Atomic monolayers represent a novel class of materials for studying localized and free excitons in two dimensions and for engineering optoelectronic devices based on their significant optical response. Here, we investigate the role of the substrate in the photoluminescence response of and monolayers exfoliated either on or epitaxially grown InGaP substrates. In the case of , we observe a significant qualitative modification of the emission spectrum, which is widely dominated by the trion resonance on InGaP substrates. However, the effects of inhomogeneous broadening of the emission features are strongly reduced. Even more striking, in sheets of , we could routinely observe emission lines from localized excitons with linewidths down to the resolution limit of 70 μeV. This is in stark contrast to reference samples featuring monolayers on surfaces, where the emission spectra from localized defects are widely dominated by spectral diffusion and blinking behavior. Our experiment outlines the enormous potential of III–V monolayer hybrid architectures to obtain high quality emission signals from atomic monolayers, which are simple to integrate into nanophotonic and integrated optoelectronic devices.
© 2017 Optical Society of America
Monolayers of transition metal dichalcogenides have moved into the focus of solid-state spectroscopy, since these new materials feature a variety of unique optical properties. Monolayers composed of the transition metal Mo or W and the chalcogen Se, S, or Te crystallize in a honeycomb lattice that lacks an inversion center. This yields a characteristic bandstructure where the direct bandgap transitions are located at the K and points of the hexagonal Brillouin zone. Due to the lack of inversion symmetry, these points are inequivalent and are occupied by charges of opposite spin. This leads to the coupling of the spin and the corresponding valley, introducing a new degree of freedom that is accessible via optical selection rules. As a result, each valley can be distinctly addressed by the polarization of an injection laser, which leads to novel spinor effects in these systems [1–6]. In addition, so-called valley excitons are formed, which feature an extraordinarily high binding energy exceeding 300 meV . This is a consequence of reduced dimensions, reduced dielectric screening, and flat bands leading to a heavy exciton mass. In most of these materials, even up to ambient conditions, the absorption and luminescence spectrum is dominated by excitonic effects, rather than by direct interband transitions. While the general properties, such as the exciton frequency and the trion binding energy, are primarily determined by the monolayer itself, the surrounding environment still has considerable influence on the optical properties. For instance, it has been shown that excitons in monolayers of sensibly react to absorbed molecules on the surface , and energy shifts resulting from capping have been reported . Similarly, the choice of the substrate can have a significant effect on the luminescence properties of the free monolayer excitons, as well as the emission features from localized excitons, which were recently identified as novel sources of single-photon streams [10–14]. Furthermore, these kinds of quantum emitters have been observed in monolayers lying on top of patterned arrays of nanopillars [15,16]. Here, we study the excitonic properties of exfoliated monolayers of and at cryogenic temperatures, which have been transferred onto as well as InGaP/GaAs heterostructures. In the case of , we observe a strong reduction of the inhomogeneous broadening of the dominant trion feature as epitaxial substrates are utilized. In monolayers of , we focus on the emission of localized excitons. These quantum-dot-like features are strongly broadened and disturbed by their environment on the insulating glass substrates. In stark contrast, the semiconducting InGaP/GaAs substrates have a suitable band alignment with respect to and (compared to GaAs) and less heavy surface oxidation (compared to AlGaAs), facilitating dramatically reduced charge fluctuations and yielding stable and robust emitters of single photons on an epitaxial platform. Furthermore, InGaP is a well-established material platform for integrated photonic devices, such that our work can easily be extended to utilize monolayer materials in more complex, integrated schemes.
2. SAMPLE STRUCTURE AND SETUP
The investigated monolayers were produced by mechanical exfoliation from a or a bulk crystal with scotch tape. After their monolayer nature was confirmed via their distinct photoluminescence (PL) and the color contrast in an optical microscope, they were transferred onto the designated target substrate via the dry-stamp method . Using this technique, flake sizes of around were fabricated. Two different sample types were implemented, which are shown in Fig. 1(a). The monolayers were transferred onto substrates composed of a 90 nm layer on top of a Si substrate. The other substrate was made of a 250 nm thick layer that was grown lattice-matched on a semi-insulating GaAs by means of gas-source molecular beam epitaxy.
In order to get an impression of the samples’ surface quality, we performed atomic force microscope (AFM) measurements, which can be seen in Fig. 1(b). The root-mean-squared roughness of both samples is of the same magnitude. Specifically, the surface is characterized by a roughness of , while the InGaP surface features a comparable value of 0.29 nm. Optical characterization was carried out in a standard microphotoluminescence setup. The samples were attached to the cold finger of a liquid helium flow cryostat, and the luminescence from the flake was collected by a objective () in a confocal microscope system. The structures were excited by a continuous-wave (cw) 532 nm laser. Photoluminescence measurements were performed using a Princeton Instruments SP2750i spectrometer equipped with a liquid nitrogen cooled CCD and a grating () for the high-resolution images or a grating for overview spectra. The PL could also be collected in a fiber-coupled Hanbury Brown and Twiss (HBT) setup with a timing resolution of approximately 570 ps to measure the second-order field correlation of the emission, after passing through a pair of bandpass filters (1 nm bandwidth).
3. EXPERIMENTAL RESULTS AND DISCUSSION
First, we investigate the impact of the two aforementioned substrates on the emission characteristics of monolayers. Figure 2 depicts a series of PL spectra of the structure, which were recorded sequentially under nominally the same conditions and without blanking the laser. The spectra were taken over a time span of 10 min at a constant laser power of 50 μW. We observe the common spectral signatures of monolayers. At 1.657 eV, the free exciton () is clearly visible. On the low-energy side, the negatively charged trion () emerges at 1.625 meV, yielding a trion binding energy of 32 meV. Notably, during the series, the initial exciton intensity decreases and the trion intensity increases until both signals converge to a constant intensity ratio after roughly 5 min. This behavior can be explained by a photo-induced doping effect that introduces new free carriers into the system, enhancing the formation of trions [18,19] at the expense of free neutral excitons. At all accessible pump powers, we observe emissions from both the and the resonance in this case. A significantly more in-depth analysis of the interplay between excitons and trions in on insulating substrates can be found in Ref. . Conversely, on the heterostructures, the free exciton is not visible at 50 μW laser power, and only the trion-attributed resonance can be clearly observed, with an energy of 1.632 eV. At high pump powers we see a strongly suppressed signal from the exciton at around 1.665 eV, which is about two orders of magnitude weaker in intensity than the trion. The spectral energy shift of compared to the stack occurs reproducibly in different flakes, and is most likely a consequence of the modified dielectric environment. Remarkably, the overall intensity of this trion resonance does not change with time, indicating that the system inherently has access to a great amount of free carriers. We note that both monolayers originate from the same bulk crystal, and therefore we can rule out inherent doping of the flake itself as a reason for this behavior. In Fig. 3 we depict the results of a power series from both samples. The nonresonant excitation power was ramped up from 50 μW up to 6 mW, and we plot the trions’ integrated intensity and linewidth. By taking into account realistic parameters (absorption coefficient , lifetime ), we estimate an upper bound of the exciton density on the order of in this experiment. With increasing power, the intensity of the observed resonances rises approximately linearly, as shown in Fig. 3(a). Fitting the data (red lines) to a straight line gives a slope of 0.96 for and 0.92 for InGaP, in good agreement with the expected slope of 1 for charged excitons. At higher output powers (), the emissions start to show a saturation behavior independent of the substrate used, caused by exciton annihilation . Another important parameter is the corresponding full width at half-maximum (FWHM) of the signal studied, which is plotted in Fig. 3(b). On the glass substrate the linewidth of the trion reaches a value of around 13 meV for a low laser power. Increasing the power yields a progressive broadening of the emission line, reaching approximately 16 meV at 6 mW pump power. We assume that this power-induced broadening of the trion resonance is a consequence of local heating from the pump laser, but it could be also induced by additional charges that accumulate in the monolayer and at random positions at the heterointerface . This charge puddling effect is known to occur on surfaces , which induce a randomly varying inhomogeneity in the PL response. Conversely, the linewidth on the InGaP sample is as small as 6.5 meV, surpassing its counterpart by a factor of 2. Even more remarkable, the linewidth stays nearly constant with increasing power and reaches just 7 meV at 6 mW laser output. This is due to the higher thermal conductivity of InGaP compared to , leading to lower local heating at the laser spot. This is also supported by an overall reduced spectral shift of InGaP during the power series. Overall, these results already outline the reduction of charge-induced fluctuations in monolayer InGaP devices, and illustrate the impact the right substrate can have on the excitonic properties of . While monolayers of are specifically suitable for studying the effects of free excitons and trions, the observation of single-photon emissions from localized excitons have brought monolayers of into the center of solid-state quantum photonics. Figure 4 shows a typical PL spectrum from such a localized exciton in a monolayer on top of a substrate, which was excited by a cw 532 nm laser at an excitation power of 30 μW and a nominal sample temperature of 4.2 K. The PL spectrum consists of several sharp peaks with linewidths of 2 meV, centered at 1.52 eV. Such a spectral feature, which is redshifted 180 meV from the free valley exciton (1.7 eV), is comparable to previously reported localized emission signals in monolayers . Compared to the weak, broad PL spectrum from the localized exciton in the monolayer exfoliated on the substrate, several bright, spectral-resolution-limited (70 μeV) PL peaks were observed from sheets transferred onto the InGaP/GaAs substrate (temperature of 4.5 K). Here, the PL excitation power in Fig. 4(b) is around 70 nW, which is almost 3 orders of magnitude smaller than the nominal 30 μW in Fig. 4(a). Additionally, the inset in Fig. 4(b) shows the cw-pumped autocorrelation histogram for the marked peak in Fig. 4(b). The emission is spectrally filtered by a pair of bandpass filters and then coupled into a fiber-based HBT setup to measure the second-order autocorrelation. Clear antibunching is observed around that drops well below 0.5 and therefore proves single-photon emission. In order to account for the finite time resolution of our setup, we fit the measured data with a two-sided exponential decay convolved with a Gaussian distribution according to
Following this, we extract a deconvoluted value of .
To assess the influence of the spectral wandering on the macroscopic time scale on the emission features depicted in Figs. 4(a) and 4(b), we record various spectra every second and combine them in the contour graph in Figs. 5(a) and 5(b). In Fig. 5(a), clear spectral wandering and jumps on a time scale of seconds are observed. Each frame is then fitted with a Lorentzian function, and the statistics of the peak energies are plotted in Fig. 5(c). We find a direct contribution as large as from the long-term spectral diffusion. This characteristic slow spectral jitter of such large magnitude is commonly observed for self-assembled quantum emitters close to surfaces or interfaces that yield the capability of trapping and releasing charges. Thus, and in agreement in principle with the studies presented in Fig. 3 for the case, we conclude that the spectral jumps are induced by carriers trapped via dangling bonds on the surface. Compared to the monolayer on substrate, no obvious spectral wandering is observed in Fig. 5(b), where the monolayer is transferred onto the InGaP substrate. The corresponding statistics of the spectral wandering in Fig. 5(d) yield a value around 5.5 μeV, which is within the linewidth fitting uncertainty. The narrowing could be attributed to fewer charge fluctuations in a semiconducting environment, which allows the transfer of trapped charges, leading to a suppression of the long scale spectral jitter. Additionally, it has been shown that InP surfaces can be effectively passivated by S or Se, saturating many dangling bonds of the substrate [27,28]. In this scenario, one would expect to observe a more stable photon emission from the InGaP hybrid structure.
Last, we perform a statistical study of the influence of the different substrates on the spectral linewidth of the localized excitons in the monolayers. A statistical histogram for 37 randomly localized emitters from 10 different monolayers on substrate is presented in Fig. 6(a). The extracted linewidths randomly fluctuate between 147 μeV and 3.3 meV. Similarly, the statistical histogram for 251 randomly localized emitters from 10 different monolayers on InGaP/GaAs is depicted in Fig. 6(b). Although linewidths of 100 μeV sharp peaks could be measured, they are not necessarily representative. Here the median linewidth of on InGaP (120 μeV) is close to the spectrometer resolution limit (70 μeV), which is about 10 times smaller compared to the structure (1150 μeV). Therefore, the resolution-limited, jitter-free PL strongly indicates that the InGaP substrate could greatly enhance the emission properties of the localized excitons in the monolayer.
In conclusion, we have studied the influence of the substrate on the emission properties of monolayers of and at cryogenic temperatures. On our reference substrate, the luminescence of the free exciton and trion in is notably inhomogeneously broadened, and is sensitive to power broadening. The investigated localized defects occurring in monolayers are subject to a long-term spectral diffusion induced by a slowly varying charge environment. In stark contrast, InGaP substrates show a notable effect on the charge environment, which directly leads to a reduced broadening of the trionic emission in and in many cases eliminates the slow spectral diffusion acting on localized emission centers in . Together with the highly developed photonic processing technology of InGaP/GaAs structures, this makes –InGaP heterostacks very interesting for novel nanophotonic and integrated monolayer-based quantum photonic architectures. Furthermore, we have observed a significantly enhanced formation of free trions in monolayers on InGaP, which makes such a platform highly suitable for studying interactions of monolayer excitations with electron gases, and likely represents a new, simpler approach to trion polaritons without the necessity of electrostatic gating.
State of Bavaria; H2020 European Research Council (ERC) (Project Unlimit-2D).
1. X. Xu, W. Yao, D. Xiao, and T. F. Heinz, “Spin and pseudospins in layered transition metal dichalcogenides,” Nat. Phys. 10, 343–350 (2014). [CrossRef]
2. A. M. Jones, H. Yu, N. J. Ghimire, S. Wu, G. Aivazian, J. S. Ross, B. Zhao, J. Yan, D. G. Mandrus, D. Xiao, W. Yao, and X. Xu, “Optical generation of excitonic valley coherence in monolayer WSe2,” Nat. Nanotechnol. 8, 634–638 (2013). [CrossRef]
3. D. Xiao, G. B. Liu, W. Feng, X. Xu, and W. Yao, “Coupled spin and valley physics in monolayers of MoS2 and other Group-VI dichalcogenides,” Phys. Rev. Lett. 108, 196802 (2012). [CrossRef]
4. K. F. Mak, K. He, J. Shan, and T. F. Heinz, “Control of valley polarization in monolayer MoS2 by optical helicity,” Nat. Nanotechnol. 7, 494–498 (2012). [CrossRef]
5. T. Cao, G. Wang, W. Han, H. Ye, C. Zhu, J. Shi, Q. Niu, P. Tan, E. Wang, B. Liu, and J. Feng, “Valley-selective circular dichroism of monolayer molybdenum disulphide,” Nat. Commun. 3, 887 (2012). [CrossRef]
6. H. Zeng, J. Dai, W. Yao, D. Xiao, and X. Cui, “Valley polarization in MoS2 monolayers by optical pumping,” Nat. Nanotechnol. 7, 490–493 (2012). [CrossRef]
7. A. Ramasubramaniam, “Large excitonic effects in monolayers of molybdenum and tungsten dichalcogenides,” Phys. Rev. B 86, 115409 (2012). [CrossRef]
8. J. He, K. Wu, R. Sa, Q. Li, and Y. Wei, “Magnetic properties of nonmetal atoms absorbed MoS2 monolayers,” Appl. Phys. Lett. 96, 082504 (2010). [CrossRef]
9. D. Sercombe, S. Schwarz, O. Del Pozo-Zamudio, F. Liu, B. J. Robinson, E. Chekhovich, and I. Tartakovskii, “Optical investigation of the natural electron doping in thin MoS2 films deposited on dielectric substrates,” Sci. Rep. 3, 3489 (2013). [CrossRef]
10. Y.-M. He, G. Clark, J. R. Schaibley, Y. He, M.-C. Chen, Y.-J. Wei, X. Ding, Q. Zhang, W. Yao, X. Xu, C.-Y. Lu, and J.-W. Pan, “Single quantum emitters in monolayer semiconductors,” Nat. Nanotechnol. 10, 497–502 (2015). [CrossRef]
11. M. Koperski, K. Nogajewski, A. Arora, V. Cherkez, P. Mallet, J.-Y. Veuillen, J. Marcus, P. Kossacki, and M. Potemski, “Single photon emitters in exfoliated WSe2 structures,” Nat. Nanotechnol. 10, 503–506 (2015). [CrossRef]
12. C. Chakraborty, L. Kinnischtzke, K. M. Goodfellow, R. Beams, and A. N. Vamivakas, “Voltage-controlled quantum light from an atomically thin semiconductor,” Nat. Nanotechnol. 10, 507–511 (2015). [CrossRef]
13. A. Srivastava, M. Sidler, A. V. Allain, D. S. Lembke, A. Kis, and A. Imamoğlu, “Optically active quantum dots in monolayer WSe2,” Nat. Nanotechnol. 10, 491–496 (2015). [CrossRef]
14. P. Tonndorf, R. Schmidt, R. Schneider, J. Kern, M. Buscema, G. A. Steele, A. Castellanos-Gomez, H. S. J. van der Zant, S. M. de Vasconcellos, and R. Bratschitsch, “Single-photon emission from localized excitons in an atomically thin semiconductor,” Optica 2, 347–352 (2015). [CrossRef]
15. A. Branny, S. Kumar, R. Proux, and B. D. Gerardot, “Deterministic strain-induced arrays of quantum emitters in a two-dimensional semiconductor,” arXiv:1610.01406 (2016).
16. C. Palacios-Berraquero, D. M. Kara, A. R.-P. Montblanch, M. Barbone, P. Latawiec, D. Yoon, A. K. Ott, M. Loncar, A. C. Ferrari, and M. Atature, “Large-scale quantum-emitter arrays in atomically thin semiconductors,” arXiv:1609.04427 (2016).
17. A. Castellanos-Gomez, M. Buscema, R. Molenaar, V. Singh, L. Janssen, J. S. J. van der Zant, and G. Steele, “Deterministic transfer of two-dimensional materials by all-dry viscoelastic stamping,” 2D Mater. 1, 11002 (2014).
18. G. V. Astakhov, D. R. Yakovlev, V. P. Kochereshko, W. Ossau, W. Faschinger, J. Puls, and A. Waag, “Binding energy of charged excitons in ZnSe-based quantum wells,” Phys. Rev. B 65, 165335 (2002). [CrossRef]
19. F. Cadiz, C. Robert, G. Wang, W. Kong, X. Fan, M. Blei, D. Lagarde, M. Gay, M. Manca, T. Taniguchi, K. Watanabe, T. Amand, X. Marie, P. Renucci, S. Tongay, and B. Urbaszek, “Ultralow power threshold for laser induced changes in optical properties of 2D molybdenum dichalcogenides,” 2D Mater. 3, 045008 (2016).
20. N. Lundt, E. Cherotchenko, O. Iff, X. Fan, Y. Shen, P. Bigenwald, A. Kavokin, S. Höfling, and C. Schneider, “The interplay between excitons and trions in a monolayer of MoSe2,” arXiv:1702.04231 (2017).
21. Y. V. Morozov and M. Kuno, “Optical constants and dynamic conductivities of single layer MoS2, MoSe2, and WSe2,” Appl. Phys. Lett. 107, 083103 (2015). [CrossRef]
22. C. Robert, D. Lagarde, F. Cadiz, G. Wang, B. Lassagne, T. Amand, A. Balocchi, P. Renucci, S. Tongay, B. Urbaszek, and X. Marie, “Exciton radiative lifetime in transition metal dichalcogenide monolayers,” Phys. Rev. B 93, 205423 (2016). [CrossRef]
23. N. Kumar, Q. Cui, F. Ceballos, D. He, Y. Wang, and H. Zhao, “Exciton-exciton annihilation in MoSe2 monolayers,” Phys. Rev. B 89, 125427 (2014). [CrossRef]
24. Z. M. Ao, W. T. Zheng, and Q. Jiang, “The effects of electronic field on the atomic structure of the graphene/α-SiO2 interface,” Nanotechnology 19, 275710 (2008). [CrossRef]
25. Y. K. Koh and D. G. Cahill, “Frequency dependence of the thermal conductivity of semiconductor alloys,” Phys. Rev. B 76, 075207 (2007). [CrossRef]
26. Y.-M. He, S. Höfling, and C. Schneider, “Phonon induced line broadening and population of the dark exciton in a deeply trapped localized emitter in monolayer WSe2,” Opt. Express 24, 8066–8073 (2016). [CrossRef]
27. S. Tian, Z. Wei, Y. Li, H. Zhao, X. Fang, J. Tang, D. Fang, L. Sun, G. Liu, B. Yao, and X. Ma, “Surface state and optical property of sulfur passivated InP,” Mater. Sci. Semicond. Process. 17, 33–37 (2014). [CrossRef]
28. C. E. J. Mitchell, I. G. Hill, A. B. McLean, and Z. H. Lu, “Sulfur passivated InP(100): surface gaps and electron counting,” Appl. Surf. Sci. 104–105, 434–440 (1996). [CrossRef]