pubs.acs.org/NanoLett Letter Artificial Graphene Spin Polarized Electrode for Magnetic Tunnel Junctions Victor Zatko,* Regina Galceran, Marta Galbiati, Julian Peiro, Florian Godel, Lisa-Marie Kern, David Perconte, Fatima Ibrahim, Ali Hallal, Mairbek Chshiev, Benjamin Martinez, Carlos Frontera, Lluìs Balcells, Piran R. Kidambi, John Robertson, Stephan Hofmann, Sophie Collin, Frédéric Petroff, Marie-Blandine Martin, Bruno Dlubak,* and Pierre Seneor* Cite This: Nano Lett. 2023, 23, 34−41 Read Online See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. ACCESS Metrics & More Article Recommendations Downloaded via 158.109.58.130 on October 11, 2023 at 11:11:01 (UTC). ABSTRACT: 2D materials offer the ability to expose their electronic structure to manipulations by a proximity effect. This could be harnessed to craft properties of 2D interfaces and van der Waals heterostructures in devices and quantum materials. We explore the possibility to create an artificial spin polarized electrode from graphene through proximity interaction with a ferromagnetic insulator to be used in a magnetic tunnel junction (MTJ). Ferromagnetic insulator/graphene artificial electrodes were fabricated and integrated in MTJs based on spin analyzers. Evidence of the emergence of spin polarization in proximitized graphene layers was observed through the occurrence of tunnel magnetoresistance. We deduced a spin dependent splitting of graphene’s Dirac band structure (∼15 meV) induced by the proximity effect, potentially leading to full spin polarization and opening the way to gating. The extracted spin signals illustrate the potential of 2D quantum materials based on proximity effects to craft spintronics functionalities, from vertical MTJs memory cells to logic circuits. KEYWORDS: proximity effects, 2D materials, graphene, spin polarization I n recent years, spintronics has seen a rapid development of its application in data storage circuits, such as Magnetic Random Access Memories (MRAMs), targeting low energy bidimensional materials as a spin polarized electrode could provide sharp interfaces with a high tunability thanks to induced proximity effects and gating, leading to unprecedented cost and nonvolatility.1,2 In this regard, spintronics already has control over carrier properties such as density and spin emerged as a promising candidate for low power and more polarization. Pioneering studies started to use intrinsic 2D efficient electronics, as the spin degree of freedom naturally ferromagnets as spin polarized electrodes.16 Interestingly, unifies memory functions with classical logic (in-memory among 2D materials, graphene shows an impressive suitability logics), and even further with beyond CMOS technologies for spintronics circuits with demonstrated efficient spin such as stochastic, neuromorphic, and quantum architec- transport and spin filtering leading to high TMR ratios.11,17,18 tures.3−7 This potential of spintronics is sustaining the quest Furthermore, it exhibits low electrical resistivity and high for novel efficient and versatile material spin platforms. As stability to an ambient atmosphere. While graphene’s band such, introducing 2D materials in spintronics structures is structure is not intrinsically spin dependent, magnetic foreseen to increase control over spin properties and extracted proximity effects are expected to further provide an efficient spin signals:8,9 graphene, h-BN, and 2D semiconductors such control for band structure engineering as discussed in the case as transition metal dichalcogenides have shown their ability for of lateral spin transport devices.19−26 As a result of proximity spin device tailoring (in terms of atomic thickness control, effects, spin polarized graphene thus appears promising to diffusion barrier in heterostructures, spin filtering, enhance- ment of magnetic anisotropy, spin orbit-toque effect...) while embedded in spin valves.10−15 Received: August 5, 2022 In this direction, the use of 2D materials as a replacement for Revised: November 24, 2022 conventional spin sources could lead to improved control over Published: December 19, 2022 device design. The basic building block of spintronics, the magnetic tunnel junction (MTJ), is made of two ferromagnetic spin sources separated by a thin insulating tunnel layer. Using © 2022 American Chemical Society https://doi.org/10.1021/acs.nanolett.2c03113 34 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter implement a 2D spin polarized conductive electrode in spin the role of an in-plane spin polarized electrode in this spin valve structures. These electrodes could then be used in valve architecture, with spin polarization properties closely vertical MTJ memory cells, and in lateral spin transport as well related to LCoMO’s magnetic behavior. The origin of the as logic circuits, paving the road toward an efficient universal obtained TMR spin signals is discussed, extracting the energy material spin platform. Although theoretically predicted for splitting in the graphene band structure due to exchange graphene/magnetic insulator lateral devices, successful extrac- interaction with the FMI (Figure 4). We finally discuss the tion of spins from proximitized graphene remains to be performance of these systems, highlighting conditions leading experimentally demonstrated in a vertical MTJ device to almost full spin polarization, hence toward very high TMR configuration.27 spin signals beyond the current technological state of the art In this Letter, we show that when graphene is at the vicinity (Figure 5). These results illustrate the potential of the of an insulating ferromagnetic layer, its intrinsic properties can manipulation by proximity effects of spin dependent properties be fundamentally changed into a spin polarized electrode in graphene, toward an external control of the spin transport in through hybridization. Graphene can then be used as a spin graphene-based devices including both vertical MTJ and spin source, here demonstrated in a reference MTJ. The study logic circuits. presented here is based on a graphene layer grown by chemical Figure 1 presents the device structure: a 30 nm film of vapor deposition (CVD) transferred on top of the La2Co1−xMn1+xO6 (x ∼ 0.23) is sputtered on top of a LSAT ferromagnetic insulator (FMI) LCoMO (La2Co1−xMn1+xO6 (Lanthanum aluminate−Strontium Aluminum Tantalate) (x ∼ 0.23)). A graphene/LCoMO electrode is then integrated substrate as in ref 28. The strained state of the FMI results into spin valve structures to probe the spin dependent in an in-plane ferromagnetic insulator film. An O2 plasma properties induced within the graphene layer by LCoMO cleaning step is performed on the LCoMO surface to ensure an (Figure 1). Magnetotransport measurements on such devices optimal oxygen content of the film before transferring the graphene layer on top.29 The graphene layer is grown by CVD on Cu using a C6H6 precursor and then lifted and transferred onto a LCoMO substrate by standard transfer protocols using a poly(methyl methacrylate) (PMMA) carrier film.30−32 In order to electrically isolate different devices within the same sample, individual graphene electrodes are defined by laser lithography and O 2 plasma etching. Another step of lithography is performed to design gold electrodes that are used to connect the graphene to a bias source. In order to define the top spin analyzer structure, a final lithography step is performed on top of the graphene/LCoMO electrode to define Figure 1. Optical image and schematic view of the devices. The image micrometer wide junctions: we grow a tunnel barrier (atomic has been taken with a ×50 magnification objective. A microjunction is patterned on top of a LCoMO/graphene electrode. A top spin layer deposition of 3 nm MgO or 1 nm sputtered Al2O3 analyzer (Co/Oxide) is deposited to probe the spin polarization in depending on samples) and a top 15 nm sputtered Co graphene, leading to a magnetic tunnel junction. A gold contact is ferromagnetic spin analyzer further capped with 80 nm Au.33,34 used to connect graphene. This fabrication process results in the definition of [LCoMO/Graphene]/oxide barrier/Co junctions, allowing us to probe proximity effects induced in graphene by LCoMO show a magnetoresistive behavior that arises from the thanks to reference tunnel spin analyzers. A tunnel behavior is proximity effects between graphene and the FMI (Figure 2). systematically measured on the junctions presented in this Furthermore, a hysteretic spin transport signal is observed in work by usual I(V) and dI/dV(V) measurements.35,36,13 several devices (Figure 3), demonstrating that graphene plays Graphene, on regular oxides, uncoupled to the environment, Figure 2. Magnetotransport measurements and simulations of the devices [LCoMO/Graphene]/MgO/Co. Inset: schematic of the device. (a) Experimental plot of the magnetoresistance (MR) as a function of in-plane external magnetic field at 2 K. The blue curve represents the measurement while sweeping from 2 T to −2 T, while the red curve is obtained when sweeping from −2 T to 2 T. We observe a clear negative MR and a hysteretic behavior below 1 T. (b) Raw SQUID data of the ferromagnetic insulator. Top inset shows SQUID data on a wider range (up to 5.5 T). Bottom inset shows simulations of the magnetoresistance of the devices. 35 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter Figure 3. ΔR, the spin induced variation of the junction’s resistance at 2 K, as a function of external in-plane magnetic field after subtraction of the general baseline. One can see a clear TMR signal: two distinct states of resistances are observed. This can be achieved only with 2 spin polarized electrodes, showing the spin polarization induced in the proximitized graphene. Devices A, B, and C are made with a MgO barrier while device D was made with an alumina barrier. is not expected to provide a spin polarized current. Here, on coupling, leading to vanishing SHE. Next, in a true 2D the contrary, by the proximity with the ferromagnetic insulator material, such as our single graphene layer, SHE would not substrate, we engineer a modification of graphene’s band create in-plane spins but would lead only to out-of-plane spins structure, lifting the degeneracy of the spin population accumulating at the edges of graphene. Those out-of-plane densities. Experimentally, if the proximity effect is strong spins would not be impacted by the reversal of our FMI’s in- enough, it is expected to lead to the observation of a hysteretic plane magnetization. Moreover, from the experimental point of behavior of the magnetoresistive signal corresponding to the view, the effects reported here are at least 1 order of magnitude classic response of a MTJ highlighting that the graphene layer higher than the highest resistance variation that could be has become spin polarized and acts as a spin source. expected from such SHE structures. In Figure 2a we show typical in-plane magnetotransport To further correlate the magnetotransport curve’s general measurements at 2 K for the devices with a MgO barrier. The shape with proximity induced modification of graphene resistance vs in-plane magnetic field curves of the junctions properties and model the magnetic field dependence, we display a negative magnetoresistance (MR). This effect is not show the raw in-plane SQUID measurement of a LCoMO film expected for an isolated graphene layer. Moreover, the current of the devices (Figure 2b). Above its coercive field, the is applied between the graphene and the cobalt electrode. magnetization of LCoMO smoothly aligns along the direction Importantly, it does not flow in or through the insulating of the external magnetic field without being fully saturated LCoMO layer.37 The insulating character of LCoMO was even up to 5 T (the contribution of the LSAT substrate being confirmed through DC transport measurement. Devices diamagnetic, it cannot account for the increased magnetization without the graphene electrode were found to be isolated with the externally applied magnetic field). As the spin with unmeasurable resistances (beyond the limit of our setup). polarization in graphene depends on the exchange interaction Overall, the individual contributions of the isolated layers of with the FMI which is proportional to its magnetization, the the devices cannot explain the general trend of the magneto- SQUID data associated with Jullière’s model allow us to transport, suggesting proximity effects between graphene and correlate the MR of our system with the application of an the FMI leading to the evolution of the graphene band external in-plane magnetic field (more details are given below). structure. We note that our observed measurements can be Strikingly we are able to reproduce the overall evolution of the ruled out as originating from spin-Hall effect (SHE) induced resistances with respect to an applied magnetic field as shown MR such as the spin-Hall magnetoresistance (SMR) effects in the bottom inset of Figure 2b. The magnetotransport signal arising at heavy metal/FMI interfaces.38 Indeed, first graphene of the spin valves depends on the relative orientation of the is acknowledged as a material with very low spin−orbit magnetization of the two ferromagnets. As in the devices 36 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter Figure 4. Model of the spin polarization induced in graphene through proximity effects with LCoMO. (a) Schematic of the splitting of the Dirac cones in graphene due to exchange with a ferromagnetic insulator (FMI). (b) Schematic representation of the splitting induced in graphene by proximity with an FMI and an oxide barrier in a MTJ. (c) Lower panel: Schematics of the resulting MTJ where the graphene plays the role of one of the spin polarized electrodes. Upper panel: TMR and spin polarization vs gate voltage induced by exchange interaction in a graphene in proximity with a ferromagnet for different values of exchange interaction. The peaks correspond to the extreme cases leading to an almost full spin polarization. If an additional small gap is opened in graphene at the Dirac point, the effect is even more dramatic. Figure 5. Electronic structure of the graphene layer at the Gr/LCOMO interface computed from first-principles. (a) Balls and sticks representation of the interface. Carbon, lanthane, cobalt, manganese, and oxygen atoms are depicted as brown, green, blue, violet, and red balls, respectively. (b) Spin splitted (black and red) electronic band structure of the graphene (solid circles) on top of LCoMO. studied here, the top Co spin analyzer is fully aligned along the graphene. The general behavior of the devices pinpoints a external in-plane magnetic field, the smooth alignment of the strong modification of graphene’s intrinsic properties induced LCoMO along that same direction induces a smooth variation by the proximity with the FMI. of the total resistance of the device (sensor-like behavior). In a Besides the general magnetoresistive trend of the R(H) nutshell, the observed MR is governed by the relative curve (Figure 2a), a hysteretic behavior is also observed alignment of the top and bottom electrode magnetizations. between 1 T and −1 T. We highlight these features by The bottom artificial proximitized graphene electrode being subtracting the general nonhysteretic magnetoresistance from induced by the LCoMO replicates its magnetization. As such, the signal previously identified. The resulting hysteretic TMR the observed inversed “parabolic-like” shape of the MR can be signals for four different devices (devices A, B, C with MgO directly related to the LCoMO magnetization behavior barrier and device D with alumina barrier) are presented in observed in the presented additional SQUID magnetic Figure 3. One can see that the extracted TMRs closely characterizations (Figure 2b, top inset). The peculiar magnetic resemble the usual spin signals extracted from an MTJ with behavior of the insulting LCoMO layer has an effect on the two metallic ferromagnets acting as spin polarized electrodes. spin transport of the full device through a proximity effect with However, here only one electrode is a conventional metallic 37 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter ferromagnet (Co), with the other electrode being graphene. As ization is directly linked to the exchange splitting by such, the otherwise nonmagnetic graphene appears to behave E E PGrProx(E) = 2E = 2 v n . To give a first estimation of ΔE, as a spin polarized electrode thanks to its interaction with the F F underlying FMI (LCoMO). The influence of the insulating we fabricated on the same chip a Hall bar device which allowed LCoMO is further confirmed by the value of magnetic fields at us to extract a carrier density of n = −4 × 1012 cm−2 for the which the hysteresis switches are observed in the TMR. These graphene layer. Measurements of MTJ configurations based on values match the LCoMO coercive field (Figure 2b). reference spin analyzers Co/oxide described above highlight The measured spin signal thus appears to arise from an PGrProx(E) ≥ 0.8%. From this value, we can thus already infer extrinsic spin polarization induced within the graphene an exchange splitting in graphene ΔE of at least 4 meV. electrode by the proximity effect with the FMI. We can However, following the SQUID data, we pinpoint that the estimate this graphene spin polarization induced in graphene observed induced exchange is related to the remanent using Jullière’s model and our Co-based spin analyzer stack. magnetization of the FMI at its switching field of 0.8 T We have the tunneling magnetoresistance defined as which is only 23% of its value at 5 T (see top inset of Figure 2PanalyzerPGrProx 2b). Hence, at 5 T (almost saturated LCoMO), the expected TMR = 1 PanalyzerPGrProx with PGrProx and Panalyzer indicating the induced exchange splitting is 15 meV. This qualitative spin polarization of the graphene in proximity with LCoMO approach is confirmed by first-principle calculations of the and the spin polarization of the Co-based spin analyzer, LCoMO/Gr interface where one finds the emergence of an respectively. In an attempt to derive a lower bound for the induced exchange splitting�above 15 meV�in graphene induced characteristics of the graphene, we extract from associated with the presence of LCoMO. In this context, it previous works that Panalyzer ≤ 32% and from the measured clearly appears that a finely gated control of EF might lead to a TMR (∼0.48%) we can deduce the spin polarization of the large PGrProx for any given sizable splitting ΔE. This LCoMO-proximitized graphene |PGrProx| ≥ 0.8%.39,40 We note opportunity is illustrated in Figure 4c. In the extreme case that, in the case of sample D, the TMR signals were about 10 where the Fermi level is aligned with the Dirac point of one times smaller than for the other samples. This sample has been spin direction through gating, the electronic density vanishes made with a sputtered 1 nm alumina tunnel barrier known to completely comparatively to the other spin direction; this leads impact the graphene underneath whereas the others were to an artificial full spin polarization of the graphene. Here we made with a 3 nm ALD-deposited MgO tunnel barrier. The note that while eventually the charge fluctuation Δn in difference between both type of samples is consistent with the graphene should be taken into account, this should not hamper noninvasive higher quality of ALD grown barriers as noted in the ability to achieve large spin polarization. Indeed, it was shown that charge fluctuations in the <10 meV range could be previous reports.35,41 This observation is also consistent with achieved even without requiring highly efficient h-BN the fact that the spin signal comes from the graphene passivation.42 In this case, a splitting value of approximately embedded in the MTJ as we see the same general 50 meV would already lead to more than 60% spin magnetotransport behavior with both tunnel barriers. polarization, while an exchange splitting value of ∼150 meV We now discuss the origin of the TMR signals observed in would lead to ∼85% of spin polarization. It is worth noting our devices and the related induced graphene spin polarization. that 60% of spin polarization corresponds to a TMR above The proximity with the ferromagnetic insulator LCoMO 100%, while a spin polarization of 85% corresponds to TMR induces a spin dependent modification of the graphene band above 500%, the range usually observed with MgO tunnel structure. This can be described in a first approximation by a barriers. Outstandingly, an even stronger effect could be spin dependent splitting ΔE of the Dirac band structure of achieved thanks to a proximity-effect induced gap in the graphene due to the proximity interaction with the FMI (as graphene band structure at the Dirac point.43 Indeed, such an depicted in Figure 4a). While more complex in reality, a simple induced gap would strongly suppress one of the spin conceptual picture of this effect can be given by considering directions, relaxing the impact of Δn, and leading to spin the graphene as an atom-thick quantum well (Figure 4b). polarizations even closer to the ideal PGrProx = 100% case. It is When this atom-thin 2D quantum well is put into proximity important here to note that using a single layer, in addition to with a FMI, its levels’ degeneracy is lifted for the spin up and maximizing the impact of the proximity effect of the FMI on down directions. This is due to different barrier heights for the graphene, supports ruling out other thickness dependent two spin directions in the FMI surrounding the quantum well effects. (i.e., different evanescent waves in the FMI barrier). This leads These results illustrate the striking potential of the proximity to the exchange field of the FMI to be imprinted onto the 2D effects to tailor the graphene toward modulated spintronics quantum well. A picture of our case, where we assume this properties for MTJs and beyond for spin logics. The spin effect to be homogeneous over the one atom-thick graphene polarization could indeed be controlled through graphene height, is shown in Figure 4b. We can deduce the value of this gating leading to controllable spin polarizer/analyzer and spin splitting using the graphene spin polarization estimated from torque properties. We think that this artificial FM proximitized magnetotransport measurements. Indeed, the spin polarization graphene offers an interesting complement to the families of D (E F) D (E F) can be defined as PGrProx(E F) = , with Dσ(EF) 2D ferromagnets (2D FMs) being now uncovered as spin D (E F) + D (E F) sources.44 Indeed, 2D FMs have been recently highlighted as being the graphene density of states for the spin direction σ(↑, offering very interesting potential compared to usual 3D E , E ↓). Here we have D (E F) = 2 where E = EF ± 2 systems, a key one being the gating of their FM proper- ( vF) ties.45−49 As shown in Figure 4, our artificial system could also with E F = vF n being the Fermi energy, n the carrier advantageously offer this gating potential, while, in contrast to density, and VF the Fermi velocity. From this, we deduce 2D FMs, being easier to integrate. Indeed, 2D FM families still E E PGrProx(E F) = and find that the graphene spin polar- remain to be stabilized and developed (see for instance ref 50), E +E 38 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter whereas graphene brings ambient stability and eased large scale Authors fabrication. This is a clear asset for implementation that could Regina Galceran − Unité Mixte de Physique, CNRS, Thales, furthermore benefit from the bright perspective to extract Université Paris-Saclay, 91767 Palaiseau, France; CSIC and strong tunable spin polarizations from graphene. We note BIST, Campus UAB, Catalan Institute of Nanoscience and however that these assets are balanced by the fact that while Nanotechnology (ICN2), Bellaterra 08193 Barcelona, Spain numerous FMI are available, our approach requires engineer- Marta Galbiati − Unité Mixte de Physique, CNRS, Thales, ing the right heterostructure while combining them with Université Paris-Saclay, 91767 Palaiseau, France graphene. Still, the strong spin polarization dependence of such Julian Peiro − Unité Mixte de Physique, CNRS, Thales, a system anticipated by our work suggests the implementation Université Paris-Saclay, 91767 Palaiseau, France of a gate to take maximum benefit from it. Once the Florian Godel − Unité Mixte de Physique, CNRS, Thales, Université Paris-Saclay, 91767 Palaiseau, France; technological progress will allow such a geometry, a different orcid.org/0000-0003-1741-2741 path can be followed to reach the most pertinent system51,7 Lisa-Marie Kern − Unité Mixte de Physique, CNRS, Thales, such as direct backgating by adding a conductive layer below Université Paris-Saclay, 91767 Palaiseau, France; our graphene/FMI or by using both as a vertical and lateral orcid.org/0000-0001-9781-9132 spin source and exploiting the long spin lifetimes during David Perconte − Unité Mixte de Physique, CNRS, Thales, transport in graphene to reach a second electrode. This would Université Paris-Saclay, 91767 Palaiseau, France for example allow researchers to define a top gate structure on Fatima Ibrahim − Univ. Grenoble Alpes, CEA, CNRS, Spintec, top of the graphene/FMI injection point and modulate the 38000 Grenoble, France spin polarization as discussed. This would however be Ali Hallal − Univ. Grenoble Alpes, CEA, CNRS, Spintec, particularly demanding (spin transport is subtly dependent 38000 Grenoble, France on device geometry and resistance matching, see ref 52), and Mairbek Chshiev − Univ. Grenoble Alpes, CEA, CNRS, we believe this will require a collective community-wide Spintec, 38000 Grenoble, France; Institut Universitaire de exploration. France, 75231 Paris, France; orcid.org/0000-0001-9232- In conclusion, we managed to convert graphene into a spin 7622 polarized electrode by proximity effects induced by an Benjamin Martinez − Institut de Ciencia de Materials de insulating ferromagnet layer and successfully implement it in Barcelona, ICMAB-CSIC, Campus UAB, 08193 Bellaterra, functional MTJ structures. By analyzing the resulting sizable Spain spin signals in several devices, we extract the energy splitting Carlos Frontera − Institut de Ciencia de Materials de induced by FMI in the graphene band structure (∼15 meV) Barcelona, ICMAB-CSIC, Campus UAB, 08193 Bellaterra, leading to the appearance of a spin polarization in graphene. Spain; orcid.org/0000-0002-0091-4756 Lluìs Balcells − Institut de Ciencia de Materials de Barcelona, This work expands the role that can be attributed to graphene ICMAB-CSIC, Campus UAB, 08193 Bellaterra, Spain in spintronics devices: it is demonstrated as a possible spin Piran R. Kidambi − Department of Chemical and source in this proximitized configuration. Based on exper- Biomolecular Engineering, Vanderbilt University, Nashville, imentally accessible materials and conditions with low charge Tennessee 37212, United States; orcid.org/0000-0003- fluctuations below 10 meV together with predictions of 1546-5014 exchange splitting in the 100 meV range and gaps above 30 John Robertson − Department of Engineering, University of meV, spin polarizations above 75% are now foreseen in these Cambridge, Cambridge CB3 0FA, United Kingdom systems.43,53 This work opens broader perspectives toward Stephan Hofmann − Department of Engineering, University of fully 2D-based MTJs and van der Waals heterostructures that Cambridge, Cambridge CB3 0FA, United Kingdom; aim at exploiting the unique properties of 2D materials in spin orcid.org/0000-0001-6375-1459 transport. The spin polarization induced in graphene appears Sophie Collin − Unité Mixte de Physique, CNRS, Thales, as an essential step toward all-spin logic circuits: bypassing the Université Paris-Saclay, 91767 Palaiseau, France ferromagnet-channel impedance mismatch issue by selectively Frédéric Petroff − Unité Mixte de Physique, CNRS, Thales, tailoring the same material either as a spin source or an Université Paris-Saclay, 91767 Palaiseau, France efficient spin transport channel might finally lead to the well- Marie-Blandine Martin − Unité Mixte de Physique, CNRS, looked after spin logic platform.54,52 Thales, Université Paris-Saclay, 91767 Palaiseau, France Complete contact information is available at: ■ AUTHOR INFORMATION Corresponding Authors https://pubs.acs.org/10.1021/acs.nanolett.2c03113 Notes Victor Zatko − Unité Mixte de Physique, CNRS, Thales, The authors declare no competing financial interest. Université Paris-Saclay, 91767 Palaiseau, France; orcid.org/0000-0002-2475-8866; Email: victor.zatko@ cnrs-thales.fr Bruno Dlubak − Unité Mixte de Physique, CNRS, Thales, ■ ACKNOWLEDGMENTS This work has received funding from the European Union’s Université Paris-Saclay, 91767 Palaiseau, France; H2020 Future and Emerging Technologies “Graphene Flag- orcid.org/0000-0001-5696-8991; ship” (Grant Core3 No. 881603), “SINFONIA” (No. 881603) Email: bruno.dlubak@cnrs-thales.fr projects and a Grant Agreement Marie Skłodowska Curie Pierre Seneor − Unité Mixte de Physique, CNRS, Thales, “ITN Spinograph” (No. 697904), as well as from EPSRC (EP/ Université Paris-Saclay, 91767 Palaiseau, France; P005152/1, EP/K016636/1). This research is supported by a Email: pierre.seneor@cnrs-thales.fr public grant overseen by the French National Research Agency 39 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter (ANR) as part of the “Investissements d’Avenir” program (14) Dankert, A.; Pashaei, P.; Kamalakar, M. V.; Gaur, A. P. S.; Labex NanoSaclay (ANR-10-LABX-0035), as well as grants Sahoo, S.; Rungger, I.; Narayan, A.; Dolui, K.; Hoque, M. A.; Patel, R. STEM2D (ANR-19-CE24-0015), MIXES (ANR-19-CE09- S.; de Jong, M. P.; Katiyar, R. S.; Sanvito, S.; Dash, S. P. Spin- 0028), and by the Flag-ERA JTC 2019 project “SOGraph- Polarized Tunneling through Chemical Vapor Deposited Multilayer Molybdenum Disulfide. ACS Nano 2017, 11, 6389. MEM” (ANR-19-GRFI-0001-07). This work has also received (15) Zatko, V.; Galbiati, M.; Dubois, S. M.-M.; Och, M.; Palczynski, fundings from the “Spanish ministry of Science and P.; Mattevi, C.; Brus, P.; Bezencenet, O.; Martin, M.-B.; Servet, B.; Innovation” through “Severo Ochoa” (CEX2019-000917-S) Charlier, J.-C.; Godel, F.; Vecchiola, A.; Bouzehouane, K.; Collin, S.; and “OXISOT” (PID2021-128410OB-I00). Petroff, F.; Dlubak, B.; Seneor, P. Band-Structure Spin-Filtering in Vertical Spin Valves Based on Chemical Vapor Deposited WS 2. ACS ■ REFERENCES (1) Khvalkovskiy, A. V.; Apalkov, D.; Watts, S.; Chepulskii, R.; Nano 2019, 13, 14468. (16) Wang, Z.; Gutiérrez-Lezama, I.; Ubrig, N.; Kroner, M.; Gibertini, M.; Taniguchi, T.; Watanabe, K.; Imamoǧlu, A.; Giannini, Beach, R. S.; Ong, A.; Tang, X.; Driskill-Smith, A.; Butler, W. H.; E.; Morpurgo, A. F. Very Large Tunneling Magnetoresistance in Visscher, P. B.; Lottis, D.; Chen, E.; Nikitin, V.; Krounbi, M. Basic Layered Magnetic Semiconductor CrI3. Nat. Commun. 2018, 9, 2516. Principles of STT-MRAM Cell Operation in Memory Arrays. J. Phys. (17) Karpan, V. M.; Giovannetti, G.; Khomyakov, P. A.; Talanana, D. Appl. Phys. 2013, 46, 074001. M.; Starikov, A. A.; Zwierzycki, M.; van den Brink, J.; Brocks, G.; (2) Dieny, B.; Prejbeanu, I. L.; Garello, K.; Gambardella, P.; Freitas, Kelly, P. J. Graphite and Graphene as Perfect Spin Filters. Phys. Rev. P.; Lehndorff, R.; Raberg, W.; Ebels, U.; Demokritov, S. O.; Akerman, Lett. 2007, 99, 176602. J.; Deac, A.; Pirro, P.; Adelmann, C.; Anane, A.; Chumak, A. V.; (18) Dlubak, B.; Martin, M.-B.; Deranlot, C.; Servet, B.; Xavier, S.; Hirohata, A.; Mangin, S.; Valenzuela, S. O.; Onbaslı, M. C.; d’Aquino, Mattana, R.; Sprinkle, M.; Berger, C.; De Heer, W. A.; Petroff, F.; M.; Prenat, G.; Finocchio, G.; Lopez-Diaz, L.; Chantrell, R.; Anane, A.; Seneor, P.; Fert, A. Highly Efficient Spin Transport in Chubykalo-Fesenko, O.; Bortolotti, P. Opportunities and Challenges Epitaxial Graphene on SiC. Nat. Phys. 2012, 8, 557. for Spintronics in the Microelectronic Industry. Nature Electronics (19) Leutenantsmeyer, J. C.; Kaverzin, A. A; Wojtaszek, M.; van 2020, 3, 446. Wees, B. J Proximity Induced Room Temperature Ferromagnetism in (3) Behin-Aein, B.; Datta, D.; Salahuddin, S.; Datta, S. Proposal for Graphene Probed with Spin Currents. 2D Mater. 2017, 4, 014001. an All-Spin Logic Device with Built-in Memory. Nat. Nanotechnol. (20) Wang, Z.; Tang, C.; Sachs, R.; Barlas, Y.; Shi, J. Proximity- 2010, 5, 266. Induced Ferromagnetism in Graphene Revealed by the Anomalous (4) Manipatruni, S.; Nikonov, D. E.; Lin, C. C.; Gosavi, T. A.; Liu, Hall Effect. Phys. Rev. Lett. 2015, 114, 016603. H.; Prasad, B.; Huang, Y. L.; Bonturim, E.; Ramesh, R.; Young, I. A. (21) Wu, Y. F.; Song, H. D.; Zhang, L.; Yang, X.; Ren, Z.; Liu, D.; Scalable Energy-Efficient Magnetoelectric Spin−Orbit Logic. Nature Wu, H. C.; Wu, J.; Li, J. G.; Jia, Z.; Yan, B.; Wu, X.; Duan, C. G.; Han, 2019, 565, 35. G.; Liao, Z. M.; Yu, D. Magnetic Proximity Effect in Graphene (5) Borders, W. A.; Pervaiz, A. Z.; Fukami, S.; Camsari, K. Y.; Ohno, Coupled to a BiFe O3 Nanoplate. Phys. Rev. B 2017, 95, 195426. H.; Datta, S. Integer Factorization Using Stochastic Magnetic Tunnel (22) Wei, P.; Lee, S.; Lemaitre, F.; Pinel, L.; Cutaia, D.; Cha, W.; Junctions. Nature 2019, 573, 390. Katmis, F.; Zhu, Y.; Heiman, D.; Hone, J.; Moodera, J. S.; Chen, C. T. (6) Grollier, J.; Querlioz, D.; Camsari, K. Y.; Everschor-Sitte, K.; Strong Interfacial Exchange Field in the Graphene/EuS Hetero- Fukami, S.; Stiles, M. D. Neuromorphic Spintronics. Nature Electronics structure. Nat. Mater. 2016, 15, 711. 2020, 3, 360. (23) Tang, C.; Zhang, Z.; Lai, S.; Tan, Q.; Gao, W. Magnetic (7) Khokhriakov, D.; Sayed, S.; Hoque, A. M.; Karpiak, B.; Zhao, B.; Proximity Effect in Graphene/CrBr 3 van Der Waals Heterostructures. Datta, S.; Dash, S. P.Multifunctional Spin Logic Gates In Graphene Adv. Mater. 2020, 32, 1908498. Spin Circuits, 2021, 2108.12259, arXiv,https://arxiv.org/abs/2108. (24) Ghiasi, T. S.; Kaverzin, A. A.; Dismukes, A. H.; de Wal, D. K.; 12259 (accessed November 18th, 2022). Roy, X.; van Wees, B. J. Electrical and Thermal Generation of Spin (8) Scharf, B.; Xu, G.; Matos-Abiague, A.; Ž utić, I. Magnetic Currents by Magnetic Bilayer Graphene. Nat. Nanotechnol. 2021 167 Proximity Effects in Transition-Metal Dichalcogenides: Converting 2021, 16, 788. Excitons. Phys. Rev. Lett. 2017, 119, 127403. (25) Karpiak, B.; Cummings, A. W.; Zollner, K.; Vila, M.; (9) Piquemal-Banci, M.; Galceran, R.; Martin, M. B.; Godel, F.; Khokhriakov, D.; Hoque, A. M.; Dankert, A.; Svedlindh, P.; Fabian, Anane, A.; Petroff, F.; Dlubak, B.; Seneor, P. 2D-MTJs: Introducing J.; Roche, S.; Dash, S. P. Magnetic Proximity in a van Der Waals 2D Materials in Magnetic Tunnel Junctions. J. Phys. D. Appl. Phys. Heterostructure of Magnetic Insulator and Graphene. 2D Mater. 2017, 50, 203002. 2020, 7, 015026. (10) Cobas, E. D.; Van ’T Erve, O. M. J.; Cheng, S. F.; Culbertson, J. (26) Kaverzin, A. A; Ghiasi, T. S; Dismukes, A. H; Roy, X.; van C.; Jernigan, G. G.; Bussman, K.; Jonker, B. T. Room-Temperature Wees, B. J Spin Injection by Spin−Charge Coupling in Proximity Spin Filtering in Metallic Ferromagnet-Multilayer Graphene-Ferro- Induced Magnetic Graphene. 2D Mater. 2022, 9, 045003. magnet Junctions. ACS Nano 2016, 10, 10357. (27) Solis, D. A.; Hallal, A.; Waintal, X.; Chshiev, M. Proximity (11) Piquemal-Banci, M.; Galceran, R.; Dubois, S. M. M.; Zatko, V.; Magnetoresistance in Graphene Induced by Magnetic Insulators. Phys. Galbiati, M.; Godel, F.; Martin, M. B.; Weatherup, R. S.; Petroff, F.; Rev. B 2019, 100, 104402. Fert, A.; Charlier, J. C.; Robertson, J.; Hofmann, S.; Dlubak, B.; (28) Galceran, R.; López-Mir, L.; Bozzo, B.; Cisneros-Fernández, J.; Seneor, P. Spin Filtering by Proximity Effects at Hybridized Interfaces Santiso, J.; Balcells, L.; Frontera, C.; Martínez, B. Strain-Induced in Spin-Valves with 2D Graphene Barriers. Nat. Commun. 2020, 11, Perpendicular Magnetic Anisotropy in L A2CoMn O6-É Thin Films 5670. and Its Dependence on Film Thickness. Phys. Rev. B 2016, 93, (12) Dankert, A.; Venkata Kamalakar, M.; Wajid, A.; Patel, R. S.; 144417. Dash, S. P. Tunnel Magnetoresistance with Atomically Thin Two- (29) Galceran, R.; Frontera, C.; Balcells, L.; Cisneros-Fernández, J.; Dimensional Hexagonal Boron Nitride Barriers. Nano Res. 2014 84 López-Mir, L.; Roqueta, J.; Santiso, J.; Bagués, N.; Bozzo, B.; Pomar, 2015, 8, 1357. A.; Sandiumenge, F.; Martínez, B. Engineering the Microstructure and (13) Piquemal-Banci, M.; Galceran, R.; Godel, F.; Caneva, S.; Magnetism of La2CoMnO6-δ Thin Films by Tailoring Oxygen Martin, M. B.; Weatherup, R. S.; Kidambi, P. R.; Bouzehouane, K.; Stoichiometry. Appl. Phys. Lett. 2014, 105, 242401. Xavier, S.; Anane, A.; Petroff, F.; Fert, A.; Dubois, S. M. M.; Charlier, (30) Kidambi, P. R.; Ducati, C.; Dlubak, B.; Gardiner, D.; J. C.; Robertson, J.; Hofmann, S.; Dlubak, B.; Seneor, P. Insulator-to- Weatherup, R. S.; Martin, M. B.; Seneor, P.; Coles, H.; Hofmann, Metallic Spin-Filtering in 2D-Magnetic Tunnel Junctions Based on S. The Parameter Space of Graphene Chemical Vapor Deposition on Hexagonal Boron Nitride. ACS Nano 2018, 12, 4712. Polycrystalline Cu. J. Phys. Chem. C 2012, 116, 22492. 40 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41 Nano Letters pubs.acs.org/NanoLett Letter (31) Dlubak, B.; Kidambi, P. R.; Weatherup, R. S.; Hofmann, S.; Tunable Room-Temperature Ferromagnetism in Two-Dimensional Robertson, J. Substrate-Assisted Nucleation of Ultra-Thin Dielectric Fe3GeTe2. Nat. 2018 5637729 2018, 563, 94. Layers on Graphene by Atomic Layer Deposition. Appl. Phys. Lett. (46) Wang, K.; De Greve, K.; Jauregui, L. A.; Sushko, A.; High, A.; 2012, 100, 173113. Zhou, Y.; Scuri, G.; Taniguchi, T.; Watanabe, K.; Lukin, M. D.; Park, (32) Mzali, S.; Montanaro, A.; Xavier, S.; Servet, B.; Mazellier, J. P.; H.; Kim, P. Electrical Control of Charged Carriers and Excitons in Bezencenet, O.; Legagneux, P.; Piquemal-Banci, M.; Galceran, R.; Atomically Thin Materials. Nat. Nanotechnol. 2018 132 2018, 13, 128. Dlubak, B.; Seneor, P.; Martin, M. B.; Hofmann, S.; Robertson, J.; (47) Gibertini, M.; Koperski, M.; Morpurgo, A. F.; Novoselov, K. S. Cojocaru, C. S.; Centeno, A.; Zurutuza, A. Stabilizing a Graphene Magnetic 2D Materials and Heterostructures. Nat. Nanotechnol. 2019 Platform toward Discrete Components. Appl. Phys. Lett. 2016, 109, 145 2019, 14, 408. 253110. (48) Och, M.; Martin, M. B.; Dlubak, B.; Seneor, P.; Mattevi, C. (33) Kern, L. M.; Galceran, R.; Zatko, V.; Galbiati, M.; Godel, F.; Synthesis of Emerging 2D Layered Magnetic Materials. Nanoscale Perconte, D.; Bouamrane, F.; Gaufrès, E.; Loiseau, A.; Brus, P.; 2021, 13, 2157. Bezencenet, O.; Martin, M. B.; Servet, B.; Petroff, F.; Dlubak, B.; (49) Yang, H.; Valenzuela, S. O.; Chshiev, M.; Couet, S.; Dieny, B.; Seneor, P. Atomic Layer Deposition of a MgO Barrier for a Passivated Dlubak, B.; Fert, A.; Garello, K.; Jamet, M.; Jeong, D. E.; Lee, K.; Lee, Black Phosphorus Spintronics Platform. Appl. Phys. Lett. 2019, 114, T.; Martin, M. B.; Kar, G. S.; Sénéor, P.; Shin, H. J.; Roche, S. Two- 053107. Dimensional Materials Prospects for Non-Volatile Spintronic (34) Dlubak, B.; Martin, M. B.; Deranlot, C.; Bouzehouane, K.; Memories. Nat. 2022 6067915 2022, 606, 663. Fusil, S.; Mattana, R.; Petroff, F.; Anane, A.; Seneor, P.; Fert, A. (50) Galbiati, M.; Zatko, V.; Godel, F.; Hirschauer, P.; Vecchiola, A.; Homogeneous Pinhole Free 1 Nm Al2O3 Tunnel Barriers on Bouzehouane, K.; Collin, S.; Servet, B.; Cantarero, A.; Petroff, F.; Graphene. Appl. Phys. Lett. 2012, 101, 203104. Martin, M. B.; Dlubak, B.; Seneor, P. Very Long Term Stabilization of (35) Dlubak, B.; Martin, M.-B.; Weatherup, R. S.; Yang, H.; a 2D Magnet down to the Monolayer for Device Integration. ACS Deranlot, C.; Blume, R.; Schloegl, R.; Fert, A.; Anane, A.; Hofmann, Appl. Electron. Mater. 2020, 2, 3508. S.; Seneor, P.; Robertson, J. Graphene-Passivated Nickel as an (51) Wen, H.; Dery, H.; Amamou, W.; Zhu, T.; Lin, Z.; Shi, J.; Oxidation-Resistant Electrode for Spintronics. ACS Nano 2012, 6, Ž utić, I.; Krivorotov, I.; Sham, L. J.; Kawakami, R. K. Experimental 10930. Demonstration of Xor Operation in Graphene Magnetologic Gates at (36) Martin, M. B.; Dlubak, B.; Weatherup, R. S.; Piquemal-Banci, Room Temperature. Phys. Rev. Appl. 2016, 5, 044003. M.; Yang, H.; Blume, R.; Schloegl, R.; Collin, S.; Petroff, F.; Hofmann, (52) Seneor, P.; Dlubak, B.; Martin, M. B.; Anane, A.; Jaffres, H.; S.; Robertson, J.; Anane, A.; Fert, A.; Seneor, P. Protecting Nickel Fert, A. Spintronics with Graphene. MRS Bull. 2012, 37, 1245. with Graphene Spin-Filtering Membranes: A Single Layer Is Enough. (53) Holmes, A. M.; Pakniyat, S.; Gangaraj, S. A. H.; Monticone, F.; Appl. Phys. Lett. 2015, 107, 012408. Weinert, M.; Hanson, G. W. Exchange Splitting and Exchange- (37) Lopez-Mir, L.; Frontera, C.; Aramberri, H.; Bouzehouane, K.; Induced Nonreciprocal Photonic Behavior of Graphene in CrI3- Cisneros-Fernandez, J.; Bozzo, B.; Balcells, L.; Martinez, B. Graphene van Der Waals Heterostructures. Phys. Rev. B 2020, 102, Anisotropic Sensor and Memory Device with a Ferromagnetic 075435. Tunnel Barrier as the Only Magnetic Element OPEN. Sci. Rep. (54) Fert, A.; Jaffrès, H. Conditions for Efficient Spin Injection from 2018, 8, 861. a Ferromagnetic Metal into a Semiconductor. Phys. Rev. B - Condens. (38) Nakayama, H.; Althammer, M.; Chen, Y. T.; Uchida, K.; Matter Mater. Phys. 2001, 64, 184420. Kajiwara, Y.; Kikuchi, D.; Ohtani, T.; Geprägs, S.; Opel, M.; Takahashi, S.; Gross, R.; Bauer, G. E. W.; Goennenwein, S. T. B.; Saitoh, E. Spin Hall Magnetoresistance Induced by a Nonequilibrium Proximity Effect. Phys. Rev. Lett. 2013, 110, 206601. (39) Bernand-Mantel, A.; Seneor, P.; Lidgi, N.; Muñoz, M.; Cros, V.; Fusil, S.; Bouzehouane, K.; Deranlot, C.; Vaures, A.; Petroff, F.; Fert, A. Evidence for Spin Injection in a Single Metallic Nanoparticle: A Step towards Nanospintronics. Appl. Phys. Lett. 2006, 89, 062502. (40) Barraud, C.; Deranlot, C.; Seneor, P.; Mattana, R.; Dlubak, B.; Fusil, S.; Bouzehouane, K.; Deneuve, D.; Petroff, F.; Fert, A. Magnetoresistance in Magnetic Tunnel Junctions Grown on Flexible Organic Substrates. Appl. Phys. Lett. 2010, 96, 072502. (41) Martin, M. B.; Dlubak, B.; Weatherup, R. S.; Yang, H.; Deranlot, C.; Bouzehouane, K.; Petroff, F.; Anane, A.; Hofmann, S.; Robertson, J.; Fert, A.; Seneor, P. Sub-Nanometer Atomic Layer Deposition for Spintronics in Magnetic Tunnel Junctions Based on Graphene Spin-Filtering Membranes. ACS Nano 2014, 8, 7890. (42) He, H.; Kim, K. H.; Danilov, A.; Montemurro, D.; Yu, L.; Park, Y. W.; Lombardi, F.; Bauch, T.; Moth-Poulsen, K.; Iakimov, T.; Yakimova, R.; Malmberg, P.; Müller, C.; Kubatkin, S.; Lara-Avila, S. Uniform Doping of Graphene Close to the Dirac Point by Polymer- Assisted Assembly of Molecular Dopants. Nat. Commun. 2018, 9, 3956. (43) Yang, H. X.; Hallal, A.; Terrade, D.; Waintal, X.; Roche, S.; Chshiev, M. Proximity Effects Induced in Graphene by Magnetic Insulators: First-Principles Calculations on Spin Filtering and Exchange-Splitting Gaps. Phys. Rev. Lett. 2013, 110, 046603. (44) Zhao, B.; Ngaloy, R.; Hoque, A. M.; Karpiak, B.; Khokhriakov, D.; Dash, S. P.Van Der Waals Magnet Based Spin-Valve Devices at Room Temperature, 2021, 2107.00310, arXiv, https://arxiv.org/abs/ 2107.00310 (accessed November 18th, 2022). (45) Deng, Y.; Yu, Y.; Song, Y.; Zhang, J.; Wang, N. Z.; Sun, Z.; Yi, Y.; Wu, Y. Z.; Wu, S.; Zhu, J.; Wang, J.; Chen, X. H.; Zhang, Y. Gate- 41 https://doi.org/10.1021/acs.nanolett.2c03113 Nano Lett. 2023, 23, 34−41