Research Article  www.advmatinterfaces.de Passivation of Bi2Te3 Topological Insulator by Transferred CVD-Graphene: Toward Intermixing-Free Interfaces Regina Galceran,* Frédéric Bonell, Lorenzo Camosi, Guillaume Sauthier, Zewdu M. Gebeyehu, Maria José Esplandiu, Aloïs Arrighi, Iván Fernández Aguirre, Adriana I. Figueroa, Juan F. Sierra, and Sergio O. Valenzuela* 1. Introduction The investigation, and ultimate application, of topological insulators, typically involve exposure to ambient conditions or their integration with Device miniaturization and progress in nanotechnology have evidenced the cru- metals, which lead to surface oxidation or material intermixing. X-ray cial role of interfaces toward the perfor- photoelectron spectroscopy (XPS) measurements that demonstrate passi- mance of devices. Interface properties are vated and intermixing-free interfaces in the topological insulator Bi2Te3 by relevant in all current spintronic technolo- means of dry-transferred CVD graphene are reported. After air exposure, gies, as well as in many novel, potentially no traces of Bi2Te3 oxidation are found. Furthermore, it is demonstrated useful, spintronic phenomena and mate- rials.[1–3] This is the case for topological that graphene acts as a very efficient metal and chalcogen diffusion barrier insulators (TIs), which are characterized in Bi2Te3/graphene/permalloy (Py) heterostructures, which are relevant for by insulating bulk and conductive topo- spintronics. Such results are in stark contrast with the significant surface logically protected surface (or boundary) degradation observed in bare Bi2Te3 under ambient conditions and the states (TSS). The TSS are of particular deep BiTe bonding disruption that occurs in Bi2Te3/Py heterostructures. interest because of their spin-momentum These findings provide a new approach to control and engineer topological locking, which constrains the electron spin orientation perpendicular to the insulator interfaces for spintronic applications and a new platform to electron momentum and represents an investigate the combined use of graphene and topological insulator Dirac effective route for charge-to-spin intercon- states. version (CSI). When a current is applied, the TSS generate a non-equilibrium spin density that can be used to manipulate the R. Galceran, F. Bonell, L. Camosi, G. Sauthier, Z. M. Gebeyehu, magnetization orientation of a ferromagnet (FM) by means of M. J. Esplandiu, A. Arrighi, I. Fernández Aguirre, A. I. Figueroa, spin-orbit torques (SOTs), a key phenomenon for next-genera- J. F. Sierra, S. O. Valenzuela tion magnetic memory technologies.[2] Recent reports using TIs Catalan Institute of Nanoscience and Nanotechnology (ICN2) have demonstrated large SOT efficiencies.[4] However, to iden- CSIC and The Barcelona Institute of Science and Technology (BIST) Campus UAB, Bellaterra, Barcelona 08193, Spain tify the microscopic mechanism of CSI and SOTs, as well as to E-mail: rgalceran@ub.edu; SOV@icrea.cat maximize their efficiency, a deep understanding of the proper- F. Bonell ties of the TI/FM interface is needed.[5–11] Univ. Grenoble Alpes, CNRS, CEA Interface phenomena are complex, involving charge transfer, Spintec, Grenoble 38000, France hybridization, intermixing, or alloying. More so when air- Z. M. Gebeyehu, A. Arrighi, I. Fernández Aguirre sensitive or reactive elements are involved, in which case, Universitat Autònoma de Barcelona oxidation and intermixing are particularly relevant. Different Bellaterra, Barcelona 08193, Spain strategies have been used to protect the surface of TIs from S. O. Valenzuela Institució Catalana de Recerca i Estudis Avançats (ICREA) oxidation, from intercalation of Rb atoms[12] to Te capping (and Barcelona 08010, Spain latter decapping) for ex situ transfers of the TI.[13] In the case of TI/Py (Py = Ni80Fe20) heterostructures, an added challenge The ORCID identification number(s) for the author(s) of this article is the significant intermixing present: nickel was observed to can be found under https://doi.org/10.1002/admi.202201997. diffuse into Bi2Se3, leading to phase separation,[7] and tellu- © 2022 The Authors. Advanced Materials Interfaces published by rium was found to diffuse out of Bi2Te3 deep into Py.[8] In these Wiley-VCH GmbH. This is an open access article under the terms of the experiments, interface diffusion was reduced by intercalating Creative Commons Attribution License, which permits use, distribution thin nonmagnetic (metallic) spacers,[8] which were grown onto and reproduction in any medium, provided the original work is properly the TI before depositing the FM. However, it has been further cited. established that interface modifications can alter the nature of the CSI mechanism, leading to large variations in the spin DOI: 10.1002/admi.202201997 torque[8] and the CSI efficiency.[9,14,15] Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (1 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de Despite increasing efforts, and a variety of metal spacers In this work, we fabricated Bi2Te3/graphene and Bi2Te3/ investigated, interface control in TI/FM structures is still lim- graphene/FM heterostructures in order to investigate the per- ited, while to our knowledge intermixing has never been com- formance of graphene as both an oxidation and diffusion bar- pletely suppressed. Because (Bi,Sb)2(Se,Te)3 family of TIs are rier. By means of X-ray photoelectron spectroscopy (XPS) we van der Waals (vdW) materials, using another vdW material demonstrate that graphene prevents Bi2Te3 from oxidizing, as a spacer represents an attractive alternative to the metals even after exposure to ambient air and moisture. A Py film was used so far. vdW materials are atomically smooth and, as part then grown under ultra-high vacuum (UHV), forming a Bi2Te3/ of a heterostructure, they tend to interact weakly, through vdW graphene/Py heterostructure. The chemical states of Te and interaction, with minimal element intermixing. Among vdW Bi in the Bi2Te3 film were investigated by XPS and compared materials, graphene is an excellent candidate as a diffusion with results obtained with Bi2Te3/Py and Bi2Te3. Our system- barrier, given its closely packed hexagonal lattice, outstanding atic analysis found no trace of intermixing when graphene is mechanical and thermal properties, and chemical stability. intercalated, demonstrating its superior diffusion barrier per- Graphene is known to be impermeable to standard gases, formance for Ni and Te. including helium,[16] and has been used to block the diffusion of metals, such as in Si/Cu,[17] Al/Si, and Au/Ni interfaces.[18] It has also been used to passivate Ni [19] and Cu and CuNi alloy,[20] 2. Results and Discussion although its effective long-term passivation of Cu is contro- verted.[21–25] Some studies report a lack of passivation or even 2.1. TI/Graphene Heterostructure increased oxidation,[22,23] possible explanations including the presence of defects in graphene [24,26] or reversible oxygen inter- Figure 1 shows a schematic of the developed approach to calation at the graphene-Cu interface.[21] Long-term passivation fabricate TI/graphene heterostructures (see the Experimental of FMs such as Ni and Co by direct chemical vapor deposition Section for details on each step). The topological insulator, (CVD) of graphene has also been reported,[27] and used as elec- Bi2Te3, was grown by molecular beam epitaxy[33] (MBE) on a trodes in magnetic tunnel junctions.[28–30] In fact, the interac- BaF2(111) substrate and transferred, using a UHV suitcase, to tion between the graphene and the FM leads to hybridization a glovebox with inert Ar gas, where the O2 and the H2O con- (chemisorption or physisorption), both potentially useful for centrations were below 1 ppm. In order to achieve a large-area spintronic applications.[31] In addition, although TI/graphene/ scalable process, graphene was grown by CVD on a Cu foil.[34] FM heterostructures have not been investigated experimentally, Dry graphene stamps were prepared using a bilayer of poly- giant CSI and large SOTs have been predicted.[32] propylene carbonate (PPC) and polydimethylsiloxane (PDMS), Figure 1. Sketch of the sample fabrication process and basic characterization. a) A dry stamp of PDMS/PPC/graphene was introduced into the glovebox and pressed onto the sample (graphene facing down) while the latter was heated to over 100 °C to melt the PPC; b) after cleaning the resist, the final sample allowed for comparison between regions with and without graphene coverage. c) Raman spectrum confirmed the presence of graphene on the covered region, using a laser wavelength of 488 nm. d) SEM images reveal graphene full coverage in the majority of the transferred area (as in 1d1) as opposed to visible graphene discontinuities (cracks) formed during non-optimized transfer (shown as reference in 1d2). Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (2 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de Figure 2. XPS measurements comparing the regions with and without graphene coverage (closed black and open red symbols, respectively). Te 3d, Bi 4f, Bi 5d, and Te 4d peaks are shown in the panels (b–d) as labeled. In all cases, large oxide peaks are present in the areas without graphene. Such peaks are absent for the graphene-capped area. Spectra intensities have been shifted and normalized to the value of the maximum of the largest peak. followed by the removal of the Cu foil by means of wet 2.2. Graphene as Oxidation, Passivation Barrier etching. The graphene stamp and Bi2Te3 were put in contact using an optical microscope and micro-positioners (Figure 1a). Figure 2 compares XPS measurements for Bi2Te3 over regions The whole stack was then heated to over 100 °C to melt the with and without graphene coverage (black-solid and red- PPC and transfer the graphene/PPC onto Bi2Te3. Up to this open symbols, respectively). The direct comparison of the two point, the process was carried out in the glovebox to avoid oxi- regions (represented in Figure 2a) allows us to eliminate any dation of the Bi2Te3 surface. Afterwards the sample was taken characterization artifacts associated with air-exposure time and out of the glove box, the PPC was removed with acetone and ambient conditions, as well as any possible variation in the isopropanol, and then the bilayer was exposed to air for over morphology of the grown Bi2Te3 films. an hour. In the finished sample, graphene covered Bi2Te3 The XPS spectra reveal peaks at the Te 3d (Figure 2b), Bi only partially to enable a direct comparison between regions 4f (Figure 2c), and Te 4d and Bi 5d (Figure 2d) edges. Several with and without graphene coverage in the same Bi2Te3 film peaks in Figure 2b,c are directly associated to Bi2Te3, located at (Figure 1b). 572.2 eV (Te 3d5/2), 582.6 eV (Te 3d3/2), 157.6 eV (Bi 4f7/2) and Poor contrast using an optical microscope makes it dif- 162.9 eV (Bi 4f5/2). Additional peaks are observed only in bare ficult to distinguish the quality of the graphene, or even its Bi2Te3 at larger binding energies (575.9 and 586.4 eV for Te and presence, on the Bi2Te3. However, this can be readily achieved 158.8 and 164.1 eV for Bi) and are indicative of substantial oxi- with Raman spectroscopy. The Raman spectrum presented in dation. By fitting the areas of the Bi, Te and their corresponding Figure 1c shows the two characteristic peaks for Bi2Te3 at low oxides, and considering electron escape depths of λBi = 2.3 nm[35] frequencies (Raman shift < 400 cm−1), while the presence of and  λTe = 2.1 nm,[36] we estimate the oxide thickness to be graphene is confirmed by means of the G and 2D peaks, visible ≈1.3 nm and 0.7 nm for Bi and Te, respectively (see Experi- at larger frequencies (≈1600 cm−1 and ≈2700 cm−1). The faint mental Section for more details). This is in stark contrast with D peak at ≈1350 cm−1 demonstrates the successful transfer of the Bi2Te3/graphene region, in which no oxidation is discern- good-quality graphene with a relatively low density of defects. able. An equivalent analysis, yielding the same conclusion, Scanning electron microscope (SEM) images also confirm can be carried out in the Te 3d and Bi 5d edges (Figure 2d). full continuous graphene coverage over the majority of the Note that the spot size of the XPS is in the few mm range, con- transferred area when using an optimized transfer process firming that the graphene transfer has been successful and that (Figure 1d.1). This contrasts with the visible graphene cracks the TI passivation is effective over a large area. present in regions of samples for which the transfer process Additional angle-resolved photoemission spectroscopy was not optimal (Figure 1d.2). measurements (see Supplementary Information) on the Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (3 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de Bi2Te3/graphene heterostructure show visible TSS of the TI and slight electron doping. However, n-doping cannot be unam- biguously attributed to the presence of the graphene capping layer. Such doping has been reported to increase over time in in situ cleaved TIs, even while maintaining the sample inside UHV conditions.[37] The successful fabrication of TI/graphene heterostructures by dry transferring CVD graphene, while preventing oxida- tion, is an important milestone. Such heterostructures have attracted significant attention as an electronic platform to combine low-energy Dirac states with very distinct properties and due to the potential to enhance the spin-orbit coupling of graphene by proximity phenomena with the TI.[38–44] Because exposure to ambient conditions causes surface oxidation in the (Bi,Sb)2(Te,Se)3 TIs,[45–48] many experimental studies using these materials aimed at minimizing the time of exposure to air[41,43] or to water-based solutions during graphene transfer.[49] However, despite these efforts, some degree of surface modi- fication or oxidation is unavoidable.[45–48] Alternative strategies, such as the deposition of the TI on graphene (either epitaxially grown[50] or exfoliated[51,52]), have also been adopted. However, the characterization of these heterostructures must be carried out through transport experiments, as access to the buried interface is not possible, unless the substrate is removed.[40] 2.3. TI/Graphene/FM Heterostructure: Graphene as Intermixing/Diffusion Barrier Figure 3. X-ray photoemission spectroscopy at the Bi 4f 7/2 and Te 3d5/2 peaks for different stacks with and without graphene interlayer. From bottom to top: pure Bi0 and Te0 (black symbols), Bi2Te3 (red symbols), After confirming the lack of oxidation of Bi2Te3 with a graphene in situ grown Bi2Te3/Py/AlOx (green symbols), Bi2Te3/graphene (orange barrier, a bilayer of Py(2 nm)/Al(2 nm) was grown on top of symbols) and Bi2Te3/graphene/Py/AlOx (blue symbols). The peak posi- Bi2Te3/graphene in UHV (see Experimental Section for details). tions for pure Bi0 and Te0 (dotted black line), as well as those corre- The Py and Al layers were made thin enough to be able to sponding to Bi2Te3 (solid red line), are indicated as references. Data from probe the composition and chemical state of the buried TI. XPS samples not containing graphene was taken from reference [8]. is a surface-sensitive technique; the escape depth of the photo- electrons in most metals is ≈2–5 nm and thus, the XPS signal broken Bi-Te bonds are present. This is a notable result given would be strongly suppressed for larger thicknesses. the large affinity of Py for Te.[8] Suppression of the intermixing Figure 3 shows XPS spectra around the binding energies of has been achieved by intercalating aluminum, silver, and Te the Bi4f7/2 and Te3d5/2 edges. The results for Bi2Te3/graphene/ between Bi2Te3 and Py. However, the observed suppression with Py are plotted with blue symbols (top trace) while those for these materials is only partial, even with several nanometers of Bi2Te3/Py are plotted with green symbols (middle trace). Addi- Te, as concluded from the coexistence of the bare Bi2Te3 and tional measurements for pure Bi and Te (black) and for Bi2Te3 pure Bi and Te emission lines in those heterostructures.[8] The (red) and Bi2Te3/graphene (orange) are shown as references. complete suppression of intermixing in Bi2Te3/graphene/Py, as The emission lines for bare Bi2Te3 at 572.2 eV (Bi4f7/2) and demonstrated in Figure 3 and observed here for the first time, 157.6 eV (Te3d5/2) are indicated with red solid vertical lines; those therefore represents a remarkable improvement with respect to for pure Bi and Te at 156.9 eV and at 573.0 eV, respectively, are metallic interlayers. indicated with black dotted vertical lines. The observed energy shift of 0.7-0.8 eV in Bi2Te3, which is opposite for Bi4f7/2 and Te3d5/2 in relation to the pure elements, is due to the charge 3. Conclusion transfer occurring in the Bi-Te covalent bonds within the TI. The emission lines in Bi2Te3/Py are very different from those We have prepared Bi2Te3/graphene heterostructures by dry in Bi2Te3, having instead the same binding energies that char- transferring CVD-grown graphene onto MBE-grown Bi2Te acterize pure Bi and Te. As previously reported,[8] this is the under an inert atmosphere. We have shown that the graphene result of a strong modification of the chemical Bi-Te bonding layer effectively passivates the Bi2Te3 surface, even after expo- in Bi2Te3 upon deposition of Py. Angular XPS measurements sure to ambient conditions, as the XPS spectra show no traces have shown that Te diffuses out of Bi2Te3 leaving a Bi-rich of surface oxidation. interface.[8] In stark contrast, both Bi2Te3/graphene and Bi2Te3/ In addition to surface passivation, graphene acts as a graphene/Py maintain the exact peak positions of Bi2Te3, dem- very efficient metal and chalcogen diffusion barrier in rel- onstrating that all detected Bi and Te remain in Bi2Te3 and no evant Bi2Te3/graphene/Py heterostructures. By means of XPS Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (4 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de measurements, we show that the chemical bonding of the (2 nm) with effusion cells at growth rates of ≈0.24 and 0.15 nm min−1, Bi2Te3 remains intact after depositing Py on top of graphene. respectively. The thickness was monitored with a calibrated quartz This result contrasts with the deep BiTe bonding disruption crystal microbalance. After growth, the sample was taken out to air for the oxidation of the Al into AlOx. observed in Bi2Te3/Py heterostructures, or even when selected Raman measurements: Raman spectra were taken using a Witec metallic interlayers are inserted between Bi2Te3 and Py. Alpha300 R confocal Raman spectrometer and the 488 nm laser Our heterostructure fabrication method does not require line, using a power of 0.8 mW. The spectra were measured with substrate-dependent growth optimization and allows for large- 100× magnification. The grating used for probing the graphene peaks scale samples. Even though the passivation of the surface and was 600 lines mm−1, while a grating of 1800 lines mm−1 was used for intermixing-free interfaces are demonstrated for Bi2Te3, we higher resolution to investigate the spectra for Bi2Te3. XPS measurements: After exposure to air, samples were introduced expect similar results for the whole family of (Bi,Sb)2(Se,Te)3 into a UHV chamber (10−10 mbar), and XPS was carried out using a TI materials. Considering the increasing interest in quantum SPECS PHOIBOS hemispherical analyzer with a monochromatized X-ray transport in TIs and graphene, in proximity effects, and in source (al Ka line of 1486.6 eV, 300 W). Calibration was done using the spintronic applications relying on TI/FM interfaces, our results C 1s peak as reference. For the measurement of the non-oxidized Bi2Te3 surface, the film was grown in the MBE and capped by a Te layer (15 nm) offer important prospects for fundamental progress and for deposited in situ below 100 °C. Inside the XPS chamber, the Te capping future spintronics. Regarding spintronics, future work should layer was desorbed by heating the sample up to 190 °C for 10 min. For focus on the properties of the ferromagnet, which is grown on the sample with and without graphene coverage (Bi2Te3/graphene), the TI/graphene bilayer, and, for example, the efficiency of the no previous annealing was done prior to the XPS measurements. No spin-orbit torques in the finished heterostructure. change was observed in the XPS peak positions after annealing at 300 °C for 30 min, just an increase in their intensity. To estimate the oxide thickness, it was noted that Bi2Te3 was much thicker than the electron ≈ 1− exp  − ox  , where Iox and Inon−ox Iox t escape depth. This led to 4. Experimental Section Inon − ox  λ are the peak areas corresponding to the oxidized and non-oxidized TI Film Growth: Bi2Te3 films with a thickness of ten quintuple elements (Te or Bi), respectively. As stated in the main text, λ = 2.3 nm layers (10 nm) were grown by co-evaporation of Bi and Te in an MBE (2.1 nm) was taken at the Bi 4f (Te 3d) energy.[35] The analysis performed chamber with a base pressure of 2 × 10−10 mbar. The growth rate was separately at the Bi 4f and Te 3d lines yielded oxide thicknesses of 1.3 nm ≈0.25 nm min−1. The fluxes were previously measured with a quartz and 0.7 nm for Bi and Te, respectively. crystal microbalance. The substrates used, BaF2 (111) purchased at PIKEM, were previously outgassed in UHV at 800 °C for 1 h (in the same chamber) and the growth was done at 380 °C and monitored by in situ reflection high energy electron diffraction. Films were transferred, using a vacuum suitcase (≈10−8 mbar), to a glovebox under Ar atmosphere, Supporting Information where the O2 and the H2O concentrations were below 1 ppm. There, Supporting Information is available from the Wiley Online Library or graphene was transferred on top, as described in the main text. The from the author. sample temperature given here was the temperature measured with a thermocouple located between the heater and the sample holder, leading to an overestimation of the real temperature, which correspond to ≈575 °C for the degassing and 260 °C for the growth. Acknowledgements Graphene Growth and Stamp Preparation: Graphene was grown on a 25 µm thick Cu foil (99.8% or 99.999% Alfa Aesar) by CVD The authors acknowledge the support of the Spanish Research Agency using a Lindberg/Blue tube furnace. The copper foil was first (AEI), Ministry of Science and Innovation (MCIN) under contract cleaned in acetone and isopropanol and then electropolished in an no. PID2019-111773RB-I00/AEI/10.13039/501100011033, PGC2018- electrochemical cell by applying 3 V for 90 s and using the copper foil 095032-B-100, and SEV-2017-0706 Severo Ochoa, and by PCI2021- as anode. The electrolyte solution was a mixture of 150 mL of water, 122035-2A funded by MICIN and by the European Union Next 75 mL of phosphoric acid, 75 mL of ethanol, 15 mL of isopropanol, Generation EU/PRTR, and of the European Union’s Horizon 2020 and 1.5 g of urea.[53] After that, the polished Cu foil was annealed in under grant agreement 881603 (Graphene Flagship). R.G. acknowledges Ar (450 sccm) and H2 (50 sccm) for 45 min under low pressure (base funding from the European Union’s Horizon 2020 research and pressure of 20 mbar). Graphene was then grown using a CH4 flow of innovation program under the Marie Sklodowska-Curie grant agreement 30 sccm at 1000 °C for 10 min and keeping the same flow conditions No 840588 (GRISOTO) and I.F.A from a fellowship from la Caixa for H2 and Ar as in the annealing step. These growth conditions Foundation (ID 100010434) with code LCF/BQ/DI18/11660030 and produce a full multidomain graphene monolayer and also occasionally of H2020 Marie Skłodowska-Curie grant agreement No. 713673. J.F.S small domains of bilayer graphene at the center of the individual acknowledges support from MINECO under contract RYC2019-028368-I/ graphene domains.[54] For graphene transfer process, four layers of AEI/10.13039/50110001103. polypropylene carbonate (PPC) (at a concentration of 0.1 gr mL−1) were spin-coated on top of the graphene and heated to 80 °C during 1 min after each layer. Finally, a thicker polydimethylsiloxane (PDMS) film was attached to the surface of the PPC, to later provide Conflict of Interest support with the handling of the stamp. Cu was wet-etched with The authors declare no conflict of interest. diluted ammonium persulfate (0.25 m) and rinsed with water as was standard procedure[55]. The PDMS stamp with graphene was left to dry overnight and entered into the glovebox. Py/Al deposition: After the characterization of the Bi2Te3/graphene Data Availability Statement surface, the sample was reintroduced in the MBE chamber. The sample was outgassed at 280 °C for 1 h in UHV and cooled down to The data that support the findings of this study are available from the liquid nitrogen temperature for the evaporation of Py (2 nm) and Al corresponding author upon reasonable request. Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (5 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de Keywords [22] M. Schriver, W. Regan, W. J. Gannett, A. M. Zaniewski, M. F. Crommie, A. Zettl, ACS Nano 2013, 7, 5763. Bi2Te3, graphene−topological insulator interface, intermixing, [23] F. Zhou, Z. Li, G. J. Shenoy, L. Li, H. Liu, ACS Nano 2013, 7, 6939. passivation, XPS [24] D. Prasai, J. C. Tuberquia, R. R. Harl, G. K. Jennings, K. I. Bolotin, ACS Nano 2012, 6, 1102. Received: September 8, 2022 [25] C. G. Kang, S. K. Lim, S. Lee, S. K. Lee, C. Cho, Y. G. Lee, Revised: October 30, 2022 H. J. Hwang, Y. Kim, H. J. Choi, S. H. Choe, M. H. Ham, B. H. Lee, Published online: December 6, 2022 Nanotechnology 2013, 24, 115707. [26] Y. P. Hsieh, M. Hofmann, K. W. Chang, J. G. Jhu, Y. Y. Li, K. Y. Chen, C. C. Yang, W. S. Chang, L. C. Chen, ACS Nano 2014, 8, 443. [1] I. Žutić, J. Fabian, S. das Sarma, Rev. Mod. Phys. 2004, 76, 446. [27] R. S. Weatherup, L. D’Arsié, A. Cabrero-Vilatela, S. Caneva, [2] B. Dieny, I. L. Prejbeanu, K. Garello, P. Gambardella, P. Freitas, R. Blume, J. Robertson, R. Schloegl, S. Hofmann, J. Am. Chem. Soc. R. Lehndorff, W. Raberg, U. Ebels, S. O. Demokritov, J. Akerman, 2015, 137, 14358. A. Deac, P. Pirro, C. Adelmann, A. Anane, A. v. Chumak, A. Hirohata, [28] B. Dlubak, M. B. Martin, R. S. Weatherup, H. Yang, C. Deranlot, S. Mangin, S. O. Valenzuela, M. C. Onbaşlı, M. d’Aquino, G. Prenat, R. Blume, R. Schloegl, A. Fert, A. Anane, S. Hofmann, P. Seneor, G. Finocchio, L. Lopez-Diaz, R. Chantrell, O. Chubykalo-Fesenko, J. Robertson, ACS Nano 2012, 6, 10930. P. Bortolotti, Nat. Electron. 2020, 3, 446. [29] V. Zatko, S. M. M. Dubois, F. Godel, M. Galbiati, J. Peiro, A. Sander, [3] J. F. Sierra, J. Fabian, R. K. Kawakami, S. Roche, S. O. Valenzuela, C. Carretero, A. Vecchiola, S. Collin, K. Bouzehouane, B. Servet, Nat. Nanotechnol. 2021, 16, 856. F. Petroff, J. C. Charlier, M. B. Martin, B. Dlubak, P. Seneor, ACS [4] C.-F. Pai, Nat. Mater. 2018, 17, 755. Nano 2022, 16, 14007. [5] V. P. Amin, P. M. Haney, M. D. Stiles, J Appl Phys 2020, 128, 151101. [30] P. Martin, B. Dlubak, R. Mattana, P. Seneor, M. B. Martin, [6] R. Galceran, B. Tian, J. Li, F. Bonell, M. Jamet, C. Vergnaud, T. Henner, F. Godel, A. Sander, S. Collin, L. Chen, S. Suffit, F. Mallet, A. Marty, J. H. García, J. F. Sierra, M. v. Costache, S. Roche, P. Lafarge, M. L. della Rocca, A. Droghetti, C. Barraud, Nanoscale S. O. Valenzuela, A. Manchon, X. Zhang, U. Schwingenschlögl, APL 2022, 14, 12692. Mater. 2021, 9, 100901. [31] M. Piquemal-Banci, R. Galceran, S. M. M. Dubois, V. Zatko, [7] S.-J. Chang, P.-Y. Chuang, C.-W. Chong, Y.-J. Chen, J.-C. A. Huang, M. Galbiati, F. Godel, M. B. Martin, R. S. Weatherup, F. Petroff, P.-W. Chen, Y.-C. Tseng, RSC Adv. 2018, 8, 7785. A. Fert, J. C. Charlier, J. Robertson, S. Hofmann, B. Dlubak, [8] F. Bonell, M. Goto, G. Sauthier, J. F. Sierra, A. I. Figueroa, P. Seneor, Nat. Commun. 2020, 11, 5670. M. V. Costache, S. Miwa, Y. Suzuki, S. O. Valenzuela, Nano Lett. [32] M. Rodriguez-Vega, G. Schwiete, J. Sinova, E. Rossi, Phys. Rev. B 2020, 20, 5893. 2017, 96, 235419. [9] E. Longo, L. Locatelli, M. Belli, M. Alia, A. Kumar, M. Longo, [33] F. Bonell, M. G. Cuxart, K. Song, R. Robles, P. Ordejón, S. Roche, M. Fanciulli, R. Mantovan, Adv. Mater. Interfaces 2021, 8, 2101244. A. Mugarza, S. O. Valenzuela, Cryst. Growth Des. 2017, 17, 4655. [10] E. Rongione, S. Fragkos, L. Baringthon, J. Hawecker, [34] Z. M. Gebeyehu, A. Arrighi, M. v. Costache, C. M. Sotomayor-Torres, E. Xenogiannopoulou, P. Tsipas, C. Song, M. Mičica, J. Mangeney, M. J. Esplandiu, S. O. Valenzuela, RSC Adv. 2018, 8, 8234. J. Tignon, T. Boulier, N. Reyren, R. Lebrun, J. M. George, P. le Fèvre, [35] H. Shinotsuka, S. Tanuma, C. J. Powell, D. R. Penn, Surf. Interface S. Dhillon, A. Dimoulas, H. Jaffrès, Adv. Opt. Mater. 2022, 10, Anal. 2015, 47, 1132. 2102061. [36] E. Nolot, C. Sabbione, W. Pessoa, L. Prazakova, G. Navarro, Appl. [11] E. Longo, M. Belli, M. Alia, M. Rimoldi, R. Cecchini, M. Longo, Surf. Sci. 2021, 536, 147703. C. Wiemer, L. Locatelli, P. Tsipas, A. Dimoulas, G. Gubbiotti, [37] M. Bianchi, D. Guan, S. Bao, J. Mi, B. B. Iversen, P. D. C. King, M. Fanciulli, R. Mantovan, Adv. Funct. Mater. 2022, 32, 2109361. P. Hofmann, Nat. Commun. 2010, 1, 128. [12] M. Bianchi, R. C. Hatch, Z. Li, P. Hofmann, F. Song, J. Mi, B. B. Iversen, [38] J. Zhang, C. Triola, E. Rossi, Phys. Rev. Lett. 2014, 112, 096802. Z. M. Abd El-Fattah, P. Löptien, L. Zhou, A. A. Khajetoorians, [39] C. D. Spataru, F. Léonard, Phys. Rev. B 2014, 90, 085115. J. Wiebe, R. Wiesendanger, J. W. Wells, ACS Nano 2012, 6, 7009. [40] P. Lee, K. H. Jin, S. J. Sung, J. G. Kim, M. T. Ryu, H. M. Park, [13] G. Kremer, K. Zhu, T. Pierron, V. Fournee, J. Ledieu, S. Andrieu, S. H. Jhi, N. Kim, Y. Kim, S. U. Yu, K. S. Kim, D. Y. Noh, J. Chung, B. Kierren, L. Moreau, D. Malterre, K. He, Q. K. Xue, Y. Fagot- ACS Nano 2015, 9, 10861. Revurat, Y. Lu, J Phys D Appl Phys 2019, 52, 494002. [41] G. Bian, T.-F. Chung, C. Chen, C. Liu, T.-R. Chang, T. Wu, [14] R. Sun, S. Yang, X. Yang, E. Vetter, D. Sun, N. Li, L. Su, Y. Li, Y. Li, I. Belopolski, H. Zheng, S.-Y. Xu, D. S. Sanchez, N. Alidoust, Z.-Z. Gong, Z.-K. Xie, K.-Y. Hou, Q. Gul, W. He, X.-Q. Zhang, J. Pierce, B. Quilliams, P. P. Barletta, S. Lorcy, J. Avila, G. Chang, Z.-H. Cheng, Nano Lett. 2019, 19, 31. H. Lin, H.-T. Jeng, M.-C. Asensio, Y. P. Chen, M. Z. Hasan, 2D [15] H. He, L. Tai, D. Wu, H. Wu, A. Razavi, K. Wong, Y. Liu, K. L. Wang, Mater. 2016, 3, 021009. APL Mater. 2021, 9, 071104. [42] K. Song, D. Soriano, A. W. Cummings, R. Robles, P. Ordejón, [16] J. S. Bunch, S. S. Verbridge, J. S. Alden, A. M. van der Zande, S. Roche, Nano Lett. 2018, 18, 2033. J. M. Parpia, H. G. Craighead, P. L. McEuen, Nano Lett. 2008, 8, 2458. [43] S. Jafarpisheh, A. W. Cummings, K. Watanabe, T. Taniguchi, [17] J. Hong, S. Lee, S. Lee, H. Han, C. Mahata, H. W. Yeon, B. Koo, B. Beschoten, C. Stampfer, Phys. Rev. B 2018, 98, 241402. S. il Kim, T. Nam, K. Byun, B. W. Min, Y. W. Kim, H. Kim, Y. C. Joo, [44] K. Zollner, J. Fabian, Phys. Rev. B 2019, 100, 165141. T. Lee, Nanoscale 2014, 6, 7503. [45] D. Kong, J. J. Cha, K. Lai, H. Peng, J. G. Analytis, S. Meister, Y. Chen, [18] H.-Y. Kim, C. Lee, J. Kim, F. Ren, S. J. Pearton, J. Vac. Sci. Technol., B H.-J. Zhang, I. R. Fisher, Z.-X. Shen, Y. Cui, ACS Nano 2011, 5, 4698. 2012, 30, 030602. [46] L. v. Yashina, J. Sánchez-Barriga, M. R. Scholz, A. A. Volykhov, [19] Y. S. Dedkov, M. Fonin, C. Laubschat, Appl. Phys. Lett. 2008, 92, 052506. A. P. Sirotina, S. N. Vera, M. E. Tamm, A. Varykhalov, D. Marchenko, [20] S. Chen, L. Brown, M. Levendorf, W. Cai, S. Y. Ju, J. Edgeworth, G. Springholz, G. Bauer, A. Knop-Gericke, O. Rader, ACS Nano X. Li, C. W. Magnuson, A. Velamakanni, R. D. Piner, J. Kang, J. Park, 2013, 7, 5181. R. S. Ruoff, ACS Nano 2011, 5, 1321. [47] A. J. Green, S. Dey, Y. Q. An, B. O’Brien, S. O’Mullane, B. Thiel, [21] P. R. Kidambi, B. C. Bayer, R. Blume, Z. J. Wang, C. Baehtz, A. C. Diebold, J Vac Sci Technol A 2016, 34, 061403. R. S. Weatherup, M. G. Willinger, R. Schloegl, S. Hofmann, Nano [48] P. A. Sharma, T. Ohta, M. T. Brumbach, J. D. Sugar, J. Michael, ACS Lett. 2013, 13, 4769. Appl. Mater. Interfaces 2021, 13, 18218. Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (6 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH 21967350, 2022, 36, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/admi.202201997 by Csic Organizacion Central Om, Wiley Online Library on [17/04/2023]. 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 License www.advancedsciencenews.com www.advmatinterfaces.de [49] S. Rajput, Y. Y. Li, M. Weinert, L. Li, ACS Nano 2016, 10, 8450. [53] S. Chen, H. Ji, H. Chou, Q. Li, H. Li, J. W. Suk, R. Piner, L. Liao, [50] N. Kim, P. Lee, Y. Kim, J. S. Kim, Y. Kim, D. Y. Noh, S. U. Yu, W. Cai, R. S. Ruoff, Adv. Mater. 2013, 25, 2062. J. Chung, K. S. Kim, ACS Nano 2014, 8, 1154. [54] H. Zhou, W. J. Yu, L. Liu, R. Cheng, Y. Chen, X. Huang, Y. Liu, [51] D. Khokhriakov, A. M. Hoque, B. Karpiak, S. P. Dash, Nat. Commun. Y. Wang, Y. Huang, X. Duan, Nat. Commun. 2013, 4, 2096. 2020, 11, 3657. [55] A. Ibrahim, G. Nadhreen, S. Akhtar, F. M. Kafiah, T. Laoui, Carbon N [52] A. Zalic, T. Dvir, H. Steinberg, Phys. Rev. B 2017, 96, 75104. Y 2017, 123, 402. Adv. Mater. Interfaces 2022, 9, 2201997 2201997 (7 of 7) © 2022 The Authors. Advanced Materials Interfaces published by Wiley-VCH GmbH