2D Nanosheet Device Integration

August 15, 2022 – Defects decide how 2D materials behave in devices, but device measurements cannot see them. Researchers from Ulm University, Friedrich-Alexander-Universität Erlangen-Nürnberg and the Hebrew University of Jerusalem have developed a way to image and modify free-standing 2D flakes in the SALVE microscope and then transfer the same flakes onto any substrate. Built into tunneling devices and probed by photoluminescence, the electron-beam-made defects show clear electrical and optical signatures.

Every crystal has its flaws, and in materials only a few atoms thick, single flaws matter. For electronic and optical applications of two-dimensional (2D) materials and their stacked heterostructures, it is important to know where the crystallographic defects sit, how many there are and what they look like at the atomic scale. Researchers from Ulm University, Friedrich-Alexander-Universität Erlangen-Nürnberg and the Hebrew University of Jerusalem have now closed a gap between two worlds: they image and modify free-standing 2D flakes in the transmission electron microscope (TEM), then lift the very same flakes off the TEM grid, place them on any substrate and build them into devices for electronic and optical sensing. Proof-of-principle experiments show that the defects written by the electron beam leave clear signatures in electrical tunneling measurements and in photoluminescence. The transfer works reliably for single and multilayer MoS2, MoSe2, WSe2, MoTe2, hexagonal boron nitride (hBN) and graphene, and it may also open a route to defect-based sensors and single-photon sources.

2D materials show electrical and optical properties that differ from those of their bulk counterparts[1],[2]. They range from insulators through semiconductors to metals, with exceptional phenomena such as superconductivity or charge density waves[3],[4]. Group-6 transition metal dichalcogenides (TMDs) such as MoS2 change from an indirect to a direct band gap semiconductor when thinned from the bulk to a single layer, which enables new optoelectronic applications[3],[5]. Stacked vertically into heterostructures, these layers can be combined to tailor the properties of the stack[1], promising for nanodevices such as tunneling transistors[6], light-emitting diodes or photodetectors[7]. Besides the number of layers and doping, atomic defects are a promising control lever for tuning the electrical and optical properties of such devices[8].

Electronic transport through tunneling heterostructures can be very sensitive to single defects. In these devices, tunneling is mediated by quantum dots, tiny regions that trap electrons at discrete energies, which are associated with defects in the barrier[9],[10]. Such quantum dots have atomic dimensions and are regarded as highly sensitive electrometers and promising platforms for storing quantum information[11]. On the optical side, Raman and photoluminescence (PL) spectroscopy are widely used to measure how defects affect semiconducting 2D materials[12]. All these techniques, however, share one drawback: they do not reveal the positions and structures of the atomic defects that cause a particular property.

Atomic-resolution TEM fills this gap, because in 2D materials the electron beam can do two jobs at once: it images the atoms around a defect, and it creates or manipulates defects on purpose[13]. Accelerating voltages below the material-specific knock-on threshold, the energy at which an electron can knock an atom out of the lattice in a direct collision, drastically reduce unwanted beam damage. Choosing the voltage freely within the range of typically 20 to 80 kV for 2D materials makes it possible to balance imaging of the pristine structure against dose-controlled defect engineering[14],[13]. The SALVE microscope, corrected for chromatic and spherical aberration and reaching atomic resolution between 20 and 80 kV, is well suited for this purpose[15]. What was missing was a way to bring the flake back out of the microscope. TEM and the other probes usually look at entirely different specimen areas, which makes it very hard to link a structure created in the microscope to a measured property. For TEM, monolayer and few-layer flakes are suspended over holey carbon films, but this support is incompatible with most other measurements, and above all with building the flake into a stacked heterostructure device. To be contacted electrically or stacked into a van der Waals heterostructure, the flake first has to be released from the TEM grid.

2D transfer method
The team developed a TEM sample preparation that transfers a 2D flake onto a holey carbon TEM grid and releases it again after the TEM work. It has so far been used for single and few-layer MoS2, MoSe2, WSe2, MoTe2, hBN and graphene (Fig. 1). The researchers call the placement of the flake on the grid the “forward” transfer and its later release the “reverse” transfer. The forward transfer before the TEM experiment follows a standard technique with spin-coated polyvinyl alcohol (PVA) on SiO2/Si[16], extended by one extra step: a thin copper layer is sputtered onto the grid (Fig. 2). This oxygen-sensitive copper layer is what later lets the flake detach from the grid. The flake is then imaged and modified in the TEM, after which the reverse transfer follows. This step is the core of the work; to the authors' knowledge it was reported here for the first time.

▸ Forward transfer with a sacrificial copper layer

The reverse transfer uses polystyrene as the transfer medium (Fig. 3a). First, a Si/SiO2 wafer is spin-coated with polystyrene (2 wt% in toluene), which adheres strongly to both grid and flakes. The TEM grid with the electron-exposed flake is placed on it, flake facing down (step 1). A drop of IPA on the grid (step 2) bonds grid and substrate firmly as it evaporates, so that the flake is in close contact with the sputtered copper on its top surface and with the polystyrene on its bottom surface (step 3). To free the flake from the grid, a drop of ammonium peroxodisulfate (APS) solution (0.3 mol/l) is placed on the grid at the position of the flake to etch away the copper. Under the optical microscope, the APS can be seen flowing under the grid as the carbon support film separates from the substrate and the flake. The grid is then lifted off with tweezers (step 4), the remaining APS is rinsed off with ultrapure water, and the sample is dried in air (step 5). Figs. 3b and 3c show the grid before and during the process, and Fig. 3d shows a flake transferred back to polystyrene. Once on a substrate, the flake can be built into a van der Waals stack, and exactly the areas that were imaged or modified in the TEM can be studied by electronic transport, optical spectroscopy or other techniques. The method should work for all 2D materials that tolerate acids and water.

▸ Why the reverse transfer needs polystyrene

Because of these surface-tension requirements, the reverse transfer works only onto polystyrene. To measure the flake or stack it into a heterostructure, it is therefore moved once more, now with the polystyrene film itself as the carrier: the film is floated off the wafer, picked up on a membrane holder and pressed onto the target substrate, where the polystyrene is finally dissolved. A small trick makes this step precise to about 1 μm, enough to place the electron-exposed spots of the tunneling device, each 1.2 μm across, onto a gold track only 2.5 μm wide.

▸ From polystyrene to the device

Exposure to the electron beam makes this step much harder. The beam strengthens the adhesion between flake and support film, because it modifies hydrocarbon contamination from air or solvents trapped between the two[18]. Simpler approaches failed for exactly this reason: without the copper layer, electron-irradiated flakes stayed on the grid, while unexposed flakes on the same grid transferred without trouble. When water was used instead of IPA to bring grid and substrate together, one flake even tore apart along the edge of its exposed area: the exposed part remained on the grid, the rest went to the substrate. The effect is probably related to electron-beam-induced deposition, in which inelastically scattered electrons create secondary electrons that break up hydrocarbon molecules; the resulting radicals react with the surfaces of flake and support film and glue them together. After the release, a carbon contamination layer about 3 nm thick, shaped like the holey support film, stays on the flake. It was found mainly at the exposed sites, which supports this explanation. Energy-dispersive X-ray spectroscopy showed that this layer consists mainly of carbon and oxygen, with no detectable copper residues from the sacrificial layer. Beam-exposed flakes therefore need the sacrificial copper layer, whereas unexposed flakes can be lifted off the grid with simpler methods.

Microscopic features in the flake survive the whole procedure, as a marker hole in a MoS2 flake shows (Fig. 4). The hole, about 50 nm across and with a distinct shape, was drilled with a converged 80 keV electron beam (Fig. 4a). The team then transferred the flake to a substrate, picked it up again with a new TEM grid and imaged the same spot once more (Fig. 4b). The overall shape of the hole was preserved through all transfer steps.

Experiments to prove the principle
Two proof-of-principle experiments showed what the method can do. The first asked how beam-induced changes in a 2D material show up in electronic transport. A six-layer MoS2 flake on a TEM grid (Fig. 5a) was exposed to different electron doses in the SALVE microscope at 80 kV and then transferred to a polystyrene substrate. Using polystyrene again as transfer medium, the exposed areas were placed precisely on a prepatterned gold electrode, and Ti/Au top contacts were added by standard electron-beam lithography (Fig. 5b). The contacts covered both exposed and unexposed areas, the latter serving as controls.

The devices were measured with a voltage bias in a two-terminal geometry at T = 4 K (Fig. 5c). Naively, a pristine MoS2 flake would act as a tunnel barrier. Defects in the tunnel barrier would appear as discrete jumps in the DC current, caused by resonant tunneling through the discrete energy levels of defect-bound quantum dots[9]. The data show two main features. The exposed regions indeed carry the telltale current steps of defect-assisted transport. In addition, the unexposed regions conduct much better. The first result matches the expectation that TEM exposure creates deep defects; the second suggests that the beam also changes the band alignment, possibly by doping the TMD flake and lowering the tunnel barrier formed by the thin TMD layer.

The second experiment probed beam-induced defects with PL. Again, a MoS2 flake was exfoliated, transferred to a TEM grid and exposed to an 80 keV electron beam at different doses (Fig. 6a), then transferred to polystyrene (Fig. 6b), where the PL of the exposed monolayer regions was measured (Fig. 6c). Two peaks stand out. The A0 peak at about 1.85 eV comes from the recombination of free excitons, electron–hole pairs, at the direct band gap, while the AB peak at about 1.7 eV comes from the radiative recombination of bound excitons. These are trapped at defects and therefore lie lower in energy than the free excitons; the B exciton appears only weakly. With increasing irradiation, and thus increasing defect density, the A0 peak weakens while the defect-bound AB peak grows, as also described in the literature[8].

The study thus demonstrates a preparation technique that reliably moves electron-irradiated 2D flakes from a TEM grid to any substrate, so that TEM can be combined with complementary characterization. The team built devices from flakes whose defect structure had been created in the TEM and related their electrical and optical signatures to that structure. In the authors' words, “relating the atomic defect structure determined by the electron probe in the TEM to properties determined with very different probes opens up new possibilities for understanding and improving the properties of nanomaterials on an atomic scale.” The transfer could also help to assemble defect-based sensors and photon sources.

▸ Exposure and measurement conditions

Resource:
Quincke, M., Lehnert, T., Keren, I., Badlyan, N. M., Port, F., Goncalves, M., Mohn, M. J., Maultzsch, J., Steinberg, H., Kaiser, U. (2022).
Transmission-Electron-Microscopy-Generated Atomic Defects in Two-Dimensional Nanosheets and Their Integration in Devices for Electronic and Optical Sensing.
ACS Appl. Nano Mater. 5(8), 11429–11436.
https://doi.org/10.1021/acsanm.2c02491

Download this news article as PDF – for reading on paper, taking notes or filing.

References
  1. Novoselov, K. S., Mishchenko, A., Carvalho, A., Castro Neto, A. H. (2016). 2D materials and van der Waals heterostructures. Science, 353, aac9439. https://doi.org/10.1126/science.aac9439

  2. Wang, Q. H., Kalantar-Zadeh, K., Kis, A., Coleman, J. N., Strano, M. S. (2012). Electronics and Optoelectronics of Two-Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol., 7, 699–712. https://doi.org/10.1038/nnano.2012.193

  3. Chhowalla, M., Shin, H. S., Eda, G., Li, L.-J., Loh, K. P., Zhang, H. (2013). The Chemistry of Two-Dimensional Layered Transition Metal Dichalcogenide Nanosheets. Nat. Chem., 5, 263–275. https://doi.org/10.1038/nchem.1589

  4. Börner, P. C., Kinyanjui, M. K., Björkman, T., Lehnert, T., Krasheninnikov, A. V., Kaiser, U. (2018). Observation of Charge Density Waves in Free-Standing 1T-TaSe2 Monolayers by Transmission Electron Microscopy. Appl. Phys. Lett., 113, 173103. https://doi.org/10.1063/1.5052722

  5. Mak, K. F., Lee, C., Hone, J., Shan, J., Heinz, T. F. (2010). Atomically Thin MoS2: A New Direct-Gap Semiconductor. Phys. Rev. Lett., 105, 136805. https://doi.org/10.1103/PhysRevLett.105.136805

  6. Britnell, L., Gorbachev, R. V., Jalil, R., Belle, B. D., Schedin, F., Mishchenko, A., Georgiou, T., Katsnelson, M. I., Eaves, L., Morozov, S. V., et al. (2012). Field-Effect Tunneling Transistor Based on Vertical Graphene Heterostructures. Science, 335, 947–950. https://doi.org/10.1126/science.1218461

  7. Najafidehaghani, E., Gan, Z., George, A., Lehnert, T., Ngo, G. Q., Neumann, C., Bucher, T., Staude, I., Kaiser, D., Vogl, T., et al. (2021). 1D p–n Junction Electronic and Optoelectronic Devices From Transition Metal Dichalcogenide Lateral Heterostructures Grown by One‐Pot Chemical Vapor Deposition Synthesis. Adv. Funct. Mater., 31, 2101086. https://doi.org/10.1002/adfm.202101086

  8. Tongay, S., Suh, J., Ataca, C., Fan, W., Luce, A., Kang, J. S., Liu, J., Ko, C., Raghunathanan, R., Zhou, J., et al. (2013). Defects Activated Photoluminescence in Two-Dimensional Semiconductors: Interplay Between Bound, Charged and Free Excitons. Sci. Rep., 3, 2657. https://doi.org/10.1038/srep02657

  9. Keren, I., Dvir, T., Zalic, A., Iluz, A., LeBoeuf, D., Watanabe, K., Taniguchi, T., Steinberg, H. (2020). Quantum-Dot Assisted Spectroscopy of Degeneracy-Lifted Landau Levels in Graphene. Nat. Commun., 11, 3408. https://doi.org/10.1038/s41467-020-17225-1

  10. Dvir, T., Aprili, M., Quay, C. H. L., Steinberg, H. (2019). Zeeman Tunability of Andreev Bound States in van der Waals Tunnel Barriers. Phys. Rev. Lett., 123, 217003. https://doi.org/10.1103/PhysRevLett.123.217003

  11. Elzerman, J. M., Hanson, R., Willems van Beveren, L. H., Witkamp, B., Vandersypen, L. M. K., Kouwenhoven, L. P. (2004). Single-Shot Read-Out of an Individual Electron Spin in a Quantum Dot. Nature, 430, 431–435. https://doi.org/10.1038/nature02693

  12. Amani, M., Lien, D.-H., Kiriya, D., Xiao, J., Azcatl, A., Noh, J., Madhvapathy, S. R., Addou, R., KC, S., Dubey, M., et al. (2015). Near-Unity Photoluminescence Quantum Yield in MoS2. Science, 350, 1065–1068. https://doi.org/10.1126/science.aad2114

  13. Zhao, X., Kotakoski, J., Meyer, J. C., Sutter, E., Sutter, P., Krasheninnikov, A. V., Kaiser, U., Zhou, W. (2017). Engineering and Modifying Two-Dimensional Materials by Electron Beams. MRS Bull., 42, 667–676. https://doi.org/10.1557/mrs.2017.184

  14. Kretschmer, S., Lehnert, T., Kaiser, U., Krasheninnikov, A. V. (2020). Formation of Defects in Two-Dimensional MoS2 in the Transmission Electron Microscope at Electron Energies Below the Knock-on Threshold: The Role of Electronic Excitations. Nano Lett., 20, 2865–2870. https://doi.org/10.1021/acs.nanolett.0c00670

  15. Linck, M., Hartel, P., Uhlemann, S., Kahl, F., Müller, H., Zach, J., Haider, M., Niestadt, M., Bischoff, M., Biskupek, J., et al. (2016). Chromatic Aberration Correction for Atomic Resolution TEM Imaging From 20 to 80 kV. Phys. Rev. Lett., 117, 076101. https://doi.org/10.1103/PhysRevLett.117.076101

  16. Köster, J., Liang, B., Storm, A., Kaiser, U. (2021). Polymer-Assisted TEM Specimen Preparation Method for Oxidation-Sensitive 2D Materials. Nanotechnology, 32, 075704. https://doi.org/10.1088/1361-6528/abc49e

  17. Gurarslan, A., Yu, Y., Su, L., Yu, Y., Suarez, F., Yao, S., Zhu, Y., Ozturk, M., Zhang, Y., Cao, L. (2014). Surface-Energy-Assisted Perfect Transfer of Centimeter-Scale Monolayer and Few-Layer MoS2 Films Onto Arbitrary Substrates. ACS Nano, 8, 11522–11528. https://doi.org/10.1021/nn5057673

  18. Westenfelder, B., Meyer, J. C., Biskupek, J., Kurasch, S., Scholz, F., Krill, C. E., Kaiser, U. (2011). Transformations of Carbon Adsorbates on Graphene Substrates Under Extreme Heat. Nano Lett., 11, 5123–5127. https://doi.org/10.1021/nl203224z