A Local Study of the Transport Mechanisms in MoS2 Layers for

Subscriber access provided by UNIV OF DURHAM

Letter

A local study of the transport mechanisms in MoS2 layers for magnetic tunnel junctions Marta Galbiati, Aymeric Vecchiola, Samuel Mañas-Valero, Josep Canet-Ferrer, Régina Galceran, Maëlis Piquemal-Banci, Florian Godel, Alicia Forment-Aliaga, Bruno Dlubak, Pierre Seneor, and Eugenio Coronado ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b08853 • Publication Date (Web): 06 Aug 2018 Downloaded from http://pubs.acs.org on August 7, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

A local study of the transport mechanisms in MoS2 layers for magnetic tunnel junctions

Marta Galbiati1,*, Aymeric Vecchiola2, Samuel Mañas-Valero1, Josep Canet-Ferrer1,3, Regina Galceran2, Maëlis Piquemal-Banci2, Florian Godel2, Alicia Forment-Aliaga1, Bruno Dlubak2, Pierre Seneor2,*, Eugenio Coronado1,* 1

Instituto de Ciencia Molecular, Universitat de Valencia, Catedrático José Beltrán Martínez nº 2, 46980 Paterna, Spain 2

Unité Mixte de Physique CNRS/Thales, 1 Av. A. Fresnel, 91767 Palaiseau, France and Université Paris-Sud, 91405 Orsay, France

3

ICFO-Institut de Ciències Fotòniques, the Barcelona Institute of Science and Technology, 08860 Barcelona, Spain

KEYWORDS MoS2, 2D materials, spintronics, conductive tip AFM, vertical transport, spin valves, magnetic tunnel junctions.

ABSTRACT

MoS2-based vertical spintronic devices have attracted an increasing interest thanks to theoretical predictions of large magnetoresistance signals. However, experimental performances are still far from expectations. Here we carry out the local electrical

1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 16

characterization of thin MoS2 flakes in a Co/Al2O3/MoS2 structure through conductive tip AFM measurements. We show that thin MoS2 presents a metallic behaviour with a strong lateral transport contribution that hinders the direct tunnelling through thin layers. Indeed, no resistance dependence is observed with the flake thickness. These findings reveal a spin depolarization source in the MoS2-based spin valves, thus pointing to possible solutions to improve their spintronic properties.

Since the isolation of graphene in 2004, there has been an explosion in the search for other 2D materials. For instance, transition metal dichalcogenides (TMDCs) have become very popular thanks to the wide range of electronic, optical, mechanical, chemical and thermal properties they can exhibit.1 Among them, MoS2 has become a prototypical 2D material beyond graphene thanks to its stability after exfoliation and its semiconducting electrical properties, which have been very appreciated for the fabrication of field-effect transistors (FETs).2 In addition, the bandgap in MoS2 has been found to change from an indirect bandgap in the bulk to a direct one in the single layer limit, thus opening many possibilities for optoelectronic applications.3 In this direction, the transport properties of MoS2 in lateral devices have been largely investigated, while in vertical devices these properties have remained scarcely explored. In spintronics, 2D materials offer ultrathin and free of defect barriers with sharp interfaces, which are ideal to be used as tunnel spacers in magnetic tunnel junctions (MTJs), allowing for a further down-scaling and tuning of these devices.4 In this context, MoS2-based vertical spintronic devices have recently attracted an increasing interest thanks to theoretical predictions that envisaged large magnetoresistance signals.5,6 However, the experimental results so far reported on this kind of devices are still far from these expectations.7,8,9,10 In this scenario, the understanding of transport mechanisms and interface control plays a key role for 2 ACS Paragon Plus Environment

Page 3 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


signal improvement. Indeed, these points are particularly critical in spintronics, where interfacial resistances have to be strictly controlled such as with impedance mismatch issue11,12,13 and where the spin polarization has to be preserved. Recent experiments and calculations have highlighted that the interaction between metals and MoS2 thin layers at the interface evades a straightforward pattern, impeding in turn the easy understanding of devices transport.14 In fact, the presence of the metal contacts alters the electronic and chemical properties of the MoS2 monolayer, which becomes metallic. The formation of an interface dipole and the metal induced gap states lead to partial Fermi level pinning effects and the creation of a Schottky contact barrier.14 This effect can be minimized by inserting a buffering material, as for example an oxide barrier, in between the metal and the MoS2 monolayer.14,15,16 In this letter we investigate the vertical transport mechanisms in thin MoS2 layers at a local level by using conductive tip atomic force microscopy (CT-AFM). We focus our experiments on a Co/Al2O3/MoS2 structure. Co is used as bottom electrode since it is a standard ferromagnetic electrode in spintronic devices. Al2O3 tunnel barrier is used to decouple MoS2 layers from the metallic surface and thus avoid the above-mentioned alteration of the MoS2 monolayer electric properties, while providing band alignment for spin and charge injection. In this way it is possible to study the general electronic properties of MoS2 monolayer not influenced by the direct interaction with a metallic surface.14,15,16 We characterize MoS2 flakes by AFM and Raman spectroscopy to select thin layers and we measure their resistance using a diamond AFM conductive tip. We observe that no resistance variation can be appreciated with flake thickness and we demonstrate that the electric charge is completely delocalized over all the interconnected flakes instead of tunnelling directly through the thin layers. This result is a key to understand and unlock the efficient use of 2D semiconducting TMDCs such as MoS2 for spin transport in magnetic tunnel junctions.



Page 4 of 16

Samples with Co/Al2O3/MoS2 structure were fabricated by magnetron sputtering of a Co metallic layer (15 nm) over a Si/SiO2 substrate and covered with the deposition of 0.6 nm Al film without breaking the vacuum atmosphere. Subsequently, the Al film was oxidized in a 50 Torr pure O2 atmosphere for 10 min. This resulted in a high-quality 1 nm thick Al2O3 tunnel barrier as shown in previous works11,17,18,19 and in Figure S1 of Supporting Information. MoS2 bulk crystal (from Manchester Nanomaterials Company) was mechanically exfoliated using the “scotch-tape” technique20 and the resulting flakes were deposited over the Co/Al2O3 surface. Optical microscopy was used to select thin layers. Flakes contrast on this surface was much lower than the one obtained over optimized SiO2 thickness, thus making the identification of thin layers more difficult (Figure 1a). Still, the optical identification by contrast remained possible. Once thin flakes were located, their thickness was investigated by AFM (Figure 1b, Figure S1) and Raman spectroscopy was used as a complementary technique to determine the number of layers.21 Figure 1c shows the A1g and E12g peaks of the MoS2 Raman spectra for a monolayer, a bilayer and thicker layers. As reported by Li et al.,21 the distance between these two peaks allows clear identification of thinner flakes as it decreases monotonically from 5 layers to 1 layer. We obtain a distance of 19 cm!1 for the single layer, 22 cm-1 for the bilayer and 24 cm-1 for the thicker flakes. These values are comparable to the ones reported in the literature.


Page 5 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ACS Paragon Plus Environment


Page 6 of 16

that of a monolayer connected to thicker flakes. First of all, one can note an overall homogeneous resistance on all the connected flakes with no substantial variation as a function of the flake thickness. Surprisingly, this is contrary to what is expected from direct tunnelling through barriers of different height. Then, if one compares the resistances of the isolated and connected monolayers, a higher resistance can be observed on the cut out part of the monolayer (log (R1L_isolated)= 11.5 ± 0.1), while a much lower one is measured on the monolayer still connected to thicker flakes (log (R1L_connected)= 8.0 ± 0.4). In order to better understand this behavior, we have also isolated another part of the flake by performing a vertical cut (line 2) with the diamond AFM tip as shown in Figure 2c. A high resistance was then measured on both parts of the flake and constant R!A products of 307 ± 79 G"!µm2 and 289 ± 49 G"!µm2 were calculated for the top and the bottom part of the isolated flakes, respectively. These values, being very similar within the margin of error of our measurements, demonstrate the presence of a strong lateral contribution in the transport mechanism of this system. Instead of tunnelling directly through the thin layers, as it would be intuitively expected, the electric charge is completely delocalized on all the interconnected flakes. Thus, the MoS2 layer acts almost as an electrode that increases the small surface area of the AFM tip and decreases the vertical resistance through the Al2O3 barrier and through the interlayer MoS2 sheets. This result confirms the hypothesis made by Quereda et al.22 for modelling their ITO/MoS2 vertical structure measured with a conductive diamond AFM tip. They claimed that the current perpendicularly injected in their system was expected to flow in-plane and spread before reaching the bottom electrode, owing to the fact that the intralayer conductivity of MoS2 individual layers was expected to be higher than the interlayer one (200 times higher).23,24


Page 7 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60




Page 8 of 16

connected to thicker ones (blue markers) are found to be in good agreement with I-V curve behaviour measured on a multilayer flake (blue line). This proves that the thicker layer conduction dominates the transport behaviour in a monolayer connected to it. Indeed, the connected monolayer and the multilayer measured here were both part of a big flake with a total area of about 470 µm2 and their similar conduction is due to electric charges that are laterally delocalized on all the connected layers as schematized in Figure 3b. On the contrary, a different trend can be detected in the isolated monolayer (red markers of Figure 3a). From resistance images we observe a large difference in conductivity between the isolated and connected monolayers at low voltage bias, while this difference is reduced at high voltage bias. Indeed, an almost exponential increase of flake conductivity with bias voltage is observed in the case of the isolated monolayer, while this increase is less pronounced when it is connected to thicker flakes. This suggests that the ratio between lateral and direct tunnelling transport contributions in the isolated and connected layers is different and depends on bias voltage. A qualitative description of this phenomenon is that, due to the limited lateral size of the isolated monolayer and its thin thickness, a vertical transport mechanism with direct tunneling of the current through the layer is prevalent on the lateral transport at low polarization bias. On the contrary, in the case of a multilayer, lateral transport contribution becomes more relevant and energetically convenient over direct tunneling and more conduction mechanisms as hopping prevail on it. A connected monolayer appears to behave in a similar way. Since the lateral size of the flake is considerably larger than the one of the isolated flake, lateral conduction becomes more convenient than directly tunnel through the flake.


Page 9 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60





Page 10 of 16

Page 11 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


lateral transport contribution, current is not simply vertically tunneling through the junction, but a lateral transport cannot be neglected. Considering spin diffusion length in MoS2 to be in the order of few 100 nm, spin current that laterally diffuses for larger lengths loses its spin polarization (red arrows) while the one with a diffusion length lower than few 100 nm maintains its spin polarization (green arrows). As consequence, an important part of spin dependent current loses its spin polarization while flowing from bottom to top electrode.

In conclusion, we have reported in this letter a direct investigation of vertical transport mechanisms at a local level on MoS2 flakes deposited over a standard spintronic interface: Co/Al2O3. Understanding vertical transport mechanisms in MoS2 is a key to unlock its efficient use for spin transport in magnetic tunnel junctions. Here Al2O3 acts as a buffer layer to decouple MoS2 from the metallic surface. Thus, it avoids the alteration of the electric properties in the MoS2 monolayer, while providing band alignment for spin and charge injection. By comparing the same monolayer cut in two parts —one isolated and one still connected to thicker flakes— we have demonstrated that lateral transport contribution plays a fundamental role in Co/Al2O3/MoS2 system since charges are found to be distributed over all the connected flakes instead of tunnelling directly through the thinner layers as it would be intuitively expected. As a consequence, no resistance dependence with flake thickness has been observed. This lateral charge distribution is expected to have a drastic effect on spin transport in spintronic devices. Indeed, since spin diffusion length in MoS2 has been reported to be in the order of few hundreds of nanometers25, lateral distribution of spin current would lead to a spin depolarization even for thin MoS2 layers. These findings may explain the limited MR signal observed until now in cross-bar spin valves geometry incorporating MoS2 as tunnel barrier.7,9 They also suggest that having flakes of a size as close as possible to the junction area should remarkably improve the spintronic performance of these devices. This could be valid not only for MoS2-based MTJs 11 ACS Paragon Plus Environment


Page 12 of 16

but also for devices incorporating other semiconductor TMDC 2D materials. These results thus represent a precious guide information for the future fabrication of 2D-based vertical spintronic devices.

ASSOCIATED CONTENT Supporting Information. Details of AFM tapping images on Al2O3 and MoS2 surfaces. Resistance images of the flake collected using PeakForce tapping mode to investigate strain effect.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected] * E-mail: [email protected] * E-mail: [email protected]

ACKNOWLEDGMENT We acknowledge the financial support from the Spanish MINECO (MAT2017-89993-R cofinanced by FEDER and Excellence Unit “María de Maeztu” MDM-2015-0538), the EU (Graphene Flagship (696656), FET-OPEN 2D-INK (664878), COST Action MOLSPIN (CA15128) and Marie Curie project SPIN2D (H2020/2014-659378) to M.G.) and the


Page 13 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Generalidad Valenciana (Prometeo Program/2017/066). S. M.-V. thanks the Spanish MECD for a F.P.U. (Formación de Profesorado Universitario fellowship FPU014/04407). The authors are also grateful to E. Tormos-Feliu, A. Lopez-Muñoz and S. Delprat for their technical support.

REFERENCES (1) Yoffe, A. D. Low-Dimensional Systems: Quantum Size Effects and Electronic Properties of Semiconductor Microcrystallites (Zero-Dimensional Systems) and some Quasi-TwoDimensional Systems Adv. Phys. 1993, 42, 173–266. (2) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A., Coleman; J. N.; Strano, M. S. Electronics and Optoelectronics of Two-Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol. 2012, 7, 699–712. (3) Kuc A.; Zibouche N.; Heine T. Influence of Quantum Confinement on the Electronic Structure of the Transition Metal Sulfide TS2. Phys. Rev. B 2011, 83, 245213. (4) Piquemal-Banci, M.; Galceran, R.; Martin, M.; Godel, F.; Anane, A.; Petroff, F.; Dlubak, B.; Seneor, P. 2D-MTJs: Introducing 2D Materials in Magnetic Tunnel Junctions. J. Phys. D. Appl. Phys. 2017, 50, 203002. (5) Dolui, K.; Narayan, A.; Rungger, I.; Sanvito, S. Efficient Spin Injection and Giant Magnetoresistance in Fe/MoS2/Fe Junctions. Phys. Rev. B 2014, 90, 041401(R). (6) Zhang, H.; Ye, M.; Wang; Y.; Quhe, R.; Pan, Y.; Guo, Y.; Song, Z.; Yang, J.; Guoe, W.; Lu, J. Magnetoresistance in Co/2D MoS2/Co and Ni/2D MoS2/Ni Junctions. Phys. Chem. Chem. Phys. 2016, 18, 16367–16376.



Page 14 of 16

(7) Wang, W.; Narayan, A.; Tang, L; and Dolui, K. Spin-Valve Effect in NiFe/MoS2/NiFe Junctions. Nano Lett. 2015, 15, 5261–5267. (8) Wu, H.; Coileáin, C. Ó.; Abid, M.; Mauit, O.; Syrlybekov, A.; Khalid, A.; Xu, H.; Gatensby, R.; Wang J. J.; Liu H.; Yang, L. Spin-Dependent Transport Properties of Fe3O4/MoS2/Fe3O4 Junctions. Sci. Rep. 2015, 5, 15984. (9) Dankert, A.; Pashaei, P.; Kamalakar, M. V.; Gaur, A. P. S.; Sahoo, S.; Rungger, I.; Narayan, A.; Dolui, K.; Hoque, M. A.; Patel, R. S.; Jong, M. P. de; Katiyar, R. S.; Sanvito, S.; Dash, S. P. Spin-Polarized Tunneling through Chemical Vapor Deposited Multilayer Molybdenum Disulfide. ACS Nano, 2017, 11, 6389–6395. (10) Wong, W. C.; Ng, S. M.; Wong, H. F.; Mak, C. L.; Leung C. W. Spin-Valve Junction with Transfer-Free MoS2 Spacer Prepared by Sputtering. IEEE Trans. Magn. 2017, 53, 1600205. (11) Dlubak, B.; Martin, M.-B.; Deranlot, C.; Servet, B.; Xavier, S.; Mattana, R.; Sprinkle, M.; Berger, C.; De Heer, W. A.; Petroff, F.; Anane, A.; Seneor, P.; Fert, A. Highly Efficient Spin Transport in Epitaxial Graphene on SiC. Nat. Phys. 2012, 8, 557-561. (12) Schmidt, G.; Ferrand, D.; Molenkamp, L. W. Fundamental Obstacle for Electrical Spin Injection From a Ferromagnetic Metal into a Diffusive Semiconductor. Phys. Rev. B 2000, 62, 4790–4793. (13) Fert, A.; George, J.M.; Jaffres, H.; Mattana, R. Semiconductors Between Spin-Polarized Sources and Drains. IEEE Trans. Electron Devices 2007, 54, 921-932. (14) Chai, J. W.; Yang, M.; Callsen, M.; Zhou, J.; Yang, T.; Zhang Z. Tuning Contact Barrier Height between Metals and MoS2 Monolayer through Interface Engineering. Adv. Mater. Interfaces 2017, 4, 1700035.


Page 15 of 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(15) Cui, X.; Shih, E.; Jauregui, L. A.; Chae, S. H.; Kim, Y. D.; Li, B.; Seo, D.; Pistunova, K.; Yin, J.; Park, J.; Choi, H.; Lee, Y. H.; Watanabe, K.; Taniguchi, T.; Kim, P.; Dean, C. R.; Hone, J. C. Low-Temperature Ohmic Contact to Monolayer MoS2 by van der Waals Bonded Co/h-BN Electrodes. Nano Lett. 2017, 17, 4781–4786. (16) Farmanbar, M.; Brocks, G. First-Principles Study of Van der Waals Interactions and Lattice Mismatch at MoS2/Metal Interfaces. Phys. Rev. B 2016, 93, 85304. (17) Dlubak, B.; Seneor, P.; Anane, A.; Barraud, C.; Deranlot, C.; Deneuve, D.; Servet, B.; Mattana, R.; Petroff, F.; Fert, A. Are Al2O3 and MgO Tunnel Barriers Suitable for Spin Injection in Graphene? Appl. Phys. Lett. 2010, 97, 092502. (18) Dlubak, B.; Martin, M.-B.; Deranlot, C.; Bouzehouane, K.; Fusil, S.; Mattana, R.; Petroff, F.; Anane, A.; Seneor, P.; Fert, A. Homogeneous Pinhole Free 1 nm Al2O3 Tunnel Barriers on Graphene. Appl. Phys. Lett. 2012, 101, 203104. (19) Bernand-Mantel, A.; Seneor, P.; Bouzehouane, K.; Fusil, S.; Deranlot, C.; Petroff, F.; Fert, A. Anisotropic Magneto-Coulomb Effects and Magnetic Single-Electron-Transistor Action in a Single Nanoparticle. Nat. Phys. 2009, 5, 920–924. (20) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666–669. (21) Li, H.; Zhang, Q.; Yap, C. C. R.; Tay, B. K.; Edwin, T. H. T.; Olivier, A.; Baillargeat, D. From Bulk to Monolayer MoS2: Evolution of Raman Scattering. Adv. Funct. Mater. 2012, 22, 1385–1390.




Page 16 of 16

A Local Study of the Transport Mechanisms in MoS2 Layers for

Recommend Documents