Figure 2: Creation of quantum emitter arrays in 1L-WSe₂.

(a) Integrated PL intensity raster scan of the region enclosed by the green rectangle in Fig. 1d, taken under 200 nW μm⁻², 532 nm laser excitation at 10 K. Green crosses mark the position of the six nanopillars beneath the 1L-WSe₂. Colour-scale bar maximum, 160 kcounts s⁻¹. (b) PL spectra taken at each of the corresponding green crosses in a, from left to right respectively, showing the presence of narrow lines at each nanopillar location. (c) Photon correlation measurements corresponding to the filtered spectral regions (10 nm wide) enclosed by the blue, green and pink rectangles, in b, with g⁽²⁾(0)=0.087±0.065, 0.17±0.02 and 0.18±0.03, and rise times of 8.81±0.80 ns, 6.15±0.36 ns and 3.08±0.41 ns, respectively. (d) Spectrum taken from a 1L-WSe₂ on a 190 nm nanopillar, showing lower background and a single sub-nm emission peak. Higher-resolution spectrum in the inset reveals the fine-structure splitting of this QE peak. An asymmetry can be seen in the spectrum, which has been previously attributed to a phonon sideband in naturally occurring QEs³¹. (e) Probability distribution (in %) of the number of emission lines per nanopillar for samples using different nanopillar heights (60, 130 and 190 nm in white, blue and purple, respectively). A trend of higher probability of single QE emission peaks per nanopillar location with increasing height is evident, reaching 50% for 190 nm nanopillars. (f) Increasing nanopillar height also leads to a reduction of spectral wandering. Solid black circles represent the mean value of spectral wandering of several QEs for a given nanopillar height, while the error bars represent the standard deviation of each distribution, both extracted from time-resolved high-resolution spectral measurements (Supplementary Fig. 4). A total number of seven samples was used to collect the statistics necessary for Fig. 2e,f.