•The etching of two-dimensional TMDs is driven by planar SPPs•Desired layers are achieved by tuning the light power and incident direction•Strong plasmonic coupling at Au/MoS2 interface induces the oxidation dissolution Plasmonic metal/transition metal dichalcogenide (TMD) heterostructures have attracted a considerable interest owing to plasmon-induced interfacial charge transfer, which has emerged as an important topic in photocatalysis and photovoltaics. In this work, we found an exotic interaction of TMDs with surface plasmon polaritons, which can be exploited as a general method to fabricate TMDs with controllable layers and lateral sizes. Taking MoS2 as an example, the strong plasmonic coupling at the Au/MoS2 interface can accumulate holes in the valence band of MoS2 and weaken its interlayer interaction. Meanwhile, the plasmonic hot electrons would transfer on MoS2 surface to produce the oxidizing H2O2 and etch top layer MoS2 in aqueous media. By controlling the light power, monolayer, bilayer, and trilayer MoS2 can be achieved with a pristine lateral size. Our findings offer deep insights into the plasmonic coupling at metal-TMDs interfaces, opening up a new avenue for controlled fabrication of TMDs. Plasmon-induced charge transfer has drawn considerable interests and spurred rapid developments of photovoltaics and photocatalysis. Various plasmonic metal/transition metal dichalcogenide (TMD) heterostructures have been developed to promote charge separation. For plasmon-induced catalysis, the transformation of TMDs is rare, and the erosion of TMDs has not yet been reported. Here, we report the etching of two-dimensional TMDs by planar surface plasmon polaritons (SPPs). We found that SPPs can etch various TMDs into desired layers and lateral size through controlling the light power and incident direction. In aqueous media, the strong plasmonic coupling at Au/MoS2 interface generates holes in the valence band of MoS2 that can weaken its interlayer interaction. As a synergistic effect, the plasmonic hot electrons produce oxidizing H2O2 to dissolve MoS2 from the top layers. Our results provide a new perspective for understanding the plasmonic coupling at metal-TMD interfaces and a reliable route toward fabricating well-defined TMDs. Plasmon-induced charge transfer has drawn considerable interests and spurred rapid developments of photovoltaics and photocatalysis. Various plasmonic metal/transition metal dichalcogenide (TMD) heterostructures have been developed to promote charge separation. For plasmon-induced catalysis, the transformation of TMDs is rare, and the erosion of TMDs has not yet been reported. Here, we report the etching of two-dimensional TMDs by planar surface plasmon polaritons (SPPs). We found that SPPs can etch various TMDs into desired layers and lateral size through controlling the light power and incident direction. In aqueous media, the strong plasmonic coupling at Au/MoS2 interface generates holes in the valence band of MoS2 that can weaken its interlayer interaction. As a synergistic effect, the plasmonic hot electrons produce oxidizing H2O2 to dissolve MoS2 from the top layers. Our results provide a new perspective for understanding the plasmonic coupling at metal-TMD interfaces and a reliable route toward fabricating well-defined TMDs. Surface plasmons possess light-trapping and electromagnetic-field-concentrating properties and open up a wide range of applications in solar energy conversion,1Atwater H.A. Polman A. Plasmonics for improved photovoltaic devices.Nat. Mater. 2010; 9: 205-213Crossref PubMed Scopus (7053) Google Scholar, 2Brongersma M.L. Halas N.J. Nordlander P. Plasmon-induced hot carrier science and technology.Nat. Nanotechnol. 2015; 10: 25-34Crossref PubMed Scopus (1943) Google Scholar, 3Clavero C. Plasmon-induced hot-electron generation at nanoparticle/metal-oxide interfaces for photovoltaic and photocatalytic devices.Nat. Photonics. 2014; 8: 95-103Crossref Scopus (1856) Google Scholar, 4Knight M.W. Sobhani H. Nordlander P. Halas N.J. Photodetection with active optical antennas.Science. 2011; 332: 702-704Crossref PubMed Scopus (1565) Google Scholar, 5Linic S. Christopher P. Ingram D.B. Plasmonic-metal nanostructures for efficient conversion of solar to chemical energy.Nat. Mater. 2011; 10: 911-921Crossref PubMed Scopus (3589) Google Scholar, 6Mubeen S. Lee J. Singh N. Krämer S. Stucky G.D. Moskovits M. An autonomous photosynthetic device in which all charge carriers derive from surface plasmons.Nat. Nanotechnol. 2013; 8: 247-251Crossref PubMed Scopus (883) Google Scholar, 7Thomann I. Pinaud B.A. Chen Z.B. Clemens B.M. Jaramillo T.F. Brongersma M.L. Plasmon enhanced solar-to-fuel energy conversion.Nano Lett. 2011; 11: 3440-3446Crossref PubMed Scopus (427) Google Scholar metasurfaces,8Li G. Zhang S. Zentgraf T. Nonlinear photonic metasurfaces.Nat. Rev. Mater. 2017; 2: 17010Crossref Scopus (419) Google Scholar,9Meinzer N. Barnes W.L. Hooper I.R. Plasmonic meta-atoms and metasurfaces.Nat. Photonics. 2014; 8: 889-898Crossref Scopus (699) Google Scholar and ultrasensitive sensing.10Anker J.N. Hall W.P. Lyandres O. Shah N.C. Zhao J. Van Duyne R.P. Biosensing with plasmonic nanosensors.Nat. Mater. 2008; 7: 442-453Crossref PubMed Scopus (5643) Google Scholar, 11Mayer K.M. Hafner J.H. Localized surface plasmon resonance sensors.Chem. Rev. 2011; 111: 3828-3857Crossref PubMed Scopus (2826) Google Scholar, 12Zhou X.L. Yang Y. Wang S. Liu X.W. Surface plasmon resonance microscopy: from single-molecule sensing to single-cell imaging.Angew. Chem. Int. Ed. Engl. 2020; 59: 1776-1785Crossref PubMed Scopus (47) Google Scholar Localized surface plasmon resonance of metal nanostructures can occur in the metal surface in which confined free electrons oscillate upon the incident radiation, generating an intense and highly localized optical fields.2Brongersma M.L. Halas N.J. Nordlander P. Plasmon-induced hot carrier science and technology.Nat. Nanotechnol. 2015; 10: 25-34Crossref PubMed Scopus (1943) Google Scholar,13Linic S. Aslam U. Boerigter C. Morabito M. Photochemical transformations on plasmonic metal nanoparticles.Nat. Mater. 2015; 14: 567-576Crossref PubMed Scopus (996) Google Scholar An important potential application of localized surface plasmon resonance is driving chemical reactions of molecules adsorbed on metal nanosctructures.14Kazuma E. Jung J. Ueba H. Trenary M. Kim Y. Real-space and real-time observation of a plasmon-induced chemical reaction of a single molecule.Science. 2018; 360: 521-526Crossref PubMed Scopus (154) Google Scholar Electron-hole pairs could be generated in the metal nanostructures. Therefore, great efforts have been made to promote charge separation by fabricating metal-semiconductor Schottky junctions.3Clavero C. Plasmon-induced hot-electron generation at nanoparticle/metal-oxide interfaces for photovoltaic and photocatalytic devices.Nat. Photonics. 2014; 8: 95-103Crossref Scopus (1856) Google Scholar,15Tian Y. Tatsuma T. Mechanisms and applications of plasmon-induced charge separation at TiO2 films loaded with gold nanoparticles.J. Am. Chem. Soc. 2005; 127: 7632-7637Crossref PubMed Scopus (1685) Google Scholar,16Wu K. Chen J. McBride J.R. Lian T. CHARGE TRANSFER. Efficient hot-electron transfer by a plasmon-induced interfacial charge-transfer transition.Science. 2015; 349: 632-635Crossref PubMed Scopus (713) Google Scholar A variety of plasmonic metal/transition metal dichalcogenide (TMDs) heterostructures, such as Au nanorods/MoS2, Ag nanowires/MoS2, Au nanotriangles/WS2, and WSe2/Au trench, were developed with enhanced light-matter interactions to boost the efficiency of photocatalysis and photodetectors.17Sriram P. Manikandan A. Chuang F.C. Chueh Y.L. Hybridizing plasmonic materials with 2D-transition metal dichalcogenides toward functional applications.Small. 2020; 16e1904271Crossref PubMed Scopus (58) Google Scholar, 18Shi Y. Wang J. Wang C. Zhai T.T. Bao W.J. Xu J.J. Xia X.H. Chen H.Y. Hot electron of Au nanorods activates the electrocatalysis of hydrogen evolution on MoS2 nanosheets.J. Am. Chem. Soc. 2015; 137: 7365-7370Crossref PubMed Scopus (442) Google Scholar, 19Li Y. DiStefano J.G. Murthy A.A. Cain J.D. Hanson E.D. Li Q.Q. Castro F.C. Chen X.Q. Dravid V.P. Superior plasmonic photodetectors based on [email protected]2 core-shell heterostructures.ACS Nano. 2017; 11: 10321-10329Crossref PubMed Scopus (99) Google Scholar, 20Lee B. Park J. Han G.H. Ee H.S. Naylor C.H. Liu W.J. Johnson A.T.C. Agarwal R. Fano resonance and spectrally modified photoluminescence enhancement in monolayer MoS2 integrated with plasmonic nanoantenna array.Nano Lett. 2015; 15: 3646-3653Crossref PubMed Scopus (213) Google Scholar, 21Lee H.S. Luong D.H. Kim M.S. Jin Y. Kim H. Yun S. Lee Y.H. Reconfigurable exciton-plasmon interconversion for nanophotonic circuits.Nat. Commun. 2016; 7: 13663Crossref PubMed Scopus (34) Google Scholar, 22Wang M.S. Krasnok A. Zhang T.Y. Scarabelli L. Liu H. Wu Z.L. Liz-Marzán L.M. Terrones M. Alù A. Zheng Y.B. Tunable Fano resonance and plasmon-exciton coupling in single Au nanotriangles on monolayer WS2 at room temperature.Adv. Mater. 2018; 30e1705779Crossref PubMed Scopus (70) Google Scholar, 23Wang Z. Dong Z. Gu Y. Chang Y.-H. Zhang L. Li L.-J. Zhao W. Eda G. Zhang W. Grinblat G. et al.Giant photoluminescence enhancement in tungsten-diselenide-gold plasmonic hybrid structures.Nat. Commun. 2016; 7: 11283Crossref PubMed Scopus (231) Google Scholar, 24Zheng D. Zhang S.P. Deng Q. Kang M. Nordlander P. Xu H.X. Manipulating coherent plasmon-exciton interaction in a single silver nanorod on monolayer WSe2.Nano Lett. 2017; 17: 3809-3814Crossref PubMed Scopus (195) Google Scholar All these two-dimensional (2D) materials exhibited stable layered structures. The transformation of 2D TMDs during the plasmon-induced catalysis is very rare,25Kang Y. Najmaei S. Liu Z. Bao Y. Wang Y. Zhu X. Halas N.J. Nordlander P. Ajayan P.M. Lou J. Fang Z. Plasmonic hot electron induced structural phase transition in a MoS2 monolayer.Adv. Mater. 2014; 26: 6467-6471Crossref PubMed Scopus (442) Google Scholar,26Kang Y. Gong Y. Hu Z. Li Z. Qiu Z. Zhu X. Ajayan P.M. Fang Z. Plasmonic hot electron enhanced MoS2 photocatalysis in hydrogen evolution.Nanoscale. 2015; 7: 4482-4488Crossref PubMed Google Scholar whereas the erosion of TMDs has not yet been reported. The exotic layer-dependent properties of TMDs have aroused great interests in the fabrication of 2D TMDs with desired layers and lateral size. Currently, the common routes to produce 2D TMDs often suffer from low efficiency, random layer distribution, and unsatisfied lateral sizes. Mechanical exfoliation can get desired layers of TMDs, but the yield is low.27Li H. Wu J.M. Yin Z.Y. Zhang H. Preparation and applications of mechanically exfoliated single-layer and multilayer MoS₂ and WSe₂ nanosheets.Acc. Chem. Res. 2014; 47: 1067-1075Crossref PubMed Scopus (1089) Google Scholar Solution-based exfoliation often obtains random layers of TMDs and is afflicted by poor crystal crystalline owing to the introduction of defects or phase transitions.28Coleman J.N. Lotya M. O'Neill A. Bergin S.D. King P.J. Khan U. Young K. Gaucher A. De S. Smith R.J. et al.Two-dimensional nanosheets produced by liquid exfoliation of layered materials.Science. 2011; 331: 568-571Crossref PubMed Scopus (5309) Google Scholar, 29Zeng Z.Y. Yin Z.Y. Huang X. Li H. He Q.Y. Lu G. Boey F. Zhang H. Single-layer semiconducting nanosheets: high-yield preparation and device fabrication.Angew. Chem. Int. Ed. Engl. 2011; 50: 11093-11097Crossref PubMed Scopus (1388) Google Scholar, 30Lin Z.Y. Liu Y. Halim U. Ding M.N. Liu Y.Y. Wang Y.L. Jia C.C. Chen P. Duan X.D. Wang C. et al.Solution-processable 2D semiconductors for high-performance large-area electronics.Nature. 2018; 562: 254-258Crossref PubMed Scopus (402) Google Scholar Furthermore, the atomic layer etching has been used to control the MoS2 layers,31Kim K.S. Ji Y.J. Kim K.H. Choi S. Kang D.H. Heo K. Cho S. Yim S. Lee S. Park J.-H. Jung Y.S. Ultrasensitive MoS2 photodetector by serial nano-bridge multi-heterojunction.Nat. Commun. 2019; 10: 4701Crossref PubMed Scopus (28) Google Scholar but along with the experimental complexity and harsh conditions. Although chemical vapor deposition has been a prevalent method to prepare 2D TMDs and can obtain different layers, it is still plagued by the defect and polycrystalline with grain boundaries.32Shi Y.M. Li H.N. Li L.-J. Recent advances in controlled synthesis of two-dimensional transition metal dichalcogenides via vapour deposition techniques.Chem. Soc. Rev. 2015; 44: 2744-2756Crossref PubMed Google Scholar,33Yang P. Zhang Z. Sun M. Lin F. Cheng T. Shi J. Xie C. Shi Y. Jiang S. Huan Y. et al.Thickness tunable wedding-cake-like MoS2 flakes for high-performance optoelectronics.ACS Nano. 2019; 13: 3649-3658Crossref PubMed Scopus (56) Google Scholar Therefore, it remains a great challenge to achieve TMD flakes with precise control of the layer number and lateral sizes. Here, we report the etching of TMDs by planar surface plasmon polaritons (SPPs). Unlike the localized surface plasmon resonance confined in metal nanostructures, planar SPPs could be excited with a weak illumination and propagates along the planar metal interface.34Zhang J.X. Zhang L.D. Xu W. Surface plasmon polaritons: physics and applications.J. Phys. D: Appl. Phys. 2012; 45: 113001Crossref Scopus (273) Google Scholar We found the SPPs catalyzed the etching of TMDs, which was highly dependent on the SPP propagating directions. By using this capability, we demonstrated a general approach for the precise fabrication of highly uniform and crystalline TMDs with controllable layer number and lateral sizes. Taking MoS2 as an example, desired layers of MoS2 (monolayers, bilayers, and trilayers) can be controllably achieved by simply tuning the power density of exciting light. To excite SPPs, we used a home-built inverted optical microscope with a high numerical aperture objective.35Huang B. Yu F. Zare R.N. Surface plasmon resonance imaging using a high numerical aperture microscope objective.Anal. Chem. 2007; 79: 2979-2983Crossref PubMed Scopus (183) Google Scholar,36Liu X.W. Yang Y.Z. Wang W. Wang S.P. Gao M. Wu J. Tao N.J. Plasmonic-based electrochemical impedance imaging of electrical activities in single cells.Angew. Chem. Int. Ed. Engl. 2017; 56: 8855-8859Crossref PubMed Scopus (42) Google Scholar Through phase matching, red light with wavelength of 660 nm was irradiated on a gold surface at a certain incident angle, i.e., the resonant angle (Figure S1). The excited SPPs resulting from the coupling of surface electromagnetic polaritons to oscillating free electrons propagated along the Au/water interface, generating an evanescent field with a penetration depth of approximately 200 nm. We employed this bifunctional setup for triggering and monitoring layer-controlled etching of 2D TMD materials (Figure 1). In a typical procedure, we mechanically exfoliated multilayered MoS2 nanoflakes from bulk crystals and transferred them onto a gold-coated glass coverslip (Figure S2). Surprisingly, we observed an etching phenomenon in multilayered MoS2 nanoflakes upon illumination at the resonant angle under the microscope in deionized (DI) water. Optical images of a MoS2 nanoflake before and after etching for 40 min clearly show the morphological changes when we used a low red-light power density (1.7 mW·mm−2) for exciting SPPs (Figures 2A and 2B ). A video of the etching process, presented in Video S1, reveals that the top layers of a MoS2 nanoflake were gradually etched away with increasing illumination time. Finally, a thinner MoS2 nanoflake with the same lateral size as before the experiment remained on the Au substrate. Because the beam diameter (ca. 0.9 mm) is far beyond the SPP propagation length of ca. 6.5 μm, it can be expected that all the exfoliated TMD flakes within the beam diameter can be etched. https://www.cell.com/cms/asset/0fd08fc8-9779-49d6-8317-066d4fa425bc/mmc2.mp4Loading ... Download .mp4 (0.52 MB) Help with .mp4 files Video S1. A movie showing the layer-controlled etching process of a MoS2 nanoflake To quantify the layer number of the MoS2 nanoflake remaining on the Au substrate, we measured the thickness of the MoS2 nanoflake using atomic force microscopy (AFM). The height of the residual MoS2 nanoflake was approximately 2.6 nm (Figure 2C), which corresponds to trilayer MoS2. Furthermore, we also employed Raman spectroscopy to characterize the MoS2 nanoflake before and after etching. The Raman spectrum is sensitive to the layer number, by reading out a frequency difference of approximately 20, 22, and 23 cm−1 for monolayer, bilayer, and trilayer MoS2 nanoflakes, respectively.37Lee C. Yan H. Brus L.E. Heinz T.F. Hone J. Ryu S. Anomalous lattice vibrations of single- and few-layer MoS2.ACS Nano. 2010; 4: 2695-2700Crossref PubMed Scopus (3445) Google Scholar Raman spectra of pristine and residual MoS2 nanoflake exhibit two typical characteristic vibrational modes, E2g1 (in-plane) and A1g (Figure 2D), respectively. The MoS2 nanoflake after etching shows obvious smaller frequency difference of 23.1 cm−1 between these two peaks than that of pristine MoS2 nanoflake (24.9 cm−1), suggesting that the residual MoS2 nanoflakes can be identified to three layers of MoS2, consistent with the abovementioned AFM analysis. We further increased the illumination power density and found that the layer number of residual MoS2 could be precisely regulated to the bilayer and monolayer. When the power density of the red light was increased to 3.4 mW·mm−2, the residual MoS2 was more transparent and thinner after etching (Figures 2E and 2F). After etching, the AFM height of ~1.83 nm corresponded to bilayer MoS2 (Figure 2G). This layer identification was corroborated by the Raman spectra with a smaller frequency difference of 21.4 cm−1 (Figure 2H). As we continued to increase the power density to 6.8 mW·mm−2, monolayer MoS2 could be preserved on the Au substrate (Figures 2I and 2J). Both the AFM height of 0.87 nm and the frequency difference of 20.2 cm−1 validated the monolayer structure of MoS2 (Figures 2K and 2L). More examples of trilayer, bilayer, and monolayer MoS2 after SPPs-driven etching (SPPE) displayed similar frequency differences (see more details in Figure S3), implying that this approach could be a robust method for fabricating MoS2 with controlled layers. Notably, this method can precisely control the layer number and retain the lateral size of pristine MoS2 flakes, which is not feasible with other mechanical-force-driven exfoliation or bottom-up growth methods, providing a reliable preparation approach to meet the layer number and size demands of TMD nanoflakes for device fabrication. We assumed that the underlying driving force of etching in our work originated from the role of SPPs. To verify this hypothesis, we conducted control experiments to demonstrate that the SPPs was responsible for the etching of MoS2 nanoflakes. First, we changed the p-polarized light to s-polarized light and recorded unaffected MoS2 nanoflakes (Figures S4A–S4C). For metal films, the p-polarized light commonly excites the propagating SPPs much more efficiently than s-polarized light, with a larger amplitude under the surface normal electric field.38Wang L.M. Zhang L. Seideman T. Petek H. Dynamics of coupled plasmon polariton wave packets excited at a subwavelength slit in optically thin metal films.Phys. Rev. B. 2012; 86: 165408Crossref Scopus (21) Google Scholar Second, we illuminated the incident light vertically without SPPs excitation and found that the MoS2 nanoflakes remained unchanged (Figures S4D–S4F). Third, to exclude the light irradiation angle effect, we used bare glass and an indium tin oxide-coated glass slide under the same total internal reflection configuration and observed unchanged MoS2 nanoflakes (Figures S4G–S4L). Fourth, the theoretically determined relationship between the incident angle and the wavelength of p-polarized light shows that the SPPs could not be launched in the wavelength range from 520 to 570 nm (Figure S5). We used a 570 nm p-polarized light as the light source and found the MoS2 nanoflakes could not be etched. These results provided solid evidence that the excitation of SPPs was pivotal to the etching of MoS2 nanoflakes. Moreover, the local temperature variation can be calculated from the differential plasmonic image of MoS2 nanoflakes based on the quantitative dependence of local reflectivity on temperature.39Chen Z.X. Shan X.N. Guan Y. Wang S.P. Zhu J.J. Tao N.J. Imaging local heating and thermal diffusion of nanomaterials with plasmonic thermal microscopy.ACS Nano. 2015; 9: 11574-11581Crossref PubMed Scopus (52) Google Scholar We found that the temperature increase was negligible, ruling out the possibility of the photothermal effect on the etching of MoS2 nanoflakes (Figure S6). To clarify the etching process, we tried to probe the potential active oxidizing species that might be associated with plasmonic excitation and etching of the MoS2. We first used ascorbic acid and isopropanol solutions to trap oxidizing ·O2− and ·OH, which are representative reactive oxygen species (ROS). The etching of MoS2 nanoflakes still occurred, ruling out the effect of ·O2− and ·OH in our system (Figures S7A–S7D). The hydrogen peroxide can be another candidate oxidizing species, which presents strong oxidizing ability.40Waiskopf N. Ben-Shahar Y. Galchenko M. Carmel I. Moshitzky G. Soreq H. Banin U. Photocatalytic reactive oxygen species formation by semiconductor-metal hybrid nanoparticles. Toward light-induced modulation of biological processes.Nano Lett. 2016; 16: 4266-4273Crossref PubMed Scopus (81) Google Scholar,41Munkhbat B. Yankovich A.B. Baranov D.G. Verre R. Olsson E. Shegai T.O. Transition metal dichalcogenide metamaterials with atomic precision.Nat. Commun. 2020; 11: 4604Crossref PubMed Scopus (32) Google Scholar We used 0.01 mg/mL catalase solution as H2O2 scavenger and observed that the etching of MoS2 nanoflakes was significantly suppressed (Figures S7E and S7F). This finding indicated that the presence of H2O2 played a critical role in the etching process of MoS2 nanoflakes. It is known that SPPs at Au/MoS2 interface can generate hot (or highly energetic) carriers via two possible pathways: the non-radiative decay into hot carriers in Au and the resonant energy transfer to excite electrons and holes in MoS2 directly.42Petek H. Photoexcitation of adsorbates on metal surfaces: one-step or three-step.J. Chem. Phys. 2012; 137091704Crossref PubMed Scopus (51) Google Scholar, 43Tan S. Argondizzo A. Ren J. Liu L. Zhao J. Petek H. Plasmonic coupling at a metal/semiconductor interface.Nat. Photon. 2017; 11: 806-812Crossref Scopus (174) Google Scholar, 44Zhang Z.S. Liu L.H. Fang W.H. Long R. Tokina M.V. Prezhdo O.V. Plasmon-mediated electron injection from Au nanorods into MoS2: traditional versus photoexcitation mechanism.Chem. 2018; 4: 1112-1127Abstract Full Text Full Text PDF Scopus (57) Google Scholar, 45Shan H. Yu Y. Wang X. Luo Y. Zu S. Du B. Han T. Li B. Li Y. Wu J. et al.Direct observation of ultrafast plasmonic hot electron transfer in the strong coupling regime.Light Sci. Appl. 2019; 8: 9Crossref PubMed Scopus (111) Google Scholar In the former case, the hot electron with sufficient energy can transfer from Au into MoS2 by overcoming the barrier at the interfaces (Figure 3A). The plasmonic hot electron will thermalize during the transport and react with oxygen and H2O to produce H2O2 through O2+2e−+2H+→H2O2. In our case, H2O2 can induce the oxidation dissolution and etching of MoS2 multilayers into desired layers in the aqueous environment viaMoS2+9O2+6H2O+18e−→ MoS42−+2SO42−+12OH− The formation of H2O2 during SPP interaction was confirmed using a fluorescence probe (Figure S8). In addition, we did not observe a layer-by-layer etching of MoS2 nanoflake when it was immersed in a 0.1 M H2O2 solution for 2 h without the excitation of SPPs (Figure S9). Notably, the replacement of DI water to 0.01 M H2O2 aqueous solution enhanced the etching process under plasmonic excitation. Taken together, only H2O2 cannot etch the MoS2 nanoflakes layer by layer, as shown in our SPPs-induced etching. The promotion of etching with addition of H2O2 implies that the etching of MoS2 nanoflakes is a result of this oxidizing H2O2 formed during the SPPs irradiation. The participation of dissolved O2 in the dissolution of MoS2 was confirmed by irradiating the Au/MoS2 sample in the air, where the etching process was completely suppressed (Figure S10). In addition, the etching was significantly suppressed when N2-saturated deionized water was used as the medium. These results validated the presence of dissolved O2 was important for the etching of MoS2 nanoflakes. To further verify the plasmonic hot electron transfer process, we inserted a 3 nm Si3N4 insulating layer between Au and MoS2 that can effectively prohibit the electron transfer to MoS2, as schematically shown in Figure 3B. Optical images display that the etching of MoS2 was completely inhibited in the Au/Si3N4/MoS2 sandwich structure (Figure S11), which strongly supports that the charge transfer from Au to MoS2 is vital in the etching process. In contrast, the resonant energy transfer from the plasmon decay can excite electrons and holes in MoS2 directly. When the spectral functions of surface plasmon of Au film and exciton excitation in thin MoS2 overlap, the strong plasmon-exciton coupling regime can be expected.46Gonçalves P.A.D. Bertelsen L.P. Xiao S.S. Mortensen N.A. Plasmon-exciton polaritons in two-dimensional semiconductor/metal interfaces.Phys. Rev. B. 2018; 97041402Crossref Scopus (51) Google Scholar, 47Mukherjee B. Kaushik N. Tripathi R.P. Joseph A.M. Mohapatra P.K. Dhar S. Singh B.P. Kumar G.V. Simsek E. Lodha S. Exciton emission intensity modulation of monolayer MoS2 via Au plasmon coupling.Sci. Rep. 2017; 7: 41175Crossref PubMed Scopus (43) Google Scholar, 48Liu W. Lee B. Naylor C.H. Ee H.S. Park J. Johnson A.T. Agarwal R. Strong exciton-plasmon coupling in MoS2 coupled with plasmonic lattice.Nano Lett. 2016; 16: 1262-1269Crossref PubMed Scopus (269) Google Scholar, 49Lee B. Liu W. Naylor C.H. Park J. Malek S.C. Berger J.S. Johnson A.T.C. Agarwal R. Electrical tuning of exciton-plasmon polariton coupling in monolayer MoS2 integrated with plasmonic nanoantenna lattice.Nano Lett. 2017; 17: 4541-4547Crossref PubMed Scopus (97) Google Scholar The band structure calculated with a two-layer MoS2 represents the typical indirect band gap and notable interlayer features between the S atoms in the upper and lower layers (Figures 3C and 3D). Two kinds of exciton excitation can take place in MoS2 flakes, which are the direct gap excitation at the K-K point (1.7 eV) and the indirect band-gap excitation at the Γ-K points (1.2 eV) mediated by intervalley scattering.50Gusakova J. Wang X.L. Shiau L.L. Krivosheeva A. Shaposhnikov V. Borisenko V. Gusakov V. Tay B.K. Electronic properties of bulk and monolayer TMDs: theoretical study within DFT framework (GVJ-2e method).Phys. Status Solidi A. 2017; 214: 201700218Crossref Scopus (182) Google Scholar,51Zheng Q.J. Saidi W.A. Xie Y. Lan Z.G. Prezhdo O.V. Petek H. Zhao J. Phonon-assisted ultrafast charge transfer at van der Waals heterostructure interface.Nano Lett. 2017; 17: 6435-6442Crossref PubMed Scopus (120) Google Scholar Considering the SPP has an energy of 1.85 eV, both the two kinds of the exciton can be excited in MoS2 through the resonant energy transfer process. Thus, the holes are formed in the valence band at both Γ and K points, where those at K point will further decay to the valence band maximum (VBM) at Γ point. Moreover, the S-S distance between the upper and lower layers (dss) is found to increase dramatically as a function of increased hole doping in VBM (Figure 3E). Consequently, the accumulation of holes in VBM of MoS2 can weaken the interlayer interactions in MoS2, which further facilitates the oxidation dissolution of top layers of MoS2. Different from the classic mechanical exfoliation where the layers can be peeled away, in our case the interlayer-expanded MoS2 can be only removed by the oxidation dissolution. After the Si3N4 insulating layer was introduced, the hot electron induced dissolution was suppressed, although the hole induced interlayer repulsion can still exist. The Si3N4 layer does not prevent the exciton-plasmon coupling, but only prohibits the hot electron transfer from Au to H2O and O2, as shown in Figure 3B. The latter strongly limits the oxidation dissolution reaction with electrons. Therefore, we suggest that the holes in VBM of MoS2, produced via exciton-plasmon coupling, are responsible for the interlayer repulsion, and the electrons, produced in Au via plasmon decay, are responsible for the dissolution. These two processes together contribute to the overall etching process layer by layer. The residual layers of MoS2 suggest a strong interaction between the Au film and adjacent MoS2 layer, which could stem from the strong affinity of Au for sulfur atoms with high bond strength.52Häkkinen H. The gold-sulfur interface at the nanoscale.Nat. Chem. 2012; 4: 443-455Crossref PubMed Scopus (1191) Google Scholar Density functional theory (DFT) calculations show a binding energy value up to 1.43 J/m2 between Au and MoS2, confirming the strong interaction (Figure S12). We also applied this method to prepare other 2D TMD materials, including WS2, MoSe2, and ternary MoS2xSe2(1−x) (Figure S13), which showed similar etching features. More interestingly, we found that the etching of MoS2 top layers was highly dependent on the SPP propagation direction, which could be easily tuned by changing the light irradiation position in the optical path. A video of the SPPs propagation direction-dependent etching behavior is presented in Video S2. Figures 4B and 4D show several snapshots of the etching process. When the SPPs propagated from left to right, etching occurred first on the left side of the MoS2 nanoflake and gradually extended toward the center along the SPPs propagation direction (Figures 4A and 4B). When we adjusted the SPP propagation direction to be from right to left, the right edge of the nanoflake began to be etched toward the central region (Figures 4C and 4D). Finally, the whole top layers of the MoS2 nanoflake were completely etched. https://www.cell.com/cms/asset/2d2fc88b-8b9b-4445-90b8-6ef8b9568e85/mmc3.mp4Loading ... Download .mp4 (2.13 MB) Help with .mp4 files Video S2. A movie showing the surface plasmon polariton propagation direction-dependent etching behavior To reveal the anisotropic etching behavior, we captured plasmonic images of MoS2 nanoflakes. The scattering pattern of MoS2 shows a high dependence on the SPP propagation direction (Figure 4E). The edge of MoS2 encountering the SPPs exhibited a higher intensity contrast than the edge away from the SPPs. To quantify the local electrical field distribution, we employed COMSOL Multiphysics software to simulate the scattering pattern of MoS2 nanoflakes (Figure S14). The simulated images clearly show the local electric field distribution on a MoS2 nanoflake, with the electric field intensity gradually decreasing along the SPPs propagation direction (Figure 4F). The strongest electric field emerged at the edge of MoS2 encountering the SPPs, consistent with the plasmonic images. Furthermore, we also simulated the local electric field distribution on MoS2 with offset angle variation (Figure S15), which conformed to the above results. Thus, the experimental plasmonic images and theoretical simulation suggest that the electric field spatial distribution along the SPP propagation direction determines the etching direction of MoS2 nanoflakes. Based on the etching capability of the plasmonic coupling at Au/MoS2 interface, here, we demonstrated a proof-of-concept application in MoS2 patterning by manipulating the propagation direction of the SPP and the illumination duration (Figure 5A). Figure 5B presents the patterning process of a MoS2 homostructure using two different SPP propagation directions. We obtained a multilayer MoS2 domain on the top center after switching the direction of the light irradiation. Because the etching initiation was highly dependent on the propagation direction of the SPP, we designed MoS2 homostructures with different features through precise selection of the SPP propagation direction and etching time. The desired shape, size, and thickness of multilayer domains remaining on the top center of MoS2 flakes can be explicitly patterned (Figure 5C), providing new possibilities for realizing novel properties in MoS2-based electronic devices. Because the etching area is dependent on the beam diameter, we could enlarge the illumination size to achieve large-scale etching of TMD based on other setups, such as prism based configuration for SPP excitation. Our SPPE approach is versatile, being applicable not only to fabrication of TMDs with controlled layers and size but also to artificial patterning of desired architectures, which is a challenge for other techniques. Therefore, these results demonstrated that our method has great potential for sample processing on the nanoscale owing to the oxidizing capability and controlled manipulation of propagating polaritons, which can be expected to be utilized in future micro-/nanoprocessing technology. In summary, we have successfully developed a general approach for the precise control of the tuneable thickness and size of 2D TMD materials. Thick MoS2 multilayers can be precisely etched into monolayers, bilayers, and trilayers by simply varying the output power of the exciting light while preserving the pristine lateral size. In combination of the theoretical calculations, we revealed that the production of ROS from the plasmonic hot electrons and weakened interlayer interaction from the exited holes in MoS2 were responsible for the etching event, and diverse MoS2 homostructures can be patterned by manipulating the propagation direction of the SPPs. Our findings provide not only a deeper understanding of SPPs-driven photochemistry but also a new perspective on plasmon-matter interactions, offering a paradigm to precisely fabricate 2D TMD materials to obtain novel physical properties and devices.