Site-selectively generated photon emitters in monolayer MoS2 via local helium ion irradiation

Quantum light sources in solid-state systems are of major interest as a basic ingredient for integrated quantum photonic technologies. The ability to tailor quantum emitters via site-selective defect engineering is essential for realizing scalable architectures. However, a major difficulty is that defects need to be controllably positioned within the material. Here, we overcome this challenge by controllably irradiating monolayer MoS2 using a sub-nm focused helium ion beam to deterministically create defects. Subsequent encapsulation of the ion exposed MoS2 flake with high-quality hBN reveals spectrally narrow emission lines that produce photons in the visible spectral range. Based on ab-initio calculations we interpret these emission lines as stemming from the recombination of highly localized electron–hole complexes at defect states generated by the local helium ion exposure. Our approach to deterministically write optically active defect states in a single transition metal dichalcogenide layer provides a platform for realizing exotic many-body systems, including coupled single-photon sources and interacting exciton lattices that may allow the exploration of Hubbard physics.

SiO 2 /Si. The spectrum reveals emission from the neutral exciton 0 X A 1s , and charged exciton − X A 1s , and from the L-peak and furthermore dose dependent emission from the L D peak. (Data are adapted from Ref. [2]) b, Dose dependent photoluminescence of MoS 2 on hBN reveals spectrally more narrow free exciton emission and similar to a emission from the L-peak and dose dependent defect emission from the L D -peak c, Dose dependent photoluminescence of a hBN/MoS 2 /hBN van der Waals heterostructure. Besides spectrally narrow free exciton emission the spectrum also reveals spectrally sharp emission.
Since the photo-physical properties of single-layer MoS 2 strongly depend on the dielectric environment, especially on the encapsulation and passivation of the surface of the crystal, [3 6] we spectroscopically investigate the He-ion dose dependence of single-layer MoS 2 in dier-ent dielectric environments. Supplementary Figure 2 shows typical low-temperature (10 K) µ-PL spectra of single-layer MoS 2 on SiO 2 /Si, MoS 2 /hBN and fully hBN encapsulated MoS 2 .
The He-ion dose is varied between σ ∼ 10 12 ions cm −2 and σ ∼ 10 15 ions cm −2 . All spectra reveal emission from the neutral and charged exciton with much narrower linewidths on hBN substrates due to a reduced inhomogeneous linewidth.
[47] Moreover, all spectra exhibit typically observed broad low-energy emission from the L-peak in addition to a superimposed emission from the L D peak [2] that increases with increased σ. This emission is detuned by ∆E ∼ 190 meV from the neutral exciton and spectrally narrows for fully hBN encapsulated To spatially correlate the spectrally sharp emission, we perform spatially resolved µ-PL mappings. Supplementary Figure 3a shows a typical photoluminescence spectrum at 10 K.
The spectrum reveals various single defect emitters. The corresponding spatially integrated PL mapping for the red coloured emitter is shown in Supplementary Figure 3b. The arrow highlights the spatial position from which the spectrum is taken from. The emission is spatially localized with a spatial extent that is limited by the focal spot diameter (∼ 1.2 µm) of our confocal microscope. Supplementary Figure 4a shows an averaged PL spectrum with the neutral exciton set to zero detuning ∆E. Here, we average over all spectra that reveal emission from the neutral exciton and emission from localized emitters. In each spectrum the emission energy of the neutral exciton is used as a reference in order to properly sum up all spectra. The averaged spectrum shows emission from the neutral and charged exciton, a signicant contribution from the L-peak at lower energies and also a signicant contribution from the most prominent defect peak X D at ∆E ∼ 190 meV. The spectrum has similarities with a photoluminescence spectrum of He-ion bombarded MoS 2 on SiO 2 as reported recently [2] and also shown in Instead of creating defects with a xed helium ion dose, we continuously vary the ion dose over two orders of magnitude from σ = 2.2 · 10 12 cm −2 to σ = 1.4 · 10 14 cm −2 by adjusting the dwell time accordingly. Here, large elds of 4 × 8 µm are exposed for each dose in order to create reliable statistics. The creation eciency of single defect emitters has a maximum at ∼ 2 · 10 12 cm −2 corresponding to ∼ 3.5 emitters µm −2 . From the temperature dependent data shown in Fig. 4a in the main manuscript, we also investigate the temperature dependent intensity of a single defect emitter X L . The corresponding data is presented in Supplementary Figure 10 in an Arrhenius plot. The data is tted by with the emission intensity at 0 K, a tting constant C and the activation energy E A .
From the t, we obtain an activation energy of (17.18 ± 2.05) meV. The activation energy is lower compared to the thermal activation energy k B T ∼ 25 meV at room temperature.
This explains why emission is absent at elevated temperatures.
Supplementary Figure 11. Dipolar emission characteristics of a single defect emitter. a, Polar plot of the emission polarization characteristics of a single defect emitter for excitation with a xed linear polarization (blue). Data are tted with cos 2 θ. b, Corresponding spectra for co-(black) and cross-linearly (red) detection polarization as highlighted by the arrows in a.

Supplementary Note 11: Emission polarization of a single defect emitter
To further support the interpretation that localized emission originates from a single dipole, we perform polarization resolved PL spectroscopy on a typical emitter at ∆E ∼ 170 meV. We excite with a xed linear excitation polarization and probe the emission polarization by rotating an analyzer in the detection path. The corresponding emission characteristic is shown in Supplementary Figure 11a. Two spectra for emission co-and crosslinearly aligned with the emission polarization are shown in Supplementary Figure 11b. This single photoluminescence line X L exhibits a dipolar emission pattern, while the 0 X A 1s emits isotropically. In turn, the former observation provides direct evidence that the luminescence X L indeed originates from a single dipole. [9] Supplementary Note 12: Defect emitter lineshape To describe the defect emitter spectrum we utilize the independent boson model that has been succcessfully applied to describe the lineshape of quantum dot states [1012] and defect-bound excitons [13]. The Hamiltonian for localized excitons of energy E coupled to lattice vibrations reads where the interaction part represents lattice deformations due to the presence of an exciton.
The exciton-phonon matrix elements contain the exciton wave function φ(k) with q e/h been the electron/hole momentum. For a localized s-exciton we obtain where a B is the 2D exciton bohr radius. The carrier-phonon matrix elements g c/v are treated in deformation potential approximation [14,15] and we account for the coupling with LA and TA phonon modes, which are found to be the dominant source of phonon dephasing in the system. For deformation potential coupling with acoustic phonons of dispersion ω q = v s q the carrier-phonon matrix element is given by with sound velocity v s , 2D mass density ρ, and deformation potentials D c/v taken from DFT/DFPT calculations [15,16]. Here, A is the crystal area.
The spectral properties are obtained from the optical response to a classical electric eld E(t), which drives a polarization of the medium. In response to a weak optical (test) eld, the linear optical susceptibility is given by: considering the polarization P in direction of the external eld. To obtain the frequencydependent response of the medium, we use the Fourier transform of the polarization P (t), which can be written for a delta-like excitation as [10] The polarization decays due to radiative recombination with a rate γ rad as well as excitations into phonon sidebands described by the complex function where N q = 1/(exp[hω q /k B T ] − 1) is the phonon occupation, g X q the exciton-phonon matrix elements, γ ph the phonon lifetime and ω q the acoustic phonon dispersion assumed in equilibrium at lattice temperature T . Absorption spectra that are obtained from solving Eqs. (6)-(8) posses a Lorentz-broadend zero-phonon line as well as phonon sidebands due to multi-phonon processes. The latter give rise to the observed asymmetric lineshape at low temperature, where only phonon emission processes dominate.
Emission spectra are obtained as a mirror image of the absorption spectra reected across the zero-phonon line [17]. At low temperature best agreement between calculated and measured photoluminescence spectra is obtained for a Bohr radius of 2 nm, see Supplementary of 0.75, which is consistent with the phonon coupling that we have obtained from the analysis of the temperature dependent peak position (cf. main text). For the phonon lifetime 1/γ ph we used a value of 40 ps according to Ref. [18]. From the t we obtain a value for the radiative linewidth of 0.5 meV. Using this parameters, all spectra in Fig. 3 of the main text are then well described by changing the temperature according to the experimental condition.
Supplementary references