gradient Granular Film: TheCo100-qCuq composition-gradient films with a thickness of 100 nm werefabricated on 10 × 10 mm2 quartz glass substrates using a magnetronsputtering system (Comet, Inc., CMS-3200). Before deposition, the cham-ber was evacuated at a base pressure of < 6.0 × 10−6 Pa using a cry-opump. Then, the films were deposited at a process Ar gas pressure of0.4 Pa at ambient temperature using DC power sources. The previouslyestablished layer-by-layer wedge-shaped deposition process was followedto achieve a composition variation of 65 ⩽ q ⩽ 100 at.% over a lengthof 7.0 mm on a single substrate.[57–59] This process involved three con-secutive steps: I) initial deposition of a uniform Cu layer, followed by thedeposition of a wedge-shaped Cu layer using a linear moving shutter overa length of 7.0 mm, II) 180° rotation of the substrate, and III) subsequentdeposition of a wedge-shaped Co layer. After completing the steps (I)–(III), the total thickness was designed to be 0.5 nm. This sequence wasrepeated 200 times to obtain the 100 nm-thick composition-gradient film.The deposition rates of the Cu and Co layers were maintained at 0.023 and0.025 nms−1, respectively. The deposition duration at each step was opti-mized to obtain the desired composition. A 2 nm-thick Al layer was appliedas a cap to prevent oxidation. One of the room-temperature-depositedfilms was annealed at 573 K for 10 min in a vacuum atmosphere to boostthe formation of the granular structure.AcknowledgementsThis work was supported by ERATO “Magnetic Thermal ManagementMaterials” (grant no. JPMJER2201) from JST, Japan, the Grant-in-Aid forScientific Research (S) (grant no. 22H04965), and the Grant-in-Aid forEarly-Career Scientists (grant no. 21K14519) from JSPS KAKENHI, Japan,and NEC Corporation. R.M. was supported by JSPS through the “JSPSPostdoctoral Fellowship for Research in Japan (Standard)” (grant no.P21064).Conflict of InterestThe authors declare no conflict of interest.Data Availability StatementThe data that support the findings of this study are available from the cor-responding author upon reasonable request.Keywordscombinatorial deposition, giant magnetoresistance effect, granular films,high-throughput material screening, lock-in thermographyReceived: February 14, 2024Revised: April 21, 2024Published online: May 24, 2024Adv. Physics Res. 2024, 3, 2400021 2400021 (7 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.comwww.advancedsciencenews.com www.advphysicsres.com[1] M. N. Baibich, J. M. Broto, A. Fert, F. N. Van Dau, F. Petroff, P. Etienne,G. Creuzet, A. Friederich, J. Chazelas, Phys. Rev. Lett. 1988, 61, 2472.[2] G. Binasch, P. Grünberg, F. Saurenbach, W. Zinn, Phys. Rev. B 1989,39, 4828.[3] S. S. P. Parkin, N. More, K. P. Roche, Phys. Rev. Lett. 1990, 64, 2304.[4] S. S. P. Parkin, Appl. Phys. Lett. 1992, 61, 1358.[5] A. E. Berkowitz, J. R. Mitchell, M. J. Carey, A. P. Young, S. Zhang, F. E.Spada, F. T. Parker, A. Hutten, G. Thomas, Phys. Rev. Lett. 1992, 68,3745.[6] J. Q. Xiao, J. S. Jiang, C. L. Chien, Phys. Rev. Lett. 1992, 68, 3749.[7] K. B. Fathoni, Y. Sakuraba, Y. Miura, T. Sasaki, T. Nakatani, K. Hono,J. Phys. D. Appl. Phys. 2022, 55, 125001.[8] K.-M. Lenssen, H. van Kesteren, T. G. S. Rijks, J. C. Kools, M. deNooijer, R. Coehoorn, W. Folkerts, Sensors Actuators A Phys 1997, 60,90.[9] K. Kapser, M. Weinberger, W. Granig, P. Slama, Smart Sensors,Measurement and Instrumentation, Springer International Publishing,New York City 2013, 133.[10] J. C. Rife, M. M. Miller, P. E. Sheehan, C. R. Tamanaha, M. Tondra, L.J. Whitman, Sensors Actuators A Phys 2003, 107, 209.[11] S. Ota, A. Ando, D. Chiba, Nat. Electron. 2018, 1, 124.[12] S. Yamada, K. Chomsuwan, M. Iwahara, 5th IEEE Conference on Sen-sors, IEEE, Daegu, Korea (South), 2006, 927.[13] M. Ha, G. S. Cañón Bermúdez, J. A.-C. Liu, E. S. Oliveros Mata, B. A.Evans, J. B. Tracy, D. Makarov, Adv. Mater. 2021, 33, 2008751.[14] P. D. Kulkarni, H. Iwasaki, T. Nakatani, Sensors 2022, 22, 9385.[15] K. Wu, D. Tonini, S. Liang, R. Saha, V. K. Chugh, J.-P. Wang, ACS Appl.Mater. Interfaces 2022, 14, 9945.[16] S. Liang, P. Sutham, K. Wu, K. Mallikarjunan, J.-P. Wang, Sensors 2022,22, 5663.[17] Y. Dong, T. Xu, W. Jiang, Appl. Phys. Lett. 2021, 119, 192403.[18] M. Ha, G. S. Cañón Bermúdez, T. Kosub, I. Mönch, Y. Zabila, E. S.Oliveros Mata, R. Illing, Y. Wang, J. Fassbender, D. Makarov, Adv.Mater. 2021, 33, 2005521.[19] D. Su, K. Wu, K. Srinivasan, Z. Nemati, R. Zamani, V. Chugh, R. Saha,R. Franklin, J. Modiano, B. Stadler, J. Wang, Adv. Mater. Interfaces2023, 10, 2201417.[20] T. Miyazaki, N. Tezuka, J. Magn. Magn. Mater. 1995, 139, L231.[21] J. S. Moodera, L. R. Kinder, T. M. Wong, R. Meservey, Phys. Rev. Lett.1995, 74, 3273.[22] S. Yuasa, T. Nagahama, A. Fukushima, Y. Suzuki, K. Ando, Nat. Mater.2004, 3, 868.[23] S. S. P. Parkin, C. Kaiser, A. Panchula, P. M. Rice, B. Hughes, M.Samant, S.-H. Yang, Nat. Mater. 2004, 3, 862.[24] T. Scheike, Z. Wen, H. Sukegawa, S. Mitani, Appl. Phys. Lett. 2023,122, 112404.[25] L. Piraux, J. M. George, J. F. Despres, C. Leroy, E. Ferain, R. Legras, K.Ounadjela, A. Fert, Appl. Phys. Lett. 1994, 65, 2484.[26] A. Vedyayev, M. Chshiev, N. Ryzhanova, B. Dieny, C. Cowache, F.Brouers, J. Magn. Magn. Mater. 1997, 172, 53.[27] H. Yuasa, M. Yoshikawa, Y. Kamiguchi, K. Koi, H. Iwasaki, M.Takagishi, M. Sahashi, J. Appl. Phys. 2002, 92, 2646.[28] R. Saeki, S. Mizoguchi, H. Kamimura, M. Hayashida, T. Ohgai, J.Magn. Magn. Mater. 2021, 529, 167849.[29] B. Dieny, P. Humbert, V. S. Speriosu, S. Metin, B. A. Gurney, P.Baumgart, H. Lefakis, Phys. Rev. B 1992, 45, 806.[30] M. Xu, Y. Fan, G. Luo, Z. Mai, Phys. Lett. A 2000, 272, 282.[31] D. Meziane Mtalsi, M. El Harfaoui, A. Qachaou, M. Faris, J. BenYoussef, H. Le Gall, Phys. status solidi 2001, 187, 633.[32] I. Bakonyi, E. Simon, B. G. Tóth, L. Péter, L. F. Kiss, Phys. Rev. B 2009,79, 174421.[33] M. Suzuki, Y. Taga, A. Goto, H. Yasuoka, J. Magn. Magn. Mater. 1993,126, 495.[34] W. Kuch, A. C. Marley, S. S. P. Parkin, J. Appl. Phys. 1998, 83, 4709.[35] P. D. Kulkarni, T. Nakatani, Z. Li, T. Sasaki, Y. Sakuraba, J. Magn. Magn.Mater. 2022, 560, 169562.[36] K. B. Fathoni, Y. Sakuraba, T. Sasaki, Y. Miura, J. W. Jung, T. Nakatani,K. Hono, APL Mater. 2019, 7, 111106.[37] D. Taparia, K. B. Fathoni, P. D. Kulkarni, A. Srinivasan, P. Alagarsamy,Y. Sakuraba, T. Nakatani, K. Hono, J. Magn. Magn. Mater. 2021, 538,168321.[38] O. Breitenstein, W. Warta, M. Langenkamp, Lock-In Thermography:Basics and Use for Evaluating Electronic Devices and Materials, SpringerBerlin Heidelberg, Berlin, Heidelberg 2010, 10.[39] S. Daimon, R. Iguchi, T. Hioki, E. Saitoh, K. Uchida, Nat. Commun.2016, 7, 13754.[40] S. Daimon, K. Uchida, R. Iguchi, T. Hioki, E. Saitoh, Phys. Rev. B 2017,96, 024424.[41] K. Uchida, S. Daimon, R. Iguchi, E. Saitoh, Nature 2018, 558, 95.[42] A. N. Pohorilyi, A. F. Kravetz, E. V. Shipil, A. Y. Vovk, C. S. Kim, H. R.Khan, J. Magn. Magn. Mater. 1998, 186, 87.[43] G. Wen, H. Zhao, J. Zhao, X. Zhang, Mater. Sci. Eng. C 2001, 16,81.[44] I. O. Shpetnyi, D. M. Kondrakhova, S. I. Vorobiov, B. Scheibe, V. I.Grebinaha, D. O. Derecha, Y. I. Gorobets, I. Y. Protsenko, J. Magn.Magn. Mater. 2019, 474, 624.[45] Z. Yan, B. Wang, X. W. Lv, W. B. Sui, J. W. Cao, H. G. Shi, M. S. Si, D.Z. Yang, D. S. Xue, Appl. Phys. Lett. 2020, 116, 42401.[46] T. Hirai, Y. Sakuraba, K. Uchida, Appl. Phys. Lett. 2022, 121, 162404.[47] H. Straube, O. Breitenstein, J. Appl. Phys. 2011, 109, 064515.[48] T. Seki, R. Iguchi, K. Takanashi, K. Uchida, Appl. Phys. Lett. 2018, 112,152403.[49] T. Seki, R. Iguchi, K. Takanashi, K. Uchida, J. Phys. D. Appl. Phys. 2018,51, 254001.[50] T. Seki, A. Miura, K. Uchida, T. Kubota, K. Takanashi, Appl. Phys. Ex-press 2019, 12, 023006.[51] R. Das, R. Iguchi, K. Uchida, Phys. Rev. Appl. 2019, 11, 034022.[52] R. Modak, K. Uchida, Appl. Phys. Lett. 2020, 116, 032403.[53] J. Q. Xiao, J. S. Jiang, C. L. Chien, Phys. Rev. B 1992, 46, 9266.[54] S. S. P. Parkin, R. F. C. Farrow, T. A. Rabedeau, R. F. Marks, G. R. Harp,Q. Lam, C. Chappert, M. F. Toney, R. Savoy, R. Geiss, Europhys. Lett.1993, 22, 455.[55] K. Miyazaki, S. Kainuma, K. Hisatake, T. Watanabe, N. Fukumuro,Electrochim. Acta 1999, 44, 3713.[56] V. V. Hiep, N. Chau, D. M. Hong, N. H. Luong, J. Magn. Magn. Mater.2007, 310, 2524.[57] H. Masuda, R. Modak, T. Seki, K. Uchida, Y.-C. Lau, Y. Sakuraba, R.Iguchi, K. Takanashi, Commun. Mater. 2020, 1, 75.[58] R. Modak, K. Goto, S. Ueda, Y. Miura, K. Uchida, Y. Sakuraba, APLMater. 2021, 9, 031105.[59] R. Modak, Y. Sakuraba, T. Hirai, T. Yagi, H. Sepehri-Amin, W. Zhou,H. Masuda, T. Seki, K. Takanashi, T. Ohkubo, K. Uchida, Sci. Technol.Adv. Mater. 2022, 23, 767.[60] T. Sugawara, K. Takanashi, H. Fujimori, J. Magn. Magn. Mater. 1998,177–181, 951.Adv. Physics Res. 2024, 3, 2400021 2400021 (8 of 8) © 2024 The Author(s). Advanced Physics Research published by Wiley-VCH GmbH 27511200, 2024, 8, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/apxr.202400021 by National Institute For, Wiley Online Library on [11/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advphysicsres.com High-Throughput Optimization of Magnetoresistance Materials Based on Lock-In Thermography 1. Introduction 2. Results and Discussion 2.1. Validation of LIT-Based GMR Measurement Method 2.2. Demonstration of High-Throughput Magnetoresistance Material Screening 3. Conclusion 4. Experimental Section Acknowledgements Conflict of Interest Data Availability Statement Keywords