# Fileset

[Iwanaga_et_al_Nanophotonics13(2024)95.pdf](https://mdr.nims.go.jp/filesets/f85a8551-e2c9-4035-83b5-0ea97f182eff/download)

## Creator

[Masanobu Iwanaga](https://orcid.org/0000-0002-8930-6940), [Xu Yang](https://orcid.org/0000-0001-8195-5850), [Vasilios Karanikolas](https://orcid.org/0000-0002-4829-8921), [Takashi Kuroda](https://orcid.org/0000-0001-6445-7673), [Yoshiki Sakuma](https://orcid.org/0000-0001-6804-7217)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Prominently enhanced luminescence from a continuous monolayer of transition metal dichalcogenide on all-dielectric metasurfaces](https://mdr.nims.go.jp/datasets/7eb9b14c-cd55-4378-86b4-500549568ef1)

## Fulltext

Prominently enhanced luminescence from a continuous monolayer of transition metal dichalcogenide on all-dielectric metasurfacesNanophotonics 2024; 13(1): 95–105Research ArticleMasanobu Iwanaga*, Xu Yang, Vasilios Karanikolas, Takashi Kuroda and Yoshiki SakumaProminently enhanced luminescence froma continuous monolayer of transition metaldichalcogenide on all-dielectric metasurfaceshttps://doi.org/10.1515/nanoph-2023-0672Received October 9, 2023; accepted December 8, 2023;published online December 25, 2023Abstract: 2D materials such as transition metal dichalco-genides (TMDCs) are a new class of atomic-layer mate-rials possessing optical and electric properties that sig-nificantly depend on the number of layers. Electronictransitions can be manipulated in artificial resonantelectromagnetic (EM) fields using metasurfaces and otherdesigned nanostructures. Here, we demonstrate promi-nently resonant enhancement in the photoluminescence(PL) of atomic monolayer, WS2, doped with a small quantityofMo. The excitonic PL showed a strong enhancement effecton a higher-ordermagnetic resonance of all-dielectricmeta-surfaces consisting of periodic arrays of Si nanopellets. ThePL intensity witnessed a 300-fold enhancement comparedto the reference PL intensity on a flat silicon dioxide (SiO2)layer, which suggests a drastic change in the dynamics ofphotoexcited states. Confocal PL microscopy and the analy-sis revealed that the single photonswere coherently emittedfrom the TMDC monolayer on the metasurface. Further-more, examining the PL lifetime in the ps and ns timescalesclarified two exponential components at the prominentexciton PL: a short-time component decaying in 22 ps anda long-time component lasting over 10 ns. Therefore, wePresent address: XuYang, Institute ofMaterials and Systems for Sustain-ability, Nagoya University, Nagoya 464-8601, Japan.Present address: Vasilios Karanikolas, Materials Modelling, Institute ofMaterials Science, Technical University of Darmstadt, 64287 Darmstadt,Germany, E-mail: karanikv@tcd.ie.*Corresponding author: Masanobu Iwanaga, National Institutefor Materials Science (NIMS), 1-1 Namiki, Tsukuba 305-0044, Japan,E-mail: iwanaga.masanobu@nims.go.jp. https://orcid.org/0000-0002-8930-6940Xu Yang, Vasilios Karanikolas, Takashi Kuroda and Yoshiki Sakuma,National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba 305-0044, Japan. https://orcid.org/0000-0001-8195-5850 (X. Yang). https://orcid.org/0000-0002-4829-8921 (V. Karanikolas). https://orcid.org/0000-0001-6445-7673 (T. Kuroda). https://orcid.org/0000-0001-6804-7217(Y. Sakuma)can infer that the radiative components were significantlyactivated in the TMDC monolayer on the metasurfacesin comparison to the reference monolayer on a flat SiO2layer.Keywords: 2D materials; transition metal dichalcogenide;tungsten disulfide; all-dielectric metasurface; enhancedphotoluminescence; resonance enhancement; coherentphoton emitter1 IntroductionOne of the features of the transition metal dichalcogenides(TMDCs) is to constitute atomic-layer 2D materials and fur-thermore to comprises van der Waals layers. Their opticalproperties change drastically depending on the number oflayers [1, 2]; in particular, the monolayers have a directbandgap and exhibit strong excitonic luminescence. Thisfeature greatly stimulated interest in fundamental studieson the luminescent properties. It was reported to date thatthe wavelengths of excitonic photoluminescence (PL) varydepending on the surrounding circumstance [3, 4], chem-ical [5, 6] or thermal [7] treatments, substrates [5, 8–10],and oxidation [11]. In addition to the surrounding circum-stances, the reported results sometimes differed from oth-ers; for example, the substrate effect was reported suchthat PL intensity of WS2 monolayer was more than 10-foldlarger on a sapphire substrate than an SiO2/Si substrate[8] whereas that was more than 5-fold weaker on a sap-phire substrate than an SiO2/Si substrate [9]. Probably, thegrowth conditions substantially affected the results. Also,interpretations of the experimental results were sometimesinconsistent; for example, itwas claimed that a TMDCmono-layer suspended on a micrometer-diameter air hole emit-ted more intense PL than the monolayer on substrates [8]whereas it was reported that a skew monolayer showedlarger PL intensity than a flat monolayer because a poten-tial to excited carriers were induced [9]; in the former,monolayer curvature on a large air hole was neglected.Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License.https://doi.org/10.1515/nanoph-2023-0672karanikv@tcd.iemailto:iwanaga.masanobu@nims.go.jphttps://orcid.org/0000-0002-8930-6940https://orcid.org/0000-0001-8195-5850https://orcid.org/0000-0002-4829-8921https://orcid.org/0000-0002-4829-8921https://orcid.org/0000-0001-6445-7673https://orcid.org/0000-0001-6445-7673https://orcid.org/0000-0001-6804-721796 — M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfacesThus, there were sometimes confusions or inconsistencyamong the numerous studies in the early years of the TMDCstudy. However, we can summarize that the PL wavelengthscoming from the excitons vary to some extent depending onthe circumstances and treatments of the TMDC atomic lay-ers. Additionally, dopant effect can alter the PL wavelengths[9, 12, 13], as was studied for many semiconductors.To achieve a higher spontaneous emission (or PL) effi-ciency than that of the as-grown monolayers, TMDC mono-layers were transferred to metallic [14–21] or dielectric[22–27] nanostructures. However, most of the results wereprimarily limited to PL-intensity enhancement factors (EFs)in the range of ten folds or less [14–18, 20–24, 26, 27]. Afew exceptions reached more than a 100-fold enhancementfor WSe2 [19], MoS2 [14], and WS2 [25]; the monolayersof WSe2 and MoS2 have advantage at the PL wavelengthslonger than 650 nm because the wavelengths reduce theoptical loss from the constituent noble metals forming theplasmonic nanostructures [14, 19]; in the case of WS2, themonolayer was transferred onto an Si3N4 grating and emit-ted directional PL via a diffraction mode, meaning thatthe geometric emission control mainly contributed to inthe PL-intensity enhancement, whereas the claim of over300-fold PL enhancement contradicts against the measuredangle-resolved PL-intensity ratio indicating ∼25-fold [25].Although the effective interplay of the nanostructures withthe TMDC monolayers is considered to lead to obviouschanges in the PL spectra, such results have rarely beenreported, except for the plasmonic case for MoS2 [14] anddielectric nanophotonic structures for WSe2 [27, 28]. Trialswere conducted to explore the prominent PL enhancementby transferring TMDCmicrostructures (primarily triangles)to the nanostructures. Recently, the wafer-scale growth ofTMDC is in progress [29–32]; however, continuous atomiclayers have not yet been employed for the PL-enhancingstudies.We point out the difference between TMDC-transferrednano- and microstructures. Most of the nanostructureshave nano-size air gaps (∼100 nm); therefore, curvaturesof monolayers on the air gaps are suppressed comparedto the micrometer air hole (≥5 μm diameter) [8]. Thus, thepotential trap effects due to the nano air gap are consideredto be quite smaller than the micrometer hole, thereby beingnegligible.Here, we show a clear change in the PL spectrum ofa TMDC continuous atomic layer due to the resonance ofall-dielectric metasurfaces composed of Si nanopellet (orshort-height nanocolumn) array. Such a spectral change hasnot been explicitly demonstrated using any other dielec-tric nanostructures. Furthermore, we show a prominent300-fold PL-intensity enhancement in comparison with thePL intensity on the reference flat SiO2 layer. The confocal PLimage is shown to realize coherent single-mode emission,which obeys Poisson distribution and is direct evidencefor single-photon emission. Such a direct imaging of pla-nar quantum emitter has never been reported to our bestknowledge. The enhanced PL was analyzed on the temporalprofiles and consequently revealed that both fast (ps) andslow (ns) components are activated on the metasurfaces.The unusual PL responses are described in terms of excitondynamics.The all-dielectric metasurfaces in this study showhighly enhanced fluorescence (FL) that exceeded 1000-foldmagnification [33] and can be applied to FL biosensors tar-geting a wide range biomolecules from proteins to DNA[34–37]. A proof for single cell-free DNA detection wasrecently reported [38]. The metasurfaces are a counterpart of plasmon–photon hybrid metasurfaces [39] with anextreme FL-enhancing capability [40–42]. In this article,we explore capability of the all-dielectric metasurfaces toenhance the PL of the TMDC atomic monolayer.2 Results and discussionA schematic illustration of a TMDC atomic monolayer andthe TMDC-transferredmetasurface composed of Si nanopel-let array is shown in Figure 1A. The atomic layer was visu-alized using VESTA [43]. A photograph of the actual sam-ple is shown in Figure 1B. The metasurface substrate usedwas a 2 cm × 2 cm square. The four white arrows indi-cate the corners of the continuous TMDC layer transferredonto the metasurface. The three metasurfaces are alignedvertically along the centerline and covered by the atomiclayer. Each metasurface area was set to 1.2 mm × 0.6 mmdimensions. The growth and transfermethods are describedin the Section 4.1 and 4.2, respectively.2.1 Structures of TMDC-transferredmetasurfacesFigure 1C shows optical microscopy images of a TMDC-transferredmetasurface. TheTMDCfilmwasnearly amono-layer that was grown under the condition described inthe Methods section, whereas the μm-scale bilayers appearsparsely as triangle shapes. The black scale bar indicates25 μm. The Inset shows a magnified view with a scale barof 5 μm, which includes a triangle bilayer at the top left. Afurther magnified view is shown in Figure 1D, which wastaken using a scanning-electron-microscopy (SEM) instru-ment (SU8230, Hitachi High-Tech, Tokyo, Japan); the scalebar indicates 500 nm. In the top-view SEM image, a holeM. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces — 97(A) (C) (D)(B)Figure 1: Structural overviews. (A) Schematic of TMDC monolayer (Gray: transition metal atoms. Yellow: sulfur atoms) and TMDC-transferredmetasurface of Si-nanopellet array. (B) Photograph of an atomic-layer TMDC-transferred metasurface substrate of a 2-cm square. The corners ofthe TMDC are indicated by white arrows. Three metasurface areas are aligned along the centerline and under the TMDC layer. (C) Optical microscopyimage of atomic-layer TMDC transferred on an all-dielectric metasurface. The atomic layer was primarily monolayer. The scale bar indicates 25 μm.Inset is a magnified view with a scale bar of 5 μm, where a triangle-shape bilayer is observed at the top left. (D) Top-view SEM image of the transferredTMDC monolayer onto the metasurface. The scale bar indicates 500 nm. The triangle indicates a broken hole in the TMDC layer.in the TMDC layer indicated by a triangle is shown, whichis in contrast to the transferred TMDC layer that cov-ers approximately the entire top of Si-nanopellet array.Although the Si nanopellets were designed to be circularcolumns, theywere approximated as regular octagons in thenanofabrication and had slight deviation from the perfectcircles.2.2 Optical resonances of the metasurfaceTo understand electromagnetic (EM) resonant modes ofthe all-dielectric metasurface itself, the numerically cal-culated reflectance spectrum is shown in Figure 2A. Themetasurface was set in accordance with the actual sample(Figure 1D), such that the periodicity of array was 400 nmand the diameter of Si nanocolumn in the xy plane was228 nm. Inset shows the unitcell at an xy section includingthe Si nanocolumn (purple). The optical resonances of themetasurface appear in thewavelength range below 900 nm.High reflectance peaks and deep dips exhibit prominentoptical resonances. The four resonances relevant to thisstudy are indicated by triangles B–E at the wavelengthsof 592.5, 608.8, 646.9, and 718.1 nm, respectively, which arelocated near the PL wavelengths of the TMDC monolayer inthis study.The resonant EM-field distributions are shown inFigure 2B–D corresponding to the wavelengths B–D inFigure 2A, respectively. The incident EM fields were set to1.0. In Figure 2B–D, the interfaces of the Si nanopellets andSiO2 layer are shown with white lines in the xz section, andthe boundary of the Si nanopellets is shownwithwhite linesand circles; the xz sections are set at the center of the Sinanopellets (dashed line in the inset of Figure 2A), and thexy sections are set at the half height of the Si nanopellets.At the reflectance dip B, themagnetic fields are stronglylocalized in the Si nanopellet, forming a waveguide modeassociated with a 92.1-fold enhancement in the field at themaximum. The electric field shows the greatest enhance-ment at the sidewall of the Si nanopellet, exceeding 60-foldin comparison to the incident electric field.Themagnetic-field distributions at the reflectance peakC are different from those at the dip B. Multipoles areobserved in the Si nanopellet, exhibiting a higher-ordermagnetic resonance associated with more than a 120-foldenhancement in the field compared to that of the incidentfield. The electric fields are predominantly distributed at theoutermost surface of the Si nanopellet; the intensity exhibitsa 127-fold enhancement for that of the incident electric field.Thus, the enhanced electric fields available at the top ofthe Si nanopellet differ from those on the resonance inFigure 2B.Although the reflectance peaks D and E are sepa-rated by a deep dip at 682.5 nm, their EM-field distri-butions are similar to each other. Figure 2D shows astrongly localized magnetic mode inside the Si nanopel-lets. More than 100-fold enhancements in the maximumintensities are observed for both electric and magneticcomponents, as compared to the incident field intensity.The enhanced electric fields are primarily distributed onthe sidewalls of the Si nanopellets as shown in Figure 2D.Figure S1 in the Supplementary Material provides the fielddistributions at the reflectance peak E.98 — M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces(A)(C) (D)(B)Figure 2: Optical resonances of the all-dielectric metasurface. (A) Computed reflectance spectrum of the metasurface consisting of a 400-nm-periodicity square array of Si nanopellets (or low-height nanocolumn) with a diameter of 228 nm. Triangles indicate resonant wavelengths B–E,corresponding to dip and peaks of the reflectance spectrum. Inset illustrates the unitcell in the xy plane on a z section; dashed line indicates xz sectionthrough the center of the Si nanopellet. (B)–(D) Resonant electric- and magnetic-field intensity distributions, |E|2 and |H|2, at the wavelengthsindicated by the triangles in (A). The xz sections correspond to the dashed line in the inset of (A). The xy section is set at half the height of the Sinanopellet. Each intensity is represented by setting the incident field intensity to 1.0. See the Supplementary Material (Figure S1) for the |E|2 and |H|2distributions at the condition of triangle E in (A).2.3 Enhanced PL spectrum and imageFigure 3A shows the PL spectra measured under excitationat 532.0 nm. The PL spectra of the atomic layer on and offthe metasurface are shown with orange and gray curves,respectively. The PL spectrum off the metasurface is magni-fied by 10-fold, for clarity. The PLEF spectrum, defined as theratio of the PL spectra on and off the metasurface, is shownwith a black curve plotted on the right axis. The peak valueof the PL EF spectrum is as large as 300 at 614.5 nm. The peakof the PL spectrum on themetasurface is located at 619.4 nmwhereas that off the metasurface is located at 650 nm. Thus,in addition to the large PL EF, a distinct spectral shapechange occurs in the PL enhancement on the metasurface.We note that the PL peak of WS2 monolayer on flatsubstrates can largely shift in a wavelength range from 620to 670 nm, due to the substrate, transfer, and/or nm-scalestrain [8, 9]. Thus, the PL peak at 650 nm off themetasurfacein Figure 3A is a plausible result.The enhanced PL-EF peak shows a good agreementwith a reflectance dip at 616 nm in Figure 3B. The nor-malized reflectance spectrum was measured at the normalincidence on the TMDC-transferred metasurface, showingreflection of the coupled system of the TMDC monolayerand the metasurface. As a result, the reflectance is reducedbelow 650 nm, compared to the metasurface reflectance inFigure 2A, because of light absorption by the TMDC mono-layer. Further, a small reflectance dip at 605 nm in Figure 3Bis considered to originate from a reflectance peak C in thecomputation (Figure 2A). Thus, the measured peak of PL EFappeared near the resonance C of the metasurface. Notethat, although there is another resonance of the reflectancepeakD inFigure 2A, it does not contribute to the PL enhance-ment in 620–650 nm.Generally, prominent PL enhancementselectively appears at particular resonances [33, 40, 41]. Thesmall dip at 616 nm ismost likely a coupled resonance of theTMDC exciton with the optical resonance C in the metasur-face. In short, the prominent PL enhancement at 614.5 nmtakes place with help of a particular EM resonance in themetasurface.Although many reports have addressed TMDC PLenhancement using artificially designed nanostructures[14–27], this type of resonant effect associated with definitechanges of spectral shapes has not been reported, exceptfor a previous report using plasmonic nanocavities [14].This study moreover presents the largest value of PL EF forthe TMDC monolayers, compared to those obtained by thedielectric nanostructured platforms previously reported.M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces — 99(A)(B)(C) (D)Figure 3: PL measurement on the metasurface (MSF). (A) Enhanced PLspectrum measured on the metasurface (orange). PL spectrum on theflat BOX layer (gray), which is 10-fold magnified for clarity. PL EF is plottedfor the right axis (black solid). (B) Normalized reflectance spectrum of theTMDC-layer-transferred metasurface, measured at the normal incidence.(C) Confocal PL image, which is shown in orange-hot presentation. Whitescale bar indicates 200 μm. (D) Photon-number distribution (orangebars) evaluated from the confocal image in (B) and fitted using Poissondistribution (open black bars), which represents a quantum light effect,coherent single-mode emission [44].We here refer to the optical properties of the as-grownTMDC atomic layers. The Raman-scattering and PL spectraare shown in the Supplementary Material (Figure S2). TheRaman spectrum shows that the TMDC consists primarilyof WS2 with a small quantity of Mo. The PL spectrum hasa peak at 625 nm, representing A exciton of WS2. As wementioned the previous reports [8, 9], transfer of TMDC filmoften resulted in PL-peak shift, which was observed in thisstudy as well. As shown in Figure 3A, the TMDC transferredonto the SiO2 layer has a PL peak at 650 nm.High-spatial-resolution PL images were acquired usingan upright confocal FL microscope (Stellaris 5, LeicaMicrosystems, Wetzlar, Germany). One of the PL images isshown in Figure 3C, where the pseudo-color representationis orange-hot and the white scale bar indicates 200 μm.Rectangular colored area corresponds to the metasurface.Evidently, the PL becomes very dark outside the metasur-face though the TMDC atomic layer covers the outside aswell. In the confocal image, the PL intensity outside themetasurface is indistinguishable from the zero level in themeasurement. The confocal PL image was acquired in aphoton-counting manner; therefore, the PL intensity wasquantified by examining the number of emitted photons.Setting an analyzing region on the metasurface(Figure S3), the distribution of photon numbers wasobtained, as shown in Figure 3D. The vertical axisrepresents number of photon-emission events. Thephoton numbers (orange bars) are small and primarilylimited to five or fewer, including 14.4 % zero-photon and29.7 % one-photon events. The distribution was fitted usingthe Poisson distribution and reproduced fairly well, beingfound to have mean photon number of 1.47. Small deviationis probably due to structural imperfection of the TMDCmonolayer: the triangular bilayers and the broken holesas seen in Figure 1C and D, respectively. The fitted Poissondistribution is plotted in Figure 3D using open black bars.The photon distribution obeying Poisson distributionrepresents coherent single-mode photon emission [44],which is direct evidence for a quantum light emitter. Itthus turns out that the confocal PL image (Figure 3C)directly visualizes a TMDC quantum light emitter. Asa corollary of the coherent single-mode emission, thedegree of second-order coherence g(2) is derived to beunity, irrespective of the mean photon numbers [44]. Notethat the confocal PL imaging is more informative thanthe conventional g(2) measurement, which is reportedfrequently as experimental evidence of single-photonemission, and completes in much shorter time of tens ofseconds than the time-consuming g(2) measurement.2.4 Decay-time analysis of enhanced PLFigure 4A shows the PL decay profile in a ps range. ThePL was excited using laser pulses with 2-ps duration and540-nm wavelength and was detected around the PL-peakwavelengths. The instrumental response function (IRF) ofthe laser pulses in the setup used is shown with a greencurve, and it has a full width at the half maximum of10 ps. The initial response of the PL decay curve follows theIRF. The ultrafast response suggests that photocarriers areimmediately consumed via a nonradiative process. In thetime range later than 15 ps from the peak of the laser pulse,themeasured data (orange closed circles) are fitted using anexponential curve (black solid curve) such thaty = y0 + C0 exp[−(x − x0)∕𝜏s] (1)100 — M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces(A)(B)(C)Figure 4: Temporal profiles of measured PL. (A) And (B) fast and slowcomponents in ps and ns ranges, respectively. Measured data are shownwith orange closed circles and fitted curves with solid curves in both(A) and (B). The fitting exponential functions, Equations (1) and (2) for(A) and (B), respectively, are described in the text. In (A), the PL intensityis shown on the logarithmic scale, and the temporal profile of psexcitation laser pulse is drawn with a green dotted curve for reference.(C) PL profile measured off the metasurface (orange crosses). IRF to theps laser pulse (or the measured temporal profile) is also shown withgreen curves in (A) and (C).where C0 is a proportional constant, x0 is time offset, and 𝜏sis decay time in the ps range. The fitting results as shownin Figure 4A revealed the value of 𝜏s = 22.3± 0.3 ps. The PLlifetime under the excitation condition was approximately10 ps in a previous time-resolved study [45]. It was alsoreported that a WS2 monolayer grown on SiO2/Si had thedecay time of 11.4 ps and a partially oxidizedWS2 monolayerhad that of 31.5 ps [11]. Thus, the measured 𝜏s lies in aplausible range.To analyze the temporal profile of PL in a ns range,in Figure 4B, the time-integrated PL intensity over [0, T](orange dot) is plotted for the gate time T of a photon-counting detector. The growing temporal profile I(T) wasdefined, using an integrated double-exponential function,such thatI(T) =T∫0[C exp(−t∕𝜏s)+ D exp(−t∕𝜏L)]dt≈ C𝜏s[1− exp(−T∕𝜏s)]+ DT (2)and fitted the measured data (solid black curve) wherea relation 𝜏L ≫ T was assumed to derive Equation (2). C,D, and 𝜏s are the fitting parameters and the fitted curveis shown with a black curve, which well reproduces thePL-growing profile. The short lifetime 𝜏s was evaluated tobe 0.41± 0.16 ns; consequently, for gate time longer than5 ns, the profile I(T) dominantly depends on T and showsan approximately linear increase in Figure 4B. This resultindicates that the 𝜏L is sufficiently longer than the maxi-mum gate time of 11 ns. In Figure 4B, the total PL intensitycoming from the long lifetime component, D𝜏L, is largerthan that from the short component, C𝜏s, whereas the ratioof populations in the fast and slow components, i.e., C:D isdetermined to be 11.8: 1 using by Equation (2). These resultsimply that there are two possible excitonic transitions in theTMDC monolayer. A theoretical study on the exciton bandsof monolayer MoS2 showed that the A excitons have K–Kdirect and K–Γ intervalley transitions at approximately thesame energy [46], where K andΓ denote the high-symmetrypoints in the Brillouin zone. Because WS2 is consideredto have similar exciton bands to those of MoS2, the twoexcitonic transitions are attributed to the two observed PLcomponents.We mention that the 𝜏s is approximately one-orderlonger than that found in the ps range. This is because thetemporal resolution in Figure 4B is limited by the minimumgate time of 0.5 ns. Thus, the evaluated 𝜏s in the ns rangesuggests that 𝜏s is shorter than 0.4 ns and does not implyinconsistency with the analysis in the ps range (Figure 4A).The precise evaluation of 𝜏s is conducted in the ps range.The PL decay was also measured off the metasurfaceand then, the TMDC layer was on placed the SiO2 layer.These data are shown with orange crosses in Figure 4C. Theps-laser-pulse IRF is shown together with a green curve,being in agreement with the PL profile. The PL decay wasalso fitted by an exponential curve (not shown here) andthe decay time was 4.7± 0.1 ps. These results indicates thatthe PL decays together with the incident laser pulse andstrongly suggests the PL predominantly decays within 5 psvia ultrafast nonradiative process(es) because the PL inten-sity is heavily reduced in comparison with that of the TMDCmonolayer on the metasurface. In other words, the photo-carriers in the TMDC film off the metasurface are immedi-ately consumed through the nonradiative path(s).M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces — 1012.5 Resonantly enhanced PL processesFigure 5A and B shows schematics of the PL dynamics onand off the metasurface, respectively. Blue lines denote theTMDCmonolayers. Schematics of the structures and energydiagrams are drawn on the left and right sides, respec-tively. On the metasurface (Figure 5A), photoexcited car-riers are considered to have two major relaxation pathsbased on the analysis of PL decay time: one is the non-radiative process(es) and the other is the excitonic tran-sitions. Green arrows indicate excitation using the laserlight. Excited-state transfer is represented with solid blackarrows. Radiative and nonradiative transitions are shownwith red and dashed black arrows, respectively. The thick-ness of the arrows represents the probability of each pro-cess; the thicker arrows denote higher probabilities. Asshown in the temporal profiles (Figure 4), the excitonic tran-sitions have fast and slow components. However, these com-ponents are indistinguishable in the PL spectrum, becausethe transitions occur at approximately the same energy [46].In Figure 5A, the vertical red arrow denotes the direct exci-tonic transition and the oblique red arrow does the indirect(or intervalley) transition.(A)(B)Figure 5: Schematic of the PL dynamics: (A) and (B) TMDC monolayers(blue) on and off the metasurface, respectively. Structural schematics aredrawn on the left side; excitons are represented with orange rounds.Energy diagrams are depicted on the right side. Green arrows representphotoexcitation. Solid red and black arrows denote radiative andexcitation transfer processes, respectively.Dashed black arrows indicatenonradiative decay. See more details in the text.In contrast, off the metasurface (Figure 5B), the PLdynamics are attributable to dominant nonradiative pro-cess(es), indicating that nearly all the photocarriers are con-sumedwithout emitting PL. A damping oscillatormodel [47]clarified that the excitation transfer on the top of dielectricbecomes drastically fast, when the distance between theexcited state and the surface of dielectric is close within afew nm [40]; the transfer rate exceeds 105 at the distanceof 0.1 nm, is more than 103 at 0.5 nm, and is in the one-digitorder at 5 nm and more. Although the excited states in theTMDC are not simple isolated oscillators, it is probable thatthe excitation transfer from the TMDC layer to the SiO2 layertakes place in a similar manner, as indicated by the blackarrows in Figure 5B.2.6 PL EFThe PL EF in the experimental configuration is expressed as[33, 40, 48]EF = NmN0× 𝜂m𝜂0× 𝛾m(k)𝛾0(k)(3)where the subscripts m and 0 denote on and off metasur-face, respectively.N is the photoexcited populations, 𝜂 is thequantum yield, and 𝛾(k) is the radiative decay rate depen-dent on outgoing wavevector k. In particular, 𝜂 = 𝛾∕(𝛾 +𝛾NR) where 𝛾NR means nonradiative decay rate. The ratio𝛾m∕𝛾0 is often referred to as the Purcell factor [49].The electric-field intensity at the excitation wavelength(Figure S4) shows that the TMDC monolayer was excitedefficiently by a maximum of 5.3-fold, which gave the ratioNm∕N0. Other terms in Equation (3) were based on the PLprocesses and Purcell effect.The Purcell effect was often discussedwith PL enhance-ment of the TMDC atomic layers [20, 27]. In a simple schemedescribing two-level systems in optical resonators, the spon-taneous emission is expedited through the resonant elec-tric fields. We numerically explored the Purcell effect inthis configuration using the all-dielectric metasurfaces; thedetails are noted in the Section 4.6. In case of z polarization,the total Purcell factor was as large as 180 at the maximum,whereas, for the in-plane (i.e., x or y) polarization, it wasapproximately 2 in the range of 600–650 nm (Figure S5). Thetotal Purcell factor contains both radiative and nonradiativeeffects. The A exciton in the TMDC is limited to the in-planepolarization [50]. Further, a radiative effect is observed inthe PL measurement. Overall, the factor 𝛾m∕𝛾0 is at most2 in Equation (3) and the other contribution comes from theratio 𝜂m∕𝜂0, which is estimated to be 30 or less. As illustratedin Figure 5, the change in exciton relaxation dynamics is amajor factor yielding the large PL EF (Figure 3A).102 — M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces3 ConclusionsThe experimental investigation of a coupled system com-posed of all-dielectric metasurfaces and a large-area con-tinuous TMDCmonolayer revealed that the exciton PL spec-trum was 300-fold enhanced at the maximum for the refer-ence PL spectrum. This enhancement is the most significantas compared to those in the previous studies using dielectricnanostructures [22–27]. The analysis for the confocal PLimage revealed that coherent single-mode photon emissionson the metasurface. The single mode substantially includessingle-photon emission [44]. Thus, this coupled system wasfound to function as a single-photon generator at room tem-perature. The enhanced PLwas alsomeasured and analyzedwith respect to the decay time, and the fast and slow radia-tive components were identified in the ps and ns ranges,respectively. The two components were accountable fordirect and intervalley transitions of the A excitons. Overall,the highly enhanced PL was achieved based on a multipleeffect of efficient excitation, activated radiative excitonicprocesses, and Purcell effect in this resonance-tuned cou-pled system.4 Experimental section4.1 TMDC growthAtomic-layer WS2 films were grown on c-plane sapphire substrates viacold-wall chemical vapor deposition at 800 ◦C and 50 Torr for 60 min.H2S and WOCl4 were used as sulfur and tungsten precursors, respec-tively. The flow rate of H2S was 1 sccm. The feed rate of N2 carrier gasthrough the WOCl4 canister was 300 sccm, and the temperature andpressure of the canister were maintained at 45 ◦C and 760 Torr, respec-tively. The total N2 flow rate in the reactor was 2500 sccm. Additionally,0.1 sccmO2 was injected during the deposition ofWS2. Dragontrail glassserving as a catalyst reservoir was placed upstream of the sapphiresubstrate during growth, providing Na and K catalysts and promotingthe lateral growth of WS2 with enlarged grain size.Due to the previous usage of the chamber for the growth of MoS2monolayers, a trace amount of Mo was found in the Raman spectrumof the as-grown film (Figure S2); however, because the ratio of Mo isestimated to be small (10 % or less) [13] and plays a minor role in theoptical properties, we simply refer to the TMDC as WS2.4.2 Transfer process onto metasurfacesThe as-grown monolayer WS2 films were transferred from the sap-phire onto the metasurface. The procedure is schematically illustratedin Figure S6. Before the delamination, the WS2/sapphire sample wascoated with polymethyl methacrylate (PMMA) to support and protectthe ultrathin WS2. Subsequently, the PMMA/WS2 stack was peeled offfrom the sapphire surface by allowing deionized water to penetratethe interface between the sapphire and WS2. Finally, the sample ofPMMA/WS2 on the above-mentioned metasurface was immersed inacetone at 50 ◦C for 60 min to remove the PMMA and then cleaned inisopropanol, followed by drying with a nitrogen blow.After the transfer, rapid temperature annealing was performed,which started at room temperature; the temperature was raised to300 ◦C and kept for 10 min under N2 gas flow at 120 sccm; subsequently,the sample was cooled down to room temperature under N2 gas flow.This annealing process improved the PL intensity a few times at themaximum.4.3 Nanofabrication of metasurfacesMetasurfaces were fabricated based on the Si-on-insulator wafers; thetop Si layerwas 200 nm thick, themiddle SiO2 layerwas 375 nm, and thebase Si waferwas 675 μm. The top Si layerwas crystallinewith the (100)surface and had a p-type dopant. The nanofabrication procedures usingelectron-beam lithography and selective dry etching of the Si layer havebeen described elsewhere in detail [34, 42].4.4 Optical measurementThe PL of the TMDC on the metasurfaces was measured at room tem-perature using amicro-PL setup for the spectrumacquisition and a con-focal microscope for the spectrally resolved PL images. In the micro-PLsetup, single-mode continuous-wave laser light of emissionwavelength532.0 nm (Action532Q-0050, AOTK, China) was focused on the TMDCmonolayer on themetasurface using a 50× objective lenswith a numer-ical aperture (NA) 0.55 (M Plan Apo, Mitsutoyo, Kawasaki, Japan) and aworking distance of 13.0 mm. The PL from the TMDC monolayer wascollected using the objective lens. The laser power was set to 0.5 mW orless at the entrance of the objective lens. The collected PL was passedthrough a monochromator (ISO-160, Teledyne Princeton Instruments,Trenton, NJ, USA), and the PL spectra were acquired using a coolingCCD camera (ProEM1024HS, Teledyne Princeton Instruments). The laserpower was set to 0.5 mW or less at the entrance of the objective lens.When comparing the PL intensities, we kept the laser power to be aconstant value.Normal reflectance in Figure 3B was measured using the setupfor the micro-PL measurement above. A 10× objective lens of NA 0.28(M Plan Apo, Mitsutoyo) was used. Light source was a halogen lampand two apertures were inserted at the entrance of the objective lensand an imaging lens (MT-L, Mitsutoyo). The apertures were set to benarrow and the incident angle of incoming beamwas limited to be lessthan 0.3◦, ensuring the normal incidence practically. Reference signalswere measured using a calibrated Al mirror. Thus, we quantitativelyevaluated reflectance of the TMDC–metasurface coupled system.For the confocal PL image shown in Figure 3C, the excitationwavelength was set to 514.0 nm and a 10× objective lens of NA 0.32(HC PL Fluotar, Leica Microsystems) was used. The lateral resolutionwas 819.2 nm. The PL growth in the ns range as shown in Figure 4B wasmeasured by varying the gate width of the photon-counting detector inthe confocal microscope.PL decay-time measurements were conducted using a 76-MHz-repetition pulsed laser (MIRA-OPO-X ps, Coherent, Santa Clara, CA, USA)with an emission wavelength of 540 nm and a pulse width of 2 ps. Thepulsed laser light was focused using a 50× objective lens with NA 0.8and aworking distance of 1.0 mm (MPFLN50X, Olympus, Tokyo, Japan).The laser power injected in the objective lens was 20 and 100 μW forthe TMDC monolayer on and off the metasurface, respectively. The PLwas collected using the same objective lens. The detection wavelengthsM. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces — 103were set to 615–620 and 640–650 nm on and off the metasurface,respectively, in accordance with the PL peaks (Figure 3A). The PL inthe ps range was acquired using a synchronously scanning streak cam-era (C5680, Hamamatsu Photonics, Hamamatsu, Japan), attached to amonochromator (250is, Chromex, Albuquerque, NM, USA). The IRFwitha fullwidth at halfmaximumof 10 ps (Figure 4A andC)was determinedby measuring the time traces of the excitation pulses.4.5 Computation of reflectance spectra and EM-fielddistributionsThe reflectance spectrum in Figure 2 and the emittance spectrum inFigure 3 were computed based on the rigorous coupled-wave analysis(RCWA) method [51], which was combined with a scattering-matrixalgorithm [52] for numerical stability. The RCWA code was imple-mented on a supercomputer. The EM-field distributions in Figure 2were output using the RCWA method as well. A material parameter,that is, the permittivity of crystalline Si constituting the metasurfacewere taken from literature [53]. The permittivities of air and SiO2 in thewavelength range of interest were set to their representative values of1.00054 and 2.1316, respectively.4.6 Numerical evaluation of purcell factorTo evaluate the Purcell factors in Equation (3), the excitons in theWS2 layer were approximated as two-level energy quantum systems.Their Purcell factors were calculated using the macroscopic quantumelectrodynamic theory. The excitation created by an electric dipolesource is connected in quantum mechanical terms with the relaxationof the exciton. In the real terms, the relaxation of the exciton wasobtained by solving theMaxwell equations for a point dipole excitationin a complicated environment, which includes the EM response of theinvolvedmaterials. A commercial finite-difference time-domain (FDTD)software (Ansys, Canonsburg, PA, USA) was used to solve the Maxwellequations. The optical responses of the different materials were deter-mined from their experimentally measured dielectric permittivities[53].Physically, the Purcell factor, Γ(r, 𝜔), gave the enhancement orinhibition of the relaxation rate of the excitonwhen theWS2 monolayerwas placed in the nanostructured environment, compared to the refer-ence relaxation value of theWS2 monolayer. A reference valuewas usedfor the case where the WS2 layer was placed in a homogeneous media.In the reference media, the WS2 excitons followed an exponentialrelaxation, exp(−t∕𝜏ref ), from the excited to the ground state, where𝜏ref is the reference lifetime. In the presence of the nanostructuredenvironment, the relaxationwas accelerated by the Purcell factor valueto exp(−Γt∕𝜏ref ). The Purcell factor is given byΓ = 𝜏ref∕𝜏NS, where 𝜏NSis the lifetime of the exciton in the nanostructured environment. Thefactor Γ can also be expressed as Γ = 𝛾m∕𝛾0, which is a factor on theright-hand side of Equation (3).Supplementary MaterialSupplementary Material is available online. Figure S1: EM-field distributions at the reflectance peak E in Figure 2A.Figure S2: Raman-scattering and PL spectra of the as-grownTMDC atomic layer. Figure S3: Confocal PL image with theanalyzing box. Figure S4: EM-field distributions of themeta-surface at 532 nm. Figure S5: Numerically evaluated Purcellfactors. Figure S6: Illustrative description for the transferprocedure.Acknowledgments: Nanofabrication, analysis for nano-structures, and confocal PL measurement were conductedat Namiki Foundry and Nanofabrication Platform in NIMS.Numerical implementations for the optical spectra and EM-field distributions were conducted using supercomputingresources at Cyberscience Center, Tohoku University.Research funding: M.I. thanks financial supports by JSPSKAKENHIGrantNumber JP20K21134 and the support systemfor curiosity-driven research in NIMS. T.K. thanks a supportof Innovative Science and Technology Initiative for Security,Grant Number JPJ004596, ATLA, Japan.Author contributions: M.I. conceptualization, numericaldesign and nanofabrication of the metasurfaces, PL mea-surement, confocal imaging, the PL-data analysis, PL decay-time analysis, andwriting original draft. X.Y. and Y.S. growthof the atomic-layer TMDC, optical measurement for the as-grown TMDC, and transfer of the atomic layer onto themetasurface. V.K. and T.K. numerical evaluation of the Pur-cell factor. T.K. PL decay-time measurement and analysisin the ps range. All authors have accepted responsibilityfor the entire content of this manuscript and approved itssubmission.Conflict of interest: Authors state no conflicts of interest.Data availability: Data sharing is not applicable to this arti-cle as no datasets were generated or analyzed during thecurrent study.References[1] A. Splendiani, L. Sun, Y. Zhang, et al., “Emergingphotoluminescence in monolayer MoS2,” Nano Lett., vol. 10, no. 4,pp. 1271−1275, 2010..[2] K. F. Mak, C. Lee, J. Hone, J. Shan, and T. F. Heinz, “Atomically thinMoS2: a new direct-gap semiconductor,” Phys. Rev. Lett., vol. 105,no. 13, p. 136805, 2010..[3] Y. Fu, D. He, J. He, et al., “Effect of dielectric environment onexcitonic dynamics in monolayer WS2,” Adv. Mater. Interfaces,vol. 6, no. 23, p. 1901307, 2019..[4] A. O. A. Tanoh, J. Alexander-Webber, J. Xiao, et al., “Enhancingphotoluminescence and mobilities in WS2 monolayers with oleicacid ligands,” Nano Lett., vol. 19, no. 9, pp. 6299−6307, 2019..[5] K. M. McCreary, A. T. Hanbicki, S. Singh, et al., “The effect ofpreparation conditions on Raman and photoluminescence ofmonolayer WS2,” Sci. Rep., vol. 6, p. 35154, 2016..[6] T. Aihara, R. Wang, X. Yang, Y. Sakuma, A. O. Okano, and M.Ikezawa, “Observation of strain relaxing in nanoscale WS2monolayers grown on SiO2/Si by organic solvent treatment,” Jpn. J.Appl. Phys., vol. 61, no. 7, p. 071003, 2022..104 — M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces[7] X. Hua, T. Axenie, M. N. Goldaraz, et al., “Improving the opticalquality of MoSe2 and WS2 monolayers with complete h-BNencapsulation by high-temperature annealing,” ACS Appl. Mater.Interfaces, vol. 14, pp. 2255−2262, 2022..[8] Y. Yu, Y. Yu, C. Xu, et al., “Engineering substrate interactions forhigh luminescence efficiency of transition-metal dichalcogenidemonolayers,” Adv. Funct. Mater., vol. 26, no. 26, pp. 4733−4739,2016..[9] C. Cong, J. Shang, Y. Wang, and T. Yu, “Optical properties of 2Dsemiconductor WS2,” Adv. Opt. Mater., vol. 6, no. 1, p. 1700767, 2018..[10] F. Zhang, C. Erb, L. Runkle, X. Zhang, and N. Alem, “Etchant-freetransfer of 2D nanostructures,” Nanotechnology, vol. 29, no. 2,p. 025602, 2018..[11] Z. Luo, W. Zheng, N. Luo, et al., “Photoluminescence lightening:extraordinary oxygen modulated dynamics in WS2 monolayers,”Nano Lett., vol. 22, no. 5, pp. 2112−2119, 2022..[12] S. Tongay, W. Fan, J. Kang, et al., “Tuning interlayer coupling inlarge-area heterostructures with CVD-grown MoS2 and WS2monolayers,” Nano Lett., vol. 14, no. 6, pp. 3185−3190, 2014..[13] X. An, W. Zhao, Y. Yu, et al., “Resonance Raman scattering ongraded-composition WxMo1−xS2 alloy with tunable excitons,” Appl.Phys. Lett., vol. 120, no. 17, p. 172104, 2022..[14] J. Huang, G. M. Akselrod, T. Ming, J. Kong, and M. H. Mikkelsen,“Tailored emission spectrum of 2D semiconductors usingplasmonic nanocavities,” ACS Photonics, vol. 5, no. 2, pp. 552−558,2018..[15] S. Butun, S. Tongay, and K. Aydin, “Enhanced light emission fromlarge-area monolayer MoS2 using plasmonic nanodisc arrays,”Nano Lett., vol. 15, no. 4, pp. 2700−2704, 2015..[16] G. M. Akselrod, T. Ming, C. Argyropoulos, et al., “Leveragingnanocavity harmonics for control of optical processes in 2dsemiconductors,” Nano Lett., vol. 15, no. 5, pp. 3578−3584, 2015..[17] J. Kern, A. Trügler, I. Niehues, et al., “Nanoantenna-enhancedlight−matter interaction in atomically thin WS2,” ACS Photonics,vol. 2, no. 9, pp. 1260−1265, 2015..[18] S. Wang, S. Li, T. Chervy, et al., “Coherent coupling of WS2monolayers with metallic photonic nanostructures at roomtemperature,” Nano Lett., vol. 16, no. 7, pp. 4368−4374, 2016..[19] Z. Wang, Z. Dong, Y. Gu, et al., “Giant photoluminescenceenhancement in tungsten-diselenide−gold plasmonic hybridstructures,” Nat. Commun., vol. 7, p. 11283, 2016..[20] S. R. K. Indukuri, C. Frydendahl, J. Bar-David, N. Mazurski, and U.Levy, “WS2 monolayers coupled to hyperbolic metamaterialnanoantennas: broad implications for light−matter-interactionapplications,” ACS Appl. Nano Mater., vol. 3, no. 10,pp. 10226−10233, 2020..[21] B. Du, Y. Li, M. Jiang, et al., “Polarization-dependent purcellenhancement on a two-dimensional h-BN/WS2 light emitter with adielectric plasmonic nanocavity,” Nano Lett., vol. 22, no. 4,pp. 1649−1655, 2022..[22] H. Chen, S. Nanz, A. Abass, et al., “Enhanced directional emissionfrom monolayer WSe2 integrated onto a multiresonantsilicon-based photonic structure,” ACS Photonics, vol. 4, no. 12,pp. 3031−3038, 2017..[23] A. F. Cihan, A. G. Curto, S. Raza, P. G. Kik, and M. L. Brongersma,“Silicon mie resonators for highly directional light emission frommonolayer MoS2,” Nat. Photonics, vol. 12, no. 5, pp. 284−290,2018..[24] T. Bucher, A. Vaskin, R. Mupparapu, et al., “Tailoringphotoluminescence from MoS2 monolayers by Mie-resonantmetasurfaces,” ACS Photonics, vol. 6, pp. 1002−1009, 2019..[25] H. Li, J. Wang, Y. Ma, et al., “Enhanced directional emission ofmonolayer tungsten disulfide (WS2) with robust linear polarizationvia one-dimensional photonic crystal (PhC) slab,” Nanophotonics,vol. 9, no. 14, pp. 4337−4345, 2020..[26] J. Yan, Z. Zheng, Z. Lou, J. Li, B. Mao, and B. Li, “Enhancement ofexciton emission in WS2 based on the Kerker effect from the modeengineering of individual Si nanostripes,” Nanoscale Horiz., vol. 5,no. 10, pp. 1368−1377, 2020..[27] S. I. Azzam, K. Parto, and G. Moody, “Purcell enhancement andpolarization control of single-photon emitters in monolayer WSe2using dielectric nanoantennas,” Nanophotonics, vol. 12, no. 3,pp. 477−484, 2023..[28] S. Wu, S. Buckley, J. R. Schaibley, et al., “Monolayer semiconductornanocavity lasers with ultralow thresholds,” Nature, vol. 520,pp. 69−72, 2015..[29] P. Yang, X. Zou, Z. Zhang, et al., “Batch production of 6-inchuniform monolayer molybdenum disulfide catalyzed by sodium inglass,” Nat. Commun., vol. 9, no. 1, p. 979, 2018..[30] X. Yang, S. Li, N. Ikeda, and Y. Sakuma, “Oxide scale sublimationchemical vapor deposition for controllable growth of monolayerMoS2 crystals,” Small Methods, vol. 6, no. 2, p. 202101107, 2022..[31] C.-C. Huang, H. Wang, Y. Cao, et al., “Facilitating uniformlarge-scale MoS2, WS2 monolayers, and their heterostructuresthrough van der Waals epitaxy,” ACS Appl. Mater. Interfaces, vol. 14,no. 37, pp. 42365−42373, 2022..[32] X. Yang, S. Li, N. Ikeda, A. Ohtake, and Y. Sakuma, “Scalable growthof atomically thin MoS2 layers in a conventional MOCVD systemusing molybdenum dichloride dioxide as the molybdenumsource,” Appl. Surf. Sci., vol. 636, no. 1, p. 157756, 2023..[33] M. Iwanaga, “All-dielectric metasurfaces withhigh-fluorescence-enhancing capability,” Appl. Sci., vol. 8, no. 8,p. 1328, 2018..[34] M. Iwanaga, “All-dielectric metasurface fluorescence biosensorsfor high-sensitivity antibody/antigen detection,” ACS Nano, vol. 14,no. 12, pp. 17458−17467, 2020..[35] M. Iwanaga, “High-sensitivity high-throughput detection ofnucleic-acid targets on metasurface fluorescence biosensors,”Biosensors, vol. 11, no. 2, p. 33, 2021..[36] M. Iwanaga and W. Tangkawsakul, “Two-way detection ofCOVID-19 spike protein and antibody using all-dielectricmetasurface fluorescence sensors,” Biosensors, vol. 12, no. 11,p. 981, 2022..[37] M. Iwanaga, “Rapid detection of attomolar SARS-CoV-2 nucleicacids in all-dielectric metasurface biosensors,” Biosensors, vol. 12,no. 11, p. 987, 2022..[38] M. Iwanaga, T. Hironaka, N. Ikeda, T. Sugasawa, and K. Takekoshi,“Metasurface biosensors enabling single-molecule sensing ofcell-free DNA,” Nano Lett., vol. 23, no. 12, pp. 5755−5761, 2023..[39] M. Iwanaga and B. Choi, “Heteroplasmon hybridization in stackedcomplementary plasmo−photonic crystals,” Nano Lett., vol. 15,no. 3, pp. 1904−1910, 2015..[40] B. Choi, M. Iwanaga, H. T. Miyazaki, Y. Sugimoto, A. Ohtake, and K.Sakoda, “Overcoming metal-induced fluorescence quenching onplasmo-photonic metasurfaces coated by a self-assembledmonolayer,” Chem. Commun., vol. 51, no. 57, pp. 11470−11473, 2015..M. Iwanaga et al.: Prominently enhanced luminescence from a TMDC monolayer on metasurfaces — 105[41] M. Iwanaga, B. Choi, H. T. Miyazaki, and Y. Sugimoto, “The artificialcontrol of enhanced optical processes in fluorescent molecules onhigh-emittance metasurfaces,” Nanoscale, vol. 8, no. 21,pp. 11099−11107, 2016..[42] M. Iwanaga, “Highly sensitive wide-range target fluorescencebiosensors of high-emittance metasurfaces,” Biosens. Bioelectron.,vol. 190, no. 1, p. 113423, 2021..[43] K. Momma and F. Izumi, “Vesta 3 for three-dimensionalvisualization of crystal, volumetric and morphology data,” J. Appl.Crystallogr., vol. 44, pp. 1272−1276, 2011..[44] R. Loudon, The Quantum Theory of Light, 3rd ed. Oxford, UK, OxfordUniv. Press, 2000, chap. 5.[45] T. Kuroda, Y. Hoshi, S. Masubuchi, et al., “Dark-state impact on theexciton recombination of WS2 monolayers as revealed bymulti-timescale pump-probe spectroscopy,” Phys. Rev. B, vol. 102,no. 19, p. 195407, 2020..[46] F. Wu, F. Qu, and A. H. MacDonald, “Exciton band structure ofmonolayer MoS2,” Phys. Rev. B, vol. 91, no. 7, p. 075310, 2015..[47] R. R. Chance, A. Prock, and R. Silbey, “Molecular fluorescence andenergy transfer near interfaces,” Adv. Chem. Phys., vol. 37,pp. 1−65, 1978.[48] M. Iwanaga, Plasmonic Resonators: Fundamentals, Advances, andApplications, Singapore, Pan Stanford Publishing, 2016.[49] E. M. Purcell, “Spontaneous emission probabilities at radiofrequencies,” Phys. Rev., vol. 69, p. 681, 1946.[50] B. Munkhbat, P. Wróbel, T. J. Antosiewicz, and T. O. Shegai,“Optical constants of several multilayer transition metaldichalcogenides measured by spectroscopic ellipsometry in the300−1700 nm range: high index, anisotropy, and hyperbolicity,”ACS Photonics, vol. 9, no. 7, pp. 2398−2407, 2022..[51] L. Li, “New formulation of the Fourier modal method for crossedsurface-relief gratings,” J. Opt. Soc. Am. A, vol. 14, no. 10,pp. 2758−2767, 1997..[52] L. Li, “Formulation and comparison of two recursive matrixalgorithm for modeling layered diffraction gratings,” J. Opt. Soc.Am. A, vol. 13, no. 5, pp. 1024−1035, 1996..[53] E. D. Palik, Handbook of Optical Constants of Solids II, San Diego,USA, Academic, 1991.Supplementary Material: This article contains supplementary material(https://doi.org/10.1515/nanoph-2023-0672).https://doi.org/10.1515/nanoph-2023-0672 1 Introduction 2 Results and discussion 2.1 Structures of TMDC-transferred metasurfaces 2.2 Optical resonances of the metasurface 2.3 Enhanced PL spectrum and image 2.4 Decay-time analysis of enhanced PL 2.5 Resonantly enhanced PL processes 2.6 PL EF 3 Conclusions 4 Experimental section 4.1 TMDC growth 4.2 Transfer process onto metasurfaces 4.3 Nanofabrication of metasurfaces 4.4 Optical measurement 4.5 Computation of reflectance spectra and EM-field distributions 4.6 Numerical evaluation of purcell factor Supplementary Material<<  /ASCII85EncodePages false  /AllowTransparency false  /AutoPositionEPSFiles true  /AutoRotatePages /None  /Binding /Left  /CalGrayProfile (Dot Gain 20%)  /CalRGBProfile (sRGB IEC61966-2.1)  /CalCMYKProfile (Euroscale Coated v2)  /sRGBProfile (sRGB IEC61966-2.1)  /CannotEmbedFontPolicy /Warning  /CompatibilityLevel 1.7  /CompressObjects /Tags  /CompressPages true  /ConvertImagesToIndexed true  /PassThroughJPEGImages false  /CreateJobTicket false  /DefaultRenderingIntent /Default  /DetectBlends true  /DetectCurves 0.1000  /ColorConversionStrategy /sRGB  /DoThumbnails true  /EmbedAllFonts true  /EmbedOpenType false  /ParseICCProfilesInComments true  /EmbedJobOptions true  /DSCReportingLevel 0  /EmitDSCWarnings false  /EndPage -1  /ImageMemory 1048576  /LockDistillerParams false  /MaxSubsetPct 35  /Optimize true  /OPM 1  /ParseDSCComments true  /ParseDSCCommentsForDocInfo true  /PreserveCopyPage true  /PreserveDICMYKValues true  /PreserveEPSInfo true  /PreserveFlatness false  /PreserveHalftoneInfo false  /PreserveOPIComments false  /PreserveOverprintSettings true  /StartPage 1  /SubsetFonts true  /TransferFunctionInfo /Apply  /UCRandBGInfo /Remove  /UsePrologue false  /ColorSettingsFile ()  /AlwaysEmbed [ true  ]  /NeverEmbed [ true  ]  /AntiAliasColorImages false  /CropColorImages false  /ColorImageMinResolution 300  /ColorImageMinResolutionPolicy /OK  /DownsampleColorImages true  /ColorImageDownsampleType /Bicubic  /ColorImageResolution 300  /ColorImageDepth -1  /ColorImageMinDownsampleDepth 1  /ColorImageDownsampleThreshold 1.50000  /EncodeColorImages true  /ColorImageFilter /DCTEncode  /AutoFilterColorImages true  /ColorImageAutoFilterStrategy /JPEG  /ColorACSImageDict <<    /QFactor 0.15    /HSamples [1 1 1 1] /VSamples [1 1 1 1]  >>  /ColorImageDict <<    /QFactor 0.15    /HSamples [1 1 1 1] /VSamples [1 1 1 1]  >>  /JPEG2000ColorACSImageDict <<    /TileWidth 256    /TileHeight 256    /Quality 10  >>  /JPEG2000ColorImageDict <<    /TileWidth 256    /TileHeight 256    /Quality 30  >>  /AntiAliasGrayImages false  /CropGrayImages false  /GrayImageMinResolution 300  /GrayImageMinResolutionPolicy /OK  /DownsampleGrayImages true  /GrayImageDownsampleType /Bicubic  /GrayImageResolution 300  /GrayImageDepth -1  /GrayImageMinDownsampleDepth 2  /GrayImageDownsampleThreshold 1.50000  /EncodeGrayImages true  /GrayImageFilter /DCTEncode  /AutoFilterGrayImages true  /GrayImageAutoFilterStrategy /JPEG  /GrayACSImageDict <<    /QFactor 0.15    /HSamples [1 1 1 1] /VSamples [1 1 1 1]  >>  /GrayImageDict <<    /QFactor 0.15    /HSamples [1 1 1 1] /VSamples [1 1 1 1]  >>  /JPEG2000GrayACSImageDict <<    /TileWidth 256    /TileHeight 256    /Quality 30  >>  /JPEG2000GrayImageDict <<    /TileWidth 256    /TileHeight 256    /Quality 30  >>  /AntiAliasMonoImages false  /CropMonoImages false  /MonoImageMinResolution 600  /MonoImageMinResolutionPolicy /OK  /DownsampleMonoImages true  /MonoImageDownsampleType /Bicubic  /MonoImageResolution 1000  /MonoImageDepth -1  /MonoImageDownsampleThreshold 1.10000  /EncodeMonoImages true  /MonoImageFilter /CCITTFaxEncode  /MonoImageDict <<    /K -1  >>  /AllowPSXObjects false  /CheckCompliance [    /None  ]  /PDFX1aCheck false  /PDFX3Check false  /PDFXCompliantPDFOnly false  /PDFXNoTrimBoxError false  /PDFXTrimBoxToMediaBoxOffset [    0.00000    0.00000    0.00000    0.00000  ]  /PDFXSetBleedBoxToMediaBox true  /PDFXBleedBoxToTrimBoxOffset [    0.00000    0.00000    0.00000    0.00000  ]  /PDFXOutputIntentProfile (None)  /PDFXOutputConditionIdentifier ()  /PDFXOutputCondition ()  /PDFXRegistryName ()  /PDFXTrapped /False  /CreateJDFFile false  /Description <<    /DEU <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>    /ENU ()    /ENN ()  >>  /Namespace [    (Adobe)    (Common)    (1.0)  ]  /OtherNamespaces [    <<      /AsReaderSpreads false      /CropImagesToFrames true      /ErrorControl /WarnAndContinue      /FlattenerIgnoreSpreadOverrides false      /IncludeGuidesGrids false      /IncludeNonPrinting false      /IncludeSlug false      /Namespace [        (Adobe)        (InDesign)        (4.0)      ]      /OmitPlacedBitmaps false      /OmitPlacedEPS false      /OmitPlacedPDF false      /SimulateOverprint /Legacy    >>    <<      /AllowImageBreaks true      /AllowTableBreaks true      /ExpandPage false      /HonorBaseURL true      /HonorRolloverEffect false      /IgnoreHTMLPageBreaks false      /IncludeHeaderFooter false      /MarginOffset [        0        0        0        0      ]      /MetadataAuthor ()      /MetadataKeywords ()      /MetadataSubject ()      /MetadataTitle ()      /MetricPageSize [        0        0      ]      /MetricUnit /inch      /MobileCompatible 0      /Namespace [        (Adobe)        (GoLive)        (8.0)      ]      /OpenZoomToHTMLFontSize false      /PageOrientation /Portrait      /RemoveBackground false      /ShrinkContent true      /TreatColorsAs /MainMonitorColors      /UseEmbeddedProfiles false      /UseHTMLTitleAsMetadata true    >>    <<      /AddBleedMarks false      /AddColorBars false      /AddCropMarks false      /AddPageInfo false      /AddRegMarks false      /BleedOffset [        0        0        0        0      ]      /ConvertColors /ConvertToCMYK      /DestinationProfileName (ISO Coated v2 \(ECI\))      /DestinationProfileSelector /UseName      /Downsample16BitImages true      /FlattenerPreset <<        /ClipComplexRegions true        /ConvertStrokesToOutlines false        /ConvertTextToOutlines false        /GradientResolution 300        /LineArtTextResolution 1200        /PresetName <FEFF005B0048006F006800650020004100750066006C00F600730075006E0067005D>        /PresetSelector /HighResolution        /RasterVectorBalance 1      >>      /FormElements true      /GenerateStructure false      /IncludeBookmarks false      /IncludeHyperlinks false      /IncludeInteractive false      /IncludeLayers false      /IncludeProfiles false      /MarksOffset 8.503940      /MarksWeight 0.250000      /MultimediaHandling /UseObjectSettings      /Namespace [        (Adobe)        (CreativeSuite)        (2.0)      ]      /PDFXOutputIntentProfileSelector /UseName      /PageMarksFile /RomanDefault      /PreserveEditing true      /UntaggedCMYKHandling /LeaveUntagged      /UntaggedRGBHandling /UseDocumentProfile      /UseDocumentBleed false    >>  ]>> setdistillerparams<<  /HWResolution [600 600]  /PageSize [595.276 841.890]>> setpagedevice