Tailoring coercive fields and the Curie temperature via proximity coupling in WSe2/Fe3GeTe2 van der Waals heterostructures
Abstract
Hybrid structures consisting of two-dimensional (2D) magnets and semiconductors have exhibited extensive functionalities in spintronics and opto-spintronics. In this work, we have fabricated WSe2/Fe3GeTe2 van der Waals (vdW) heterostructures and investigated the proximity effects on 2D magnetism. Through reflective magnetic circular dichroism (RMCD), we have observed a temperature-dependent modulation of magnetic order in the heterostructure. For temperatures above , WSe2-covered Fe3GeTe2 exhibits a larger coercive field than that observed in bare Fe3GeTe2, accompanied by a noticeable enhancement of the Curie temperature by . This strengthening suggests an increase in magnetic anisotropy in the interfacial Fe3GeTe2 layer, which can be attributed to the spin-orbit coupling (SOC) proximity effect induced by the adjacent WSe2 layers. However, at much lower temperatures (), a non-monotonic modification of the coercive field is observed, showing both reduction and enhancement, which depends on the thickness of the WSe2 and Fe3GeTe2 layers. Moreover, an unconventional two-step magnetization process emerges in the heterostructure, indicating the short-range nature of SOC proximity effects. Our findings revealing proximity effects on 2D magnetism may shed light on the design of future spintronic and memory devices based on 2D magnetic heterostructures.
, ,
-
April 30, 2024
-
Keywords: vdW heterostructure, 2D magnetism, spin-orbit coupling, proximity effect
1 Introduction
The integration of magnetic materials with semiconductors has been extensively explored in spintronics through active manipulations of the spin degrees of freedom[1, 2]. This exploration has led the investigation of various prototypical devices, such as spin field-effect transistors[3] and magnetic tunnel junctions[4, 5], to unlock fascinating functionalities within these hybrid structures. The interfaces between magnets and semiconductors play a critical role in determining these novel properties. Recently, the utilization of two-dimensional materials to form van der Waals (vdW) magnetic heterostructures has provided a new platform for the development of ultrathin spintronic devices[6, 7, 8, 9], owing to their high-quality crystalline structures and bond-free interfaces[10, 11, 12].
Incorporating 2D magnets with 2D semiconductors in a vdW heterostructure enables a wide range of intriguing phenomena through magnetic proximity effects. For instance, WSe2/CrI3 heterostructures demonstrate a spontaneous emergence of both valley Zeeman splitting and polarization[13, 14]. The former is highly sensitive to the overall magnetization while the latter is dominated by the interfacial layer of CrI3 and can be controlled through electrical gating[15, 16]. Additionally, the magnetic proximity effect in MoSe2/CrBr3 exhibits a dependency on the charge state[17]. The examples mentioned above focus on the impact of magnetic layers on nonmagnetic semiconductors. Conversely, there is growing interest in modulating 2D magnets through adjacent nonmagnetic layers[18, 19, 20, 21]. In 2D magnets, spin-orbit coupling (SOC) from heavy elements is crucial for magnetic anisotropy, which sustains long-range magnetic order even at the monolayer limit[22, 23]. It has been proposed that combining 2D magnets with materials that exhibit strong SOC can effectively enhance the magnetic anisotropy[24, 25, 26, 27, 21], resulting in a modification of the magnetic order. However, experimental studies are few, which require extensive efforts.
Here, we investigate the influence of proximity effects on a magnetic layer in vdW heterostructures, consisting of a 2D itinerant magnet Fe3GeTe2 (FGT) and a 2D semiconductor WSe2 with strong SOC[28, 7]. Through reflective magnetic circular dichroism (RMCD) studies, we observe a pronounced enhancement of the coercive field and Curie temperature in the WSe2/Fe3GeTe2 heterostructure compared with bare FGT, arising from SOC occurring at the interface. However, a non-monotonic tuning of coercive fields, including both reduction and enhancement, emerges in WSe2/Fe3GeTe2 regions at much lower temperatures (), depending on the thickness of WSe2 and Fe3GeTe2 layers. Moreover, an unconventional magnetic stratification is observed in the heterostructure with monolayer WSe2, leading to an obvious two-step magnetization reversal, which reveals the short-range nature of SOC proximity effects.
2 Results and discussions
Figure 1a shows the schematic of the WSe2/Fe3GeTe2 heterostructure, assembled by an all-dry viscoelastic stam** method[29]. In our experiments, we exfoliated Fe3GeTe2 flakes onto a Si substrate with SiO2. The thin WSe2 flakes were exfoliated on polydimethylsiloxane (PDMS). Then we transferred the WSe2 onto the Fe3GeTe2 flake via a high-precision transfer setup to construct the final heterostructure. The fabrication was carried out inside a glovebox with an argon atmosphere to avoid the reaction with oxygen and moisture. Figure 1b shows an optical image of the heterostructure H1 with few-layer FGT and bilayer WSe2. The thickness of the FGT flake is (about 7 layers) determined by atomic force microscopy (AFM) (see the inset of Fig. 1b). The room-temperature photoluminescence (PL) spectra of the WSe2 flakes (Fig. 1c) exhibit a distinct dependence on thickness, in which a strong single peak centered at emerges in the monolayer WSe2[30]. This allows us to identify the layer numbers of the corresponding WSe2 flakes in heterostructures.
To investigate the WSe2-induced proximity effect on magnetic order, we performed a series of RMCD measurements by an optical cryostat. A HeNe laser with excitation power of was used and focused onto the sample (Fig. 1a) with a spot size of about . Figures 1d,e show a comparison of the RMCD signals of WSe2-covered FGT (WSe2+FGT) region and the bare FGT region from the same heterostructure H1 (bilayer WSe2/6.2 nm FGT), at the temperature of and , respectively. The typical rectangular hysteresis loops emerge in both regions. Notably, we observe an opposite effect on the coercive field at different temperatures. At , the loop in the WSe2+FGT region (red curve at the bottom) is broader than that measured in bare FGT (blue curve at the top), as shown in Fig. 1d, while exhibits a reduction in the WSe2+FGT region at .
Figures 2a,b present the temperature dependence of magnetic hysteresis loops in sample H1, compared between the bare FGT and WSe2+FGT regions. As temperature increases, the hysteresis loop shrinks rapidly in both regions due to the stronger thermal fluctuation. It disappears at in bare FGT, implying a transition from the ferromagnetic phase to paramagnetic phase, but retains a rectangular loop in the WSe2+FGT region under the same conditions. Figure 2c shows the remanent RMCD signal as a function of temperature, in which we also compare the results for bare FGT (blue squares) and WSe2+FGT (red circles). The corresponding dashed curves are plotted by the critical power-law form , by which we extract the Curie temperature and the critical exponent . The of WSe2-covered FGT is increased by due to the WSe2 proximity effect ( in bare FGT and for WSe2+FGT). The value of we obtain is , consistent with the previous report on a few layers of FGT[23]. Figure 2d displays the corresponding extracted as a function of temperature for both two regions. The enhancement of can be observed in the same temperature range. The significant increase in and the simultaneous enhancement of confirm the improved magnetism in FGT, which can be attributed to the proximity effect from the adjacent WSe2.
In contrast to the enhancement of coercive fields at higher temperatures (), as shown in Fig. 2, WSe2-covered FGT demonstrates a distinct reduction of () at (see Fig. 3a). We measured the layer dependence of WSe2 on , as shown in Fig. 3a, demonstrating a discernible reduction in with the increase in layer numbers. Figures 3b,c depict the difference of coercive fields () as a function of temperature in both 2L-WSe2/FGT and 5L-WSe2/FGT regions, where and represent the extracted coercive field in WSe2-covered FGT and bare FGT, respectively. As shown in Fig. 3b, the negative associated with the suppression of magnetic anisotropy is evident in both 2L- and 5L-WSe2 covered regions at low temperatures, persisting as the temperature increases until . in 5L-WSe2+FGT is smaller than that in 2L-WSe2+FGT, suggesting a stronger suppression of magnetic anisotropy. Additionally, as the temperature rises above , of the 2-layer (red circles) and the 5-layer (yellow triangles) WSe2+FGT are almost equal within the margin of error. Figure 3c presents the results for the overall temperature range below . The sign of changes from negative (blue panel) to positive (red panel) at approximately , and the regions with different thicknesses of WSe2 show no pronounced difference at elevated temperatures.
Thus far, our observations have revealed the coexistence of the proximity-induced enhancement and suppression of magnetic anisotropy in FGT, suggesting the likelihood of multiple underlying mechanisms. In 2D magnets, magnetic anisotropy arising from spin-orbit coupling plays a crucial role in stabilizing magnetic order. It has been proposed that attaching a nonmagnetic layer with strong SOC to 2D magnets can effectively enhance the magnetic anisotropy[27]. Recent experiments involving analogous heterostructures[18, 21] reveal the stabilization of 2D magnetism in thick FGT flakes, which is induced by the SOC proximity effect. Therefore, it is reasonable to ascribe the enhancing magnetic anisotropy to the SOC proximity effect. On the other hand, the magnetic anisotropy of FGT can also be modulated by strain engineering[31, 32]. Given that the lattice constant in WSe2 is smaller than that in FGT, a slightly compressive strain may occur in the -plane of FGT[24], potentially contributing to the suppression of magnetic anisotropy at low temperatures.
Figure 4a shows the RMCD measurements in two WSe2+FGT heterostructures with varying thicknesses of FGT but the same thickness of WSe2 flakes (2 layers). The heterostructure with a thinner FGT exhibits a more pronounced enhancement of at . We present the relative difference as a function of temperature in Fig. 4b. It is noteworthy that the reduction of at low temperatures is only observable in the thinnest FGT (), but is absent in the heterostructures with thicker FGT, which implies that the potential strain effect rapidly diminishes as the FGT thickness increases. Figure 4c illustrates the dependence of on FGT thicknesses. At the thickness of , of WSe2/FGT and bare FGT are nearly identical; however, a significant difference is observed at the thickness of 6.2 nm, showing that the enhancement induced by SOC proximity is strengthened with decreasing thickness of FGT.
To explore more details of this interfacial SOC proximity effect, we fabricated another heterostructure H2 for RMCD measurements, consisting of a monolayer WSe2 and a -thick FGT layer. A thin h-BN was used to covered the heterostructure as a comparison. Figures 5a,b show the normalized RMCD loops at several fixed temperatures (solid curves for WSe2-covered FGT and dashed curves for bare FGT). As the temperature increases, H2 generates four distinct regions marked by Roman numerals. In region I, the enhancement of is present but diminishes above in region II. No observable difference of persists until . In region III, the enhanced recovers at , accompanied by the emergence of a two-step magnetization reversal (as marked by red arrows in Fig. 5b). Unlike the rectangle loop in the bare FGT, the WSe2-covered FGT exhibits an evident two-step hysteresis loop at about , with two distinct coercive fields and () during the magnetization process. The two-step loop is unstable at elevated temperatures, vanishing at but reappearing in region IV. We depict the modulation as a function of temperature in Fig. 5c, defining the modulation strength as the ratio of in the WSe2-covered FGT to that in the bare FGT. The two-step loop is only observable with separate coercive fields at higher temperatures. In addition, the enhancement of in H2 with hBN encapsulation confirms that the altered magnetic order results from the interfacial proximity effect rather than the degeneration (see Figure S3).
The two-step magnetization reversal in the ferromagnetic layer can be achieved by spin exchange coupling with an adjacent antiferromagnetic layer[33, 34]. However, this is not the scenario in this work as WSe2 is a nonmagnetic semiconductor. Figure 5d shows the two-step hysteresis loop manifesting four stable magnetic states, labeled by red numbers from 1 to 4. The additional intermediate states (state 2 and state 4) could be explained by the short-range nature of the SOC proximity effect. Assuming that the FGT layers fall into two parts due to the proximity of WSe2 (see the schematics in Fig. 5e). For the FGT layers at the interface, the SOC proximity effect is dominant, leading to enhanced magnetic anisotropy and a larger coercive field. While the underneath FGT layers are minimally affected by the interface, resulting in a smaller magnetic field required for flip**. Thus the arrangement of magnetic moments can be illustrated by schematics in Fig. 5e. States 1 and 3 exhibit fully spin-polarized magnetization with all-spin-up and all-spin-down configurations, respectively. While the magnetizations in the layers adjacent to and far away from the interface point in opposite directions, resulting in intermediate states 2 and 4. A similar two-step hysteresis loop induced by SOC proximity was recently observed in the MoS2/FGT heterostructure[19].
At low temperatures, characterized by reduced thermal fluctuation, the interlayer exchange coupling is sufficiently strong to maintain a uniform in FGT layers. However, the interlayer exchange coupling diminishes with strong thermal fluctuation at elevated temperatures, resulting in non-uniform coercive fields for the interfacial and underneath FGT layers, i.e., the two-step magnetization reversal process. At much lower temperatures (), the proximity effect exhibits a non-monotonic modification depending on the thickness of the WSe2 and the FGT layers. The enhancement of emerges in monolayer-WSe2/FGT, but disappears in others. While the reduced is observed with the thinnest FGT (). This may suggest a complex competition between different mechanisms including interlayer exchange coupling, SOC proximity, interfacial strain, thermal fluctuation and so on, which requires further investigation in the future.
In conclusion, we systematically investigated the proximity effect on 2D magnetism in WSe2/FGT vdW heterostructures. Both and can be noticeably enhanced in the WSe2-covered region, which is more pronounced as the FGT thickness decreases. This enhancement is attributed to the SOC proximity effect induced from the adjacent WSe2. At low temperatures, we observed a reduction of the coercive field in the heterostructure with a thick FGT flake, potentially due to the strain effect from the interface. In addition, the presence of unconventional magnetic stratification demonstrates the short-range nature of the SOC proximity effect. Our findings offer a practical method for modifying 2D magnetic order by combining a 2D nonmagnetic semiconductor with strong SOC, thereby promising for advancing the field of spintronics based on vdW magnetic heterostructures.
References
- [1] Prinz G A 1990 Science 250 1092–1097
- [2] uti I, Fabian J and Das Sarma S 2004 Rev. Mod. Phys. 76 323–410
- [3] Datta S and Das B 1990 Appl. Phys. Lett. 56 665–667
- [4] Moodera J S, Kinder L R, Wong T M and Meservey R 1995 Phys. Rev. Lett. 74 3273–3276
- [5] Miyazaki T and Tezuka N 1995 J. Magn. Magn. Mater. 139 L231–L234
- [6] Ahn E C 2020 npj 2D Mater. Appl. 4 17
- [7] Sierra J F, Fabian J, Kawakami R K, Roche S and Valenzuela S O 2021 Nat. Nanotech. 16 856–868
- [8] Zhang L, Zhou J, Li H, Shen L and Feng Y P 2021 Appl. Phys. Rev. 8 021308
- [9] Ghising P, Biswas C and Lee Y H 2023 Adv. Mater. 35 2209137
- [10] Novoselov K S, Mishchenko A, Carvalho A and Castro Neto A H 2016 Science 353 aac9439
- [11] Liu Y, Weiss N O, Duan X, Cheng H C, Huang Y and Duan X 2016 Nat. Rev. Mater. 1 16042
- [12] Liu Y, Huang Y and Duan X 2019 Nature 567 323–333
- [13] Zhong D, Seyler K L, Linpeng X, Cheng R, Sivadas N, Huang B, Schmidgall E, Taniguchi T, Watanabe K, McGuire M A, Yao W, Xiao D, Fu K M C and Xu X 2017 Sci. Adv. 3 e1603113
- [14] Zhong D, Seyler K L, Linpeng X, Wilson N P, Taniguchi T, Watanabe K, McGuire M A, Fu K M C, Xiao D, Yao W and Xu X 2020 Nat. Nanotech. 15 187–191
- [15] Zollner K, Faria Junior P E and Fabian J 2019 Phys. Rev. B 100 085128
- [16] Li L, Jiang S, Wang Z, Watanabe K, Taniguchi T, Shan J and Mak K F 2020 Phys. Rev. Mater. 4 104005
- [17] Lyons T P, Gillard D, Molina-Sánchez A, Misra A, Withers F, Keatley P S, Kozikov A, Taniguchi T, Watanabe K, Novoselov K S, Fernández-Rossier J and Tartakovskii A I 2020 Nat. Commun. 11 6021
- [18] Kim S J, Choi D, Kim K W, Lee K Y, Kim D H, Hong S, Suh J, Lee C, Kim S K, Park T E and Koo H C 2021 ACS Nano 15 16395–16403
- [19] Tu Z, Zhou T, Ersevim T, Arachchige H S, Hanbicki A T, Friedman A L, Mandrus D, Ouyang M, Žutić I and Gong C 2022 Appl. Phys. Lett. 120 043102
- [20] Gong C, Zhang P, Norden T, Li Q, Guo Z, Chaturvedi A, Najafi A, Lan S, Liu X, Wang Y, Gong S J, Zeng H, Zhang H, Petrou A and Zhang X 2023 Nat. Commun. 14 3839
- [21] Choi E M, Kim T, Cho B W and Lee Y H 2023 ACS Nano 17 15656–15665
- [22] Huang B, Clark G, Navarro-Moratalla E, Klein D R, Cheng R, Seyler K L, Zhong D, Schmidgall E, McGuire M A, Cobden D H, Yao W, Xiao D, Jarillo-Herrero P and Xu X 2017 Nature 546 270–273
- [23] Fei Z, Huang B, Malinowski P, Wang W, Song T, Sanchez J, Yao W, Xiao D, Zhu X, May A F, Wu W, Cobden D H, Chu J H and Xu X 2018 Nat. Mater. 17 778–782
- [24] Wang H, Liu Y, Wu P, Hou W, Jiang Y, Li X, Pandey C, Chen D, Yang Q, Wang H, Wei D, Lei N, Kang W, Wen L, Nie T, Zhao W and Wang K L 2020 ACS Nano 14 10045–10053
- [25] Katmis F, Lauter V, Nogueira F S, Assaf B A, Jamer M E, Wei P, Satpati B, Freeland J W, Eremin I, Heiman D, Jarillo-Herrero P and Moodera J S 2016 Nature 533 513–516
- [26] Zhu W, Song C, Han L, Bai H, Wang Q, Yin S, Huang L, Chen T and Pan F 2022 Adv. Funct. Mater. 32 2108953
- [27] Dong X J, You J Y, Zhang Z, Gu B and Su G 2020 Phys. Rev. B 102 144443
- [28] Le D, Barinov A, Preciado E, Isarraraz M, Tanabe I, Komesu T, Troha C, Bartels L, Rahman T S and Dowben P A 2015 J. Phys.: Condens. Matter 27 182201
- [29] Castellanos-Gomez A, Buscema M, Molenaar R, Singh V, Janssen L, van der Zant H S J and Steele G A 2014 2D Mater. 1 011002
- [30] Zhao W, Ghorannevis Z, Chu L, Toh M, Kloc C, Tan P H and Eda G 2013 ACS Nano 7 791–797
- [31] Zhuang H L, Kent P R C and Hennig R G 2016 Phys. Rev. B 93 134407
- [32] Wang Y, Wang C, Liang S J, Ma Z, Xu K, Liu X, Zhang L, Admasu A S, Cheong S W, Wang L, Chen M, Liu Z, Cheng B, Ji W and Miao F 2020 Adv. Mater. 32 2004533
- [33] Chen P, Huang Z, Li C, Zhang B, Bao N, Yang P, Yu X, Zeng S, Tang C, Wu X, Chen J, Ding J, Pennycook S J, Ariando A, Venkatesan T V and Chow G M 2018 Adv. Funct. Mater. 28 1801766
- [34] Wu X, Lan D, Hwang I, Sun C, Zhou H, Yu X, Yang P, Yu X, Liu C, Chen P, Ding J, Chen J and Chow G M 2023 J. Alloy. Compd. 932 167582