# Fileset

[s41467-025-57391-8.pdf](https://mdr.nims.go.jp/filesets/36bc60ac-8e04-4dcf-bc08-961f44c92a5a/download)

## Creator

Qinghai Tan, Abdullah Rasmita, Zhaowei Zhang, Xuran Dai, Ruihua He, Xiangbin Cai, [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), Hongbing Cai, Wei-bo Gao

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Enhanced coherence from correlated states in WSe2/MoS2 moiré heterobilayer](https://mdr.nims.go.jp/datasets/3b361976-62e8-4b5a-8ab2-575d77d35889)

## Fulltext

Enhanced coherence from correlated states in WSe2/MoS2 moiré heterobilayerArticle https://doi.org/10.1038/s41467-025-57391-8Enhanced coherence from correlated statesin WSe2/MoS2 moiré heterobilayerQinghai Tan1,2,9 , Abdullah Rasmita 1,3,9, Zhaowei Zhang 1, Xuran Dai1,Ruihua He1, Xiangbin Cai1, Kenji Watanabe 4, Takashi Taniguchi 5,Hongbing Cai 1,6 & Wei-bo Gao 1,3,7,8Moiré superlattices in van der Waals materials have emerged as a promisingplatform for studying the correlated states in condensedmatter physics. Thesecorrelated states have substantial effects on the emission coherence, animportant parameter for quantum light applications. However, the effect ofcorrelated states on the excitonic emission coherence is largely unexplored.Here, we show that the coherence of moiré interlayer exciton emission intungsten diselenide (WSe2)/molybdenum disulfide (MoS2) heterobilayers issensitive to the correlated insulating states in this material. We demonstratethat the emission linewidth of interlayer exciton shows a dip at a particularpower range, which we attributed to the excitonic (bosonic) interaction.Moreover, such linewidthminima also appear in thedopingdependence of thephotoluminescence spectrum at the integer electronic filling factor, fel = 1,demonstrating the effect of the electronic (fermionic) correlated insulatingstate on the interlayer exciton emission coherence. Our results demonstratethe richness of exciton-exciton and exciton-electron interactions in moirésemiconductors and pave the way for engineering emission coherence bycontrolling such interactions.Understanding the impact of correlated states of particles on mate-rial properties is a crucial step in advancing solid-state devices1,2.Correlated states occur when the behaviours and properties ofmultiple particles, including fermions (e.g., electrons) and bosons(e.g., excitons (bound electron-hole pairs)), are interdependent.These states are related to many fascinating phenomena, includingsuperconductivity3,4 and correlated insulators5,6.One notable consequence of particle correlation is that theiremission will be correlated or synchronized. This behaviour modifiesthe overall coherence of the total emission from thematerial7,8. Hence,by examining the coherence of the emitted light9–12, we can directlyprobe the particle correlation unambiguously. Coherence is also animportant figure of merit for many applications, including laser andquantum light sources. Hence, it is crucial to understand the connec-tion between coherence and correlated state in solid-state physics.One of the promising platforms for performing such a study is thesemiconductor two-dimensional (2D) moiré superlattice based ontransition metal dichalcogenide (TMD)13–17. The in-plane moiré super-lattice is formed by the interference pattern resulting from the inter-action between two or more stacked monolayer TMD materials. TheReceived: 21 December 2024Accepted: 20 February 2025Check for updates1Division of Physics andApplied Physics, School of Physical andMathematical Sciences, Nanyang Technological University, Singapore, Singapore. 2School ofMicroelectonics, University of Science and Technology of China, Hefei, China. 3School of Electrical and Electronic Engineering, Nanyang TechnologicalUniversity, Singapore, Singapore. 4ResearchCenter for FunctionalMaterials, National Institute forMaterials Science, Tsukuba, Japan. 5International Center forMaterials Nanoarchitectonics, National Institute for Materials Science, Tsukuba, Japan. 6Hefei National Laboratory for Physical Sciences at the Microscale &Synergetic Innovation Center of Quantum Information & Quantum Physics, University of Science and Technology of China, Hefei Anhui, China. 7Centre forQuantum Technologies, Nanyang Technological University, Singapore, Singapore. 8Quantum Science and Engineering Centre, Nanyang TechnologicalUniversity, Singapore, Singapore. 9These authors contributed equally: Qinghai Tan, Abdullah Rasmita. e-mail: tanqh@ustc.edu.cn; coldice@ustc.edu.cn;wbgao@nstu.edu.sgNature Communications |         (2025) 16:4518 11234567890():,;1234567890():,;http://orcid.org/0000-0002-0669-7131http://orcid.org/0000-0002-0669-7131http://orcid.org/0000-0002-0669-7131http://orcid.org/0000-0002-0669-7131http://orcid.org/0000-0002-0669-7131http://orcid.org/0000-0002-0603-6763http://orcid.org/0000-0002-0603-6763http://orcid.org/0000-0002-0603-6763http://orcid.org/0000-0002-0603-6763http://orcid.org/0000-0002-0603-6763http://orcid.org/0000-0003-3701-8119http://orcid.org/0000-0003-3701-8119http://orcid.org/0000-0003-3701-8119http://orcid.org/0000-0003-3701-8119http://orcid.org/0000-0003-3701-8119http://orcid.org/0000-0002-1467-3105http://orcid.org/0000-0002-1467-3105http://orcid.org/0000-0002-1467-3105http://orcid.org/0000-0002-1467-3105http://orcid.org/0000-0002-1467-3105http://orcid.org/0000-0003-3186-1041http://orcid.org/0000-0003-3186-1041http://orcid.org/0000-0003-3186-1041http://orcid.org/0000-0003-3186-1041http://orcid.org/0000-0003-3186-1041http://orcid.org/0000-0003-3971-621Xhttp://orcid.org/0000-0003-3971-621Xhttp://orcid.org/0000-0003-3971-621Xhttp://orcid.org/0000-0003-3971-621Xhttp://orcid.org/0000-0003-3971-621Xhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-025-57391-8&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-025-57391-8&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-025-57391-8&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1038/s41467-025-57391-8&domain=pdfmailto:tanqh@ustc.edu.cnmailto:coldice@ustc.edu.cnmailto:wbgao@nstu.edu.sgwww.nature.com/naturecommunicationslocalization induced by this superlattice can result in flat bands18–20,enhancing the effect of the particle correlation on the materialbehavior21,22. Moreover, the doping and electric field in the 2Dmaterialcan be controlled precisely using electrical gating23, enabling the studyof correlated behaviour in a large range of particle density and electricfield24–27. This versatility makes it a promising tool for the quantumsimulation of correlated states, including the effectof correlated stateson emission coherence.Here, by performing excitation power, temperature, and dopingdependence study, we experimentally demonstrate the effect of cor-related state on the coherence of interlayer exciton (IX, i.e., boundelectron-hole pair with electron and hole located in different layers)photoluminescence (PL) emission in WSe2/MoS2 heterobilayer. Fromthe power-dependent PL spectrum and coherence timemeasurement,we find that the IX PL spectrum linewidth decreases while its coher-ence time increases within a particular power range, indicating theeffect of excitonic interaction28–31. The observed linewidth dip persistsuntil a temperature of 50 K, above which the phonon-induced scat-tering suppresses the linewidth reduction. The linewidth reduction isalso observed when the carrier doping, instead of excitation power, isincreased until the linewidth reaches the minimum value at the fillingfactor fel = 1, i.e., one electron has been added to eachmoiré cell. Thesefindings demonstrate the effect of excitonic and electronic correlationon the IX emission coherence.ResultsDevice structure and conceptual overviewThe device structure is shown in Fig. 1a, while the optical image of thefabricated device (Device 1) is shown in Fig. 1b (see Methods for thefabrication detail). We adopt the dual capacitor structure, allowingcarrier doping control by gate voltages. The heterobilayer consists ofmonolayer WSe2 stacked onmonolayer MoS2. The MoS2 andWSe2 areconnected to a common electrical ground. The second harmonicgeneration (SHG) signal (Supplementary Fig. S1a) in the hetero-structure region is suppressed, indicating that the stacking is closer toAB (i.e., a 60o twist angle) stacking32. By performing polarization-resolved SHG (Supplementary Fig. S1b), we determine that the twistangle is close to AB stacking within ±1o. The heterobilayer is placedbetween twodielectric hexagonal boronnitride (hBN) layers, eachwitha thicknessof around 11 to 12 nm (Supplementary Fig. S1c). The top andbottom graphite electrodes sandwich the hBN layers, completing thecapacitor structure. One voltage source (Vg) is connected to bothelectrodes for controlling carrier doping.Due to the different lattice constant between MoS2 and WSe2, amoiré superlattice is created even well-aligned19,33. The superlatticepotential induces not only electron localization but also excitonlocalization34, creatingmultiplemoiré IX energy levels (Fig. 1c)35–38. Theinteraction between the moiré IX and the moiré-induced quasiparticlelattice (i.e., either electron or exciton lattice) can affect themoiré IX PLemission29–31,39–41, such as the coherence (Fig. 1d, e). Our particularinterest is the effect of the superlattice on the high energy moiré IXs.These moiré IXs are less confined compared to the low-energy IX.Hence, they are more affected by the surrounding environment. Thus,their emission properties, including the linewidth and coherence, areexpected to be sensitive to correlated states involving multipleparticles.Enhanced IX coherence via IX-IX interactionWe then study the evolution of the PL spectrum when the excitationpower is varied (Fig. 2a). At high enough excitation power (above8μW), we observed the emergence of the high-energy IX peak (label-led as P2), whose emission dominates the PL signal at higher power(above 34μW). Similar results are obtained from another bilayersample (Supplementary Fig. S2a). Power dependence of the PLGroundTop gateWSe2/MoS2Bottom gateWSe2hBNTop gateBottom gatehBNMoS2VgGroundInteractionMoiré IXQuasiparticle latticeHigh coherenceab cdeLow coherenceFig. 1 | Device structure and theoretical physics picture. a Device structure. Theblack, blue, purple, and green layers indicate graphite electrodes, hBN, WSe2, andMoS2, respectively. One voltage source (Vg) is connected to both electrodes tocontrol the carrier doping. bOptical image of the fabricated device (Device 1). Thesolid line illustrates the heterostructure region. The dotted lines illustrate thegraphite region as top gate, bottom gate and ground. The scale bar is 10μm.c Illustration of interaction between moiré interlayer exciton (IX) and moiré-induced quasiparticle lattice. The linewidth of the high-energy moiré IX can beaffected by the interaction.d, eA schematic ofmoiré IX emissions in lowcoherence(d) and high coherence (e). The temporal coherence of the moiré IX is affected bythe correlated states. The different coloured arrows illustrate the IX emissions atdifferent energies.Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 2www.nature.com/naturecommunicationsspectrum under different excitation wavelengths also shows a similarbehaviour (Supplementary Fig. S3). More discussion on the origin ofthe P2 emission is given in Supplementary Note 1.In this study, we focus on the high-energy P2 emission. Figure 2bshows the power-dependent linewidth of the P2 emission obtainedfrom two-peakGaussianfitting of the PL spectra (see inset of Fig. 2b forthe fitting result at 34μW). The P2 linewidth decreases from 19.3meVat ~34μWto 17.6meV at ~110μW, resulting in a linewidth dip at the 100to 150μW power range. Such behaviour is also observed from a dif-ferent excitation location (Supplementary Fig. S4) as well as fromanother sample (Supplementary Fig. S2b). To confirm the linewidthnarrowing of the P2 emission, we measured the coherence time of theP2 emission (coming from the high-energy moiré IX decay) using aMach-Zehnder interferometer setup with an installed delay line in oneof the arms (see inset of Fig. 2c). Optical filters were applied before theinterferometer to select the P2 emission (see Methods for measure-ment setup detail). We recorded the intensity as a function of the timedelay and extracted the coherence time for various excitation powers(Supplementary Fig. S5). As shown in Fig. 2c, we observed a peak in thepower-dependent coherence time at the power range where the line-width narrowing happens. Considering that the coherence time isinversely proportional to the linewidth, such observation confirms thelinewidth narrowing at this power range.The observation of the linewidth dip above can be understood byconsidering the power dependence of the spectrum broadening. Thelinewidth of the PL emission (wPL) results from inhomogeneous andhomogeneousbroadening42,43. Considering the IX long lifetimeof ~4 ns(see Supplementary Fig. S6a for the time-resolved PL intensity), the IXhomogeneous linewidth ismainly affected by dephasing43,44 instead oflifetime. Hence, wPL can be expressed aswPL �ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiwdep2 +winh2q, ð1Þwhere wdep is the linewidth due to the dephasing process and winh isthe inhomogeneous linewidth.Figure 2d illustrates the effect of IX density on the inhomoge-neous broadening. The disorder potential induces a spatial variationofthe IX energy, resulting in inhomogeneous broadening of the PLemission linewidth. As the power increases, the exciton populationwillaccumulate at the local minimum of the disorder potential. Con-sidering IX-IX repulsion, such accumulation results in a larger blue shiftat the disorder minima location than the average blueshift. Conse-quently, the linewidth becomes narrower with increasing power45–47.However, the IX-IX repulsion also results in excitation-induceddephasing44. This dephasing results in linear-in-density linewidthbroadening, which compete with the linewidth narrowing effectdescribed before. The combinationof these twomechanisms results inthe linewidth dip observed in the experiment.Temperature and power effects on IX linewidthTo further check the validity of this reasoning, we conducted a tem-perature dependence study and compared the result with the theo-retical model. Figure 3a shows the PL spectra at constant excitationpower ( ~ 100 μW) and temperatures varying from 5K to 72 K. The PL0 0.5 1Normalized intensity T = 5 KacDelaySource DetectorP1 P2P1P2bdehehehehehehenergyspaceLow powerHigh powerDisorder potentialDisorder potentialFig. 2 | Temporal coherence probing of excitonic interaction. a Power-dependent IX photoluminescence (PL) spectrum from Device 1. The low- and high-energy moiré IXs are labelled as P1 and P2, respectively. b Linewidth vs excitationpower for P2 emission. Inset: fitted PL spectrum under 730nm 34μW excitation.cCoherence time vs excitationpowerof the P2emission. Inset: temporal coherencemeasurement setup. The arrowed lines show the two paths taken by the light. In(b, c), the symbols are the measured values, and the shaded area represents thecoherence time uncertainty. The line is the guide for the eye. d Schematic ofinhomogeneous broadening reduction by IX-IX repulsion. The dashed grey andsolid black lines are the disorderpotential at zero andfinite IX density, respectively.The disorder potential is screened by the increasing IX density at a higher power.The red and white circles labelled as ‘e’ and ‘h’ represent the electron and hole,respectively.Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 3www.nature.com/naturecommunicationsspectrum shows a redshift, which can be attributed to thetemperature-induced bandgap change. Figure 3b shows that the line-width dip gets smaller with increasing temperature and eventuallydisappears at 50K. This phenomenon can be understood by con-sidering that the inhomogeneous linewidth reduction mechanismrelies on the IX accumulation at the disorder potential minimum,which depends on IX density and mobility. For the 5 K – 72 K tem-perature range, the IX density does not change much with tempera-ture, as shown by the relatively similar PL intensity and the powersaturation curve in this temperature range (Supplementary Fig. S7).Hence, the primarymechanismbehind the temperaturedependence isIX mobility reduction with increasing temperature47. We developed atheoretical model considering the temperature dependence of themoiré IX mobility (Supplementary Note 2). As shown in Fig. 3b, themodel fits the data well, further supporting the physical picture.Enhanced IX coherence at correlated electronic stateFinally, we performed a gate (doping) dependence experiment to studythe effect of electron population on the IX linewidth, focusing on elec-tron doping case. The gate dependence of the PL spectrum is shown inFig. 4a. We then extracted the peak energy and the linewidth of the P2emission from these PL spectra. As shown in Fig. 4b, the linewidth showsa dip, while the peak energy shows a peak at fel = 1 (see SupplementaryNote 3 for discussion on charge density and filling factor). The linewidthreduction at the integer filling factor suggests that the correlated elec-trons affect the IX-IX interaction. This phenomenon can be understoodby considering the increase in IX PL intensity at integer filling factors(see Supplementary Fig. S8). The intensified PL signal can be attributedto a longer IX lifetime41, which in turn leads to a higher IX density even ata constant excitation power. Considering the IX-IX interaction, theincreased IX density at fel = 1 results in a reduction in inhomogeneousbroadening and also contributes to IX energy blueshift. From the PLmeasurement, we found that the P2 emission intensity is enhanced byaround 3 times at fel = 1. Based on the theoretical calculation, thisenhancement reduces the linewidth from 18.4meV to 14.1meV, whichagrees well with the experimental results shown in Fig. 4b. A similarphenomenon is also observed in other integer electron filling factors(see Supplementary Note 4). More discussion on the doping depen-dence of the PL spectrum and linewidth, including the contribution ofthe change in dielectric constant14,48 and charge fluctuation, is given inSupplementary Note 1 and 2.DiscussionIn summary, we have demonstrated the effect of excitation power andelectrical doping on the IX PL emission linewidth. By studying the IX PLlinewidth evolution, we have shown that the IX-IX interaction isaffected by both excitonic interaction and electronic correlation.These results could be used to design nonlinear excitonic devicesFig. 4 | Temporal coherence probing of electronic correlation. a PL spectra atvarious gate voltages at 21.7μW excitation power from Device 1. b P2 peak energyand linewidth vs gate voltage. At fel = 1, the PL linewidth shows a dipwhile its energyshows a peak behaviour. The symbols are themeasurement results, and the shadedarea represents the parameter uncertainty obtained from two-peakGaussian fittingof the PL spectrum.The lines are guides for the eye. c Schematic of inhomogeneousbroadening reduction by correlated electronic state. The longer lifetime of IX at fel= 1 results in increased IX density, which in turn reduces the inhomogeneousbroadening.Fig. 3 | Temperature dependence of linewidth reduction. a Temperaturedependence of the PL spectrum from Device 1. The excitation power used is~100μW. b P2 linewidth vs power at various temperatures. The dots and lines arethe data and the fitting with the theoretical model, respectively. The shaded arearepresents the linewidth uncertaintyobtained from the two-peakGaussianfittingofthe PL spectrum.Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 4www.nature.com/naturecommunicationswhichcanbe controlledbyboth excitation power andelectrical gating.We show that it is possible to reduce the IX linewidth by utilizingcorrelated states. Reducing the linewidth can reduce the couplingstrength required to reach lasing49,50 or the strong-coupling regimebetween IXs and other quasiparticles (including other excitons andphotons)51, where strong nonlinear behaviour can be observed. Fur-thermore, the gate tunability of the linewidth shows the possibility ofcontrolling the cross-over from linear to nonlinear regime using gatevoltages (see also Supplementary Fig. S10).Our demonstration of the long-range IX-IX interaction effect on thehigh-energymoiré IX opens possible optical manipulation of the higher-energymoiré band.We note that the high-energymoiré IX becomes thelowest IX energy level at electron full filling (see Fig. 4a). Ourwork showsthat the disorder effect on this IX can be minimized, which is beneficialin realizingmany-body excitonic state52–54, such as exciton Bose-Einsteincondensate. Additionally, compared to the low-energy IX, the high-energy moiré IX is affected more by the higher-energy moiré bands.Hence, utilizing the IX-electron interaction involving the higher-energymoiré bands and the optical addressability of this IX, it could be possibleto detect and manipulate the topological properties of the higher-energy moiré band optically. Such properties play a crucial role, forexample, concerning the non-Abelian anyon in moiré superlattices55.Finally, our result of electrically tunable linewidth demonstratesthat exciton coherence in the high-energy moiré band is stronglyinfluenced by the correlated electrons in the lower-energymoiré band.The interaction between multiple moiré IX and electronic bands canlead to novel correlated states involving exciton and electron,including quantum interference between these quasiparticles. Whilehere we focus on the PL linewidth, the rich exciton-exciton andexciton-electron interaction can also affect the lineshape. For example,it could be possible to electrically drive two different excitonic statesinto a resonance condition, resulting in asymmetric Fano-like PLlineshape56 (see Supplementary Note 1 for more discussion). Ourfinding shows that the coherence of IX emission is an indispensabletool to uncover phenomena related to correlated excitonic (bosonic)and electronic (fermionic) states in the 2D heterostructure.MethodsSample fabricationMonolayers WSe2 and MoS2, few-layer graphite (FLG), and thin hBNwere first mechanically exfoliated from their bulk crystal. Subse-quently, we used a layer-by-layer dry-transfer technique to prepareMoS2/WSe2 heterostructures onto the SiO2/Si substrate with ultralowdoping Si. FLG was used to contact the heterostructure (as ground)and top/bottom gates, and thin hBN (around 11 to 12 nm thick, seeSupplementary Fig. S1c) was used as the dielectric layers. The crystalorientation ofmonolayerWSe2 andMoS2was determined by the sharpedge of samples57. We then annealed the samples under an ultrahighvacuum (around 10−7 mbar) at 200 °C for 3 hours.CW and time-resolved PL measurementsAll PL spectra under CW excitation and time-resolved PL measurementswere performed in a home-built confocal optical microscope in thereflection geometry. The PL spectra were obtained by using a spectro-meter (Princeton) with a liquid nitrogen-cooled charge-coupled device(CCD) InGaAs photodetector (range: 800nm–1600nm). A 50× long-work distance infrared objective lens (LCPLN50XIR) with a spot size ofaround 1 µm [numerical aperture (NA) =0.65] was used. The sampletemperature was cooled/controlled using a cryostat (Montana Instru-ments and Attocube, Attodry 2100). A 726nm pulse laser (sourceFWHM<80ps with amaximum repetition rate of 80MHz) was used fortime-resolved IX PL measurement, and a 730nm CW laser was used forobtaining PL spectra. An infrared superconducting single-photondetector (SSPD, range: up to 1550nm) was used as a photodetectorfor time-resolved PL and photon count measurement. For energy- andtime-resolved PLmeasurement, a 950nm long pass filter combinedwith1200nm short pass (for P1 emission measurement) or 1200nm longpass filters (for P2 emission measurement) were used to cut the signal.Time-coherence measurementThe time coherence measurements were performed in a home-builtcoherence system.AnSSPDwith a rangeofup to 1550nmwasused as aphotodetector.SHG measurementThe SHGmeasurementswereperformed at room temperature. A pulselaser from a Ti:Sapphire oscillator (Spectra Physics, Tsunami) with apeak at around 880 nm, a repetition rate of 80MHz, and a pulseduration of 100 fs was used as the excitation source. The laser powerwas around 0.8mW.Data availabilityRelevant data supporting the key findings of this study are availablewithin the article and the Supplementary Information file. All raw datagenerated during the current study are available from the corre-sponding authors upon request.References1. Basov, D. N., Averitt, R. D., van der Marel, D., Dressel, M. & Haule, K.Electrodynamics of correlated electron materials. Rev. Mod. Phys.83, 471–541 (2011).2. Quintanilla, J. & Hooley, C. The strong-correlations puzzle. Phys.World 22, 32 (2009).3. Cao, Y. et al. Unconventional superconductivity in magic-anglegraphene superlattices. Nature 556, 43–50 (2018).4. Bednorz, J. G. & Müller, K. A. Possible highTc superconductivity inthe Ba−La−Cu−O system. Z. Phys. B 64, 189–193 (1986).5. Boer, J. H. D. & Verwey, E. J. W. Semi-conductors with partially andwith completelyfilled3d-lattice bands.Proc. Phys. Soc.49, 59 (1937).6. Cao, Y. et al. Correlated insulator behaviour at half-filling in magic-angle graphene superlattices. Nature 556, 80–84 (2018).7. Dicke, R. H. Coherence in spontaneous radiation processes. Phys.Rev. 93, 99–110 (1954).8. Laikhtman, B. & Rapaport, R. Exciton correlations in coupledquantum wells and their luminescence blue shift. Phys. Rev. B 80,195313 (2009).9. Richard, M. et al. Experimental evidence for nonequilibrium Bosecondensation of exciton polaritons. Phys. Rev. B 72, 201301 (2005).10. Anankine, R. et al. Temporal coherenceof spatially indirect excitonsacross Bose–Einstein condensation: the role of free carriers. New J.Phys. 20, 073049 (2018).11. Lagoin, C., Suffit, S., Baldwin, K., Pfeiffer, L. & Dubin, F. Mott insu-lator of strongly interacting two-dimensional semiconductor exci-tons. Nat. Phys 18, 149–153 (2022).12. Sigl, L. et al. Signatures of a degenerate many-body state of inter-layer excitons in a van der Waals heterostack. Phys. Rev. Res 2,042044 (2020).13. Tang, Y. et al. Simulation of Hubbard model physics in WSe2/WS2moiré superlattices. Nature 579, 353–358 (2020).14. Xu, Y. et al. Correlated insulating states at fractional fillings ofmoirésuperlattices. Nature 587, 214–218 (2020).15. Regan, E. C. et al. Mott and generalized Wigner crystal states inWSe2/WS2 moiré superlattices. Nature 579, 359–363 (2020).16. Wang, L. et al. Correlated electronic phases in twisted bilayertransition metal dichalcogenides. Nat. Mater. 19, 861–866 (2020).17. Chu, Z. et al. Nanoscale conductivity imaging of correlated elec-tronic states in WSe2/WS2 moiré superlattices. Phys. Rev. Lett. 125,186803 (2020).18. Zhang, Z. et al. Flat bands in twisted bilayer transition metaldichalcogenides. Nat. Phys. 16, 1093–1096 (2020).Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 5www.nature.com/naturecommunications19. Pan, Y. et al.Quantum-confinedelectronic states arising fromthemoirépattern of MoS2–WSe2 heterobilayers. Nano Lett 18, 1849–1855 (2018).20. Li, H. et al. Imaging moiré flat bands in three-dimensional recon-structed WSe2/WS2 superlattices. Nat. Mater. 20, 945–950 (2021).21. Wu, F., Lovorn, T., Tutuc, E. & MacDonald, A. H. Hubbard modelphysics in transition metal dichalcogenide moiré bands. Phys. Rev.Lett 121, 026402 (2018).22. Li, W., Lu, X., Dubey, S., Devenica, L. & Srivastava, A. Dipolar inter-actions between localized interlayer excitons in van der Waalsheterostructures. Nat. Mater. 19, 624–629 (2020).23. Geim, A. K. & Grigorieva, I. V. Van der Waals heterostructures.Nature 499, 419–425 (2013).24. Li, T. et al. Continuous Mott transition in semiconductor moirésuperlattices. Nature 597, 350–354 (2021).25. Ghiotto, A. et al. Quantum criticality in twisted transition metaldichalcogenides. Nature 597, 345–349 (2021).26. Gu, J. et al. Dipolar excitonic insulator in a moiré lattice. Nat. Phys.18, 395–400 (2022).27. Zhang, Z. et al. Correlated interlayer exciton insulator in hetero-structures of monolayer WSe2 and moiré WS2/WSe2. Nat. Phys. 18,1214–1220 (2022).28. Brem, S. & Malic, E. Bosonic delocalization of dipolar moiré exci-tons. Nano Lett 23, 4627–4633 (2023).29. Xiong, R. et al. Correlated insulator of excitons in WSe2/WS2 moirésuperlattices. Science 380, 860–864 (2023).30. Gao, B. et al. Excitonic Mott insulator in a Bose-Fermi-Hubbardsystem of moiré WS2/WSe2 heterobilayer. Nat. Commun. 15,2305 (2024).31. Park, H. et al. Dipole ladders with large Hubbard interaction in amoiré exciton lattice. Nat. Phys. 19, 1286–1292 (2023).32. Hsu, W.-T. et al. Second harmonic generation from artificiallystacked transitionmetal dichalcogenide twistedbilayers.ACSNano8, 2951–2958 (2014).33. Zhang, C. et al. Interlayer couplings, moiré patterns, and 2D elec-tronic superlattices in MoS2/WSe2 hetero-bilayers. Sci. Adv 3,e1601459 (2017).34. Karni, O. et al. Structure of the moiré exciton captured by imagingits electron and hole. Nature 603, 247–252 (2022).35. Tran, K. et al. Evidence for moiré excitons in van der Waals het-erostructures. Nature 567, 71–75 (2019).36. Wu, F., Lovorn, T. & MacDonald, A. H. Theory of optical absorptionby interlayer excitons in transition metal dichalcogenide hetero-bilayers. Phys. Rev. B 97, 035306 (2018).37. Brem, S., Linderälv, C., Erhart, P. &Malic, E. Tunable phases ofmoiréexcitons in van der Waals heterostructures. Nano Lett 20,8534–8540 (2020).38. Tan, Q., Rasmita, A., Zhang, Z., Novoselov, K. S. & Gao, W.-B. Sig-nature of cascade transitions between interlayer excitons in amoirésuperlattice. Phys. Rev. Lett 129, 247401 (2022).39. Miao, S. et al. Strong interaction between interlayer excitons andcorrelated electrons inWSe2/WS2moiré superlattice.Nat.Commun12, 3608 (2021).40. Liu, E. et al. Excitonic and valley-polarization signatures of fractionalcorrelated electronic phases in a WSe2/WS2 moiré superlattice.Phys. Rev. Lett. 127, 037402 (2021).41. Tan, Q. et al. Layer-dependent correlated phases in WSe2/MoS2moiré superlattice. Nat. Mater 22, 605–611 (2023).42. Moody, G. et al. Intrinsic homogeneous linewidth and broadeningmechanisms of excitons in monolayer transition metal dichalco-genides. Nat. Commun. 6, 8315 (2015).43. Ovesen, S. et al. Interlayer exciton dynamics in van der Waals het-erostructures. Commun. Phys 2, 23 (2019).44. Erkensten, D., Brem, S. & Malic, E. Exciton-exciton interaction intransition metal dichalcogenide monolayers and van der Waalsheterostructures. Phys. Rev. B 103, 045426 (2021).45. High, A. A. et al. Indirect excitons in elevated traps. Nano Lett. 9,2094–2098 (2009).46. Alloing, M., Lemaître, A. & Dubin, F. Quantum signature blurred bydisorder in indirect exciton gases. Europhys. Lett. 93, 17007 (2011).47. Ivanov, A. L. Quantum diffusion of dipole-oriented indirect excitonsin coupled quantum wells. Europhys. Lett 59, 586 (2002).48. Xu, Y. et al. Creation of moiré bands in a monolayer semiconductorby spatially periodic dielectric screening. Nat. Mater 20,645–649 (2021).49. Liu, Y. et al. Room temperature nanocavity laser with interlayerexcitons in 2D heterostructures. Sci. Adv. 5, eaav4506 (2019).50. Paik, E. Y. et al. Interlayer exciton laser of extended spatial coher-ence in atomically thin heterostructures. Nature 576, 80–84 (2019).51. Zhang, L. et al. VanderWaals heterostructurepolaritonswithmoiré-induced nonlinearity. Nature 591, 61–65 (2021).52. Butov, L. V., Gossard,A.C. &Chemla, D. S.Macroscopically orderedstate in an exciton system. Nature 418, 751–754 (2002).53. Leonard, J. R. et al. Moiré pattern of interference dislocations incondensate of indirect excitons. Nat. Commun 12, 1175 (2021).54. Anton-Solanas,C. et al. Bosonic condensationof exciton–polaritonsin an atomically thin crystal. Nat. Mater. 20, 1233–1239 (2021).55. Kang, K. et al. Evidence of the fractional quantum spin Hall effect inmoiré MoTe2. Nature 628, 522–526 (2024).56. Fano, U. Effects of configuration interaction on intensities andphase shifts. Phys. Rev. 124, 1866–1878 (1961).57. Tan, Q. et al. Layer-engineered interlayer excitons. Sci. Adv 7,eabh0863 (2021).AcknowledgementsThis work is supported by the Singapore ASTAR (M21K2c0116,M24M8b0004), National Research Foundation through its CompetitiveResearchProgram (CRPAwardNo.NRF-CRP22-2019-0004,NRF-CRP30-2023-0003, NRF2023-ITC004-001 and NRF-MSG-2023-0002) and Sin-gapore Ministry of Education (MOE-T2EP50221-0005, MOE-T2EP50222-0018).We thank Leonid Butov for discussion regarding the experimentalresults.Author contributionsW.-b.G. andQ.T. conceived and designed the experiments, Q.T. fabricatedthe devices, and performed the experiments with the help of Z.Z., X.C. andX.D., K.W. and T.T. providedhigh-quality BN, A.R. performed the theoreticalanalysis, A.R. and Q.T. analyzed the data., A.R., Q.T., R.H., H.C. andW.-b.G.wrote the manuscript with inputs from all authors. W-b.G. supervised theproject. All authors contributed to the discussion of the results.Competing interestsThe authors declare no competing interests.Additional informationSupplementary information The online version containssupplementary material available athttps://doi.org/10.1038/s41467-025-57391-8.Correspondence and requests for materials should be addressed toQinghai Tan, Hongbing Cai or Wei-bo Gao.Peer review information Nature Communications thanks the anon-ymous reviewers for their contribution to the peer review of this work. Apeer review file is available.Reprints and permissions information is available athttp://www.nature.com/reprintsPublisher’s note Springer Nature remains neutral with regard to jur-isdictional claims in published maps and institutional affiliations.Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 6https://doi.org/10.1038/s41467-025-57391-8http://www.nature.com/reprintswww.nature.com/naturecommunicationsOpen Access This article is licensed under a Creative CommonsAttribution-NonCommercial-NoDerivatives 4.0 International License,which permits any non-commercial use, sharing, distribution andreproduction in any medium or format, as long as you give appropriatecredit to the original author(s) and the source, provide a link to theCreative Commons licence, and indicate if you modified the licensedmaterial. Youdonot havepermissionunder this licence toshare adaptedmaterial derived from this article or parts of it. The images or other thirdparty material in this article are included in the article’s CreativeCommons licence, unless indicated otherwise in a credit line to thematerial. If material is not included in the article’s Creative Commonslicence and your intended use is not permitted by statutory regulation orexceeds the permitted use, you will need to obtain permission directlyfrom the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.© The Author(s) 2025Article https://doi.org/10.1038/s41467-025-57391-8Nature Communications |         (2025) 16:4518 7http://creativecommons.org/licenses/by-nc-nd/4.0/http://creativecommons.org/licenses/by-nc-nd/4.0/www.nature.com/naturecommunications Enhanced coherence from correlated states in WSe2/MoS2 moiré heterobilayer Results Device structure and conceptual overview Enhanced IX coherence via IX-IX interaction Temperature and power effects on IX linewidth Enhanced IX coherence at correlated electronic state Discussion Methods Sample fabrication CW and time-resolved PL measurements Time-coherence measurement SHG measurement Data availability References Acknowledgements Author contributions Competing interests Additional information