Intense emission from an InGaN quantum disc (QDisc) embedded in a GaN nanowire p-n junction is directly resolved by performing cathodoluminescence spectroscopy. The luminescence observed from the p-type GaN region is exclusively dominated by the emission at 380 nm, which has been usually reported as the emission from Mg induced impurity bands. Here, we confirm that the robust emission from 380 nm is actually not due to the Mg induced impurity bands, but rather due to being the recombination between electrons in the QDisc and holes in the p-type GaN. This identification helps to get a better understanding of the confused luminescence from nanowires with thin QDiscs embedded for fabricating electrically driven single photon emitters.
© 2017 Optical Society of America
Reliable generation of single photons on-demand is essential for numerous applications in quantum cryptography, quantum metrology, and linear-optical quantum computing [1–3]. Low-dimensional semiconductor materials such as quantum dots/discs are prominent candidates for the realization of miniaturized single photon emitting devices, because of their very narrow emission linewidth and capability to integrate with mainstream semiconductor diode technology [4–9]. Since the early demonstration of the first quantum dot single photon light emitting diode by Yuan et al , there have been many reports on electrically driven single photon source (SPS) in narrow bandgap semiconductors (such as III-As and III-P) [11–13]. However, all those reported narrow bandgap materials only operate around cryogenic temperatures. Very recently, room temperature operation of an electrically driven SPS based on III-nitride quantum dots has been realized . III-nitride semiconductors cannot only allow the excitons to remain stable up to room temperature due to their large exciton binding energy , but also have broad emission range from deep ultraviolet to telecommunication wavelengths.
Unfortunately, due to the anisotropy in the strain profile arising from asymmetry in the quantum dot structure, piezoelectric fields, compositional inhomogeneity and alloy disorder, self-organized InGaN quantum dots face several limitations for more efficient single photon emission and higher operating speeds . A different approach, recently reported, is the use of disc-in-a-nanowire heterostructures for electrical excitation [16,17]. The nanowires (NWs) are also benefited from the fact that the radial strain relaxation during expitaxy reduces the polarization field and density of threading dislocations [18–21]. More importantly, the piezoelectric field in such NWs is significantly lower than that in self-organized quantum dots embedded in matrix and the band filling effects are also absent in this structure [16,22,23]. Nevertheless, due to the difficulties in the complicated fabrication and Ohmic contact technology, there are only few electrically driven single photon devices based on III-nitride disc-in-nanowire heterostructures have been realized. Moreover, it is also noticed that almost no investigation on the optical properties of single disc-in-nanowire p-n junction has been reported. To attain necessary control over the growth of well confined quantum disc (QDisc), and further obtain high quality electrically driven SPS, a thorough understanding of the emission characteristics and underlying physical processes is needed.
In this letter, we present micro-structure properties and emission features from GaN NW p-n junction with InGaN disc embedded as the active region by using cathodoluminescence (CL) spectroscopy in a scanning electron microscope (SEM) at various temperatures. By extracting the point spectra along the axial line of NWs, luminescence from the p-type GaN is exclusively dominated by an emission with wavelength of 380 nm, which has been usually reported for general p-type GaN as being the Mg induced impurity bands [24–26]. Coincidentally, the spectra of photoluminescence (PL) show that the emission wavelength from the thin QDisc is also around 380 nm. In order to clarify the origin of each luminescence, temperature dependent evolution of the monochromatic CL and the p-i-n GaN NWs without InGaN QDisc are characterized. The results evidence that the robust emission at 380 nm is not due to the Mg induced impurity bands in p-type GaN, but rather being the recombination between electrons in the QDisc and holes in the p-type GaN.
2. Results and discussion
The InGaN/GaN disc-in-nanowire heterostructure shown schematically for a single NW in Fig. 1(a) was grown on Si (111) substrate by plasma-assisted molecular beam epitaxy (PA-MBE). The substrate was cleaned with a hydrofluoric acid solution to remove native oxides prior to being loaded into the MBE chamber. Then the Si wafer was degassed under ultrahigh vacuum at 900°C for 30 min. Following the annealing, the substrate temperature was lowered until the reconstruction transition from (1 × 1) to (7 × 7) occurs at around 830°C. Subsequently, 1 min nitridation of the Si surface at 680°C leads to formation of a SixNy film. Catalyst-free growth of NWs was carried out under nitrogen-rich condition, maintaining a constant nitrogen flux (480 RF power @ 1.5 sccm N2 flow rate). Si-doped GaN (n-GaN) NWs with a height of ~900 nm were firstly grown at 700 °C, and then followed by a single layer of InGaN QDisc surrounded by 20 nm of undoped GaN barriers grown at 450 °C. Finally, 200-nm-long Mg-doped GaN (p-GaN) wires were grown on top at 500 °C. The relative low substrate temperature during the growth of p-GaN leads to an increase in the lateral growth rate and results in an increase in the NW diameter at the end. It is shown that the GaN NWs for measurements are around 1.1-μm-long on the average (Fig. 1(b)). A representative NW from the ensemble is depicted in a cross-sectional high angle annular dark field transmission electron microscope (HAADF-TEM) image in Fig. 1(c), where an InGaN QDisc is located at about 200 nm away from the top of the NW p-n junction. Higher magnified HAADF image of the NW (position marked with a yellow dashed rectangle) shows that the InGaN QDisc appears as a brighter layer, and GaN is darker. The thickness of the InGaN disc is around 2 nm. To reveal the optical properties of the NWs, PL measurements on as-grown NW ensembles are conducted from 77 K to 300 K in a cold finger cryostat. A 325 nm laser is utilized as the excitation source. Figure 1(d) shows the temperature dependent PL spectra. The NWs exhibit intense 77 K luminescence peaks centered at 356 nm and 381 nm respectively. With increasing temperature, the luminescence peak at 356 nm associated with the emission from the band-edge of GaN shows redshift and becomes weaker. These phenomena are due to the temperature induced bandgap shrinkage and the activation of the non-radiative recombination at higher temperature [27,28]. However, the emission band observed at around 381 nm is kept strong and predominant even up to room temperature. That indicates that that emission does not originate from Mg induced impurity bands but rather from the InGaN QDisc in NW [29,30]. The fact that the intensity of the PL emission typically decreases monotonically with increasing temperature is mainly due to non-radiative processes intrigued at high temperature . Here, the emission from the InGaN QDisc originates from the well confined states, so that they could be isolated from the non-radiative recombination centers; while the Mg induced impurity defects is distributed in the nanowires, not being isolated from the nonradiative recombination centers, and thus their luminescence should show similar temperature dependence as that for the luminescence from the band-edge of GaN. The luminescence spectrum in Fig. 1(d) is obviously not Gaussian-like due to the compositional fluctuation of QDisc in NWs and possible satellites of the luminescence.
To further investigate the optical properties of the NWs, they were removed from the as-grown sample and transferred to a doped silicon wafer, as shown in Fig. 2(a). The use of a doped substrate is required to avoid the charging effect. Measurements were carried out on an Attolight Rosa 4634 CL microscope, with a beam probe of 3 nm in diameter, an accelerating voltage of 8 kV and a beam current of 20 nA. This CL microscope tightly integrates a high numerical aperture achromatic and an aberration-corrected optical objective (NA = 0.72), combined with the electron column objective lens. The CL was spectrally resolved with a Czerny-Turner spectrometer (320 mm focal length, 150 grooves/mm grating) and measured with a UV-Vis sensor. The dispersed QDisc-NWs were cryogenically cooled using liquid helium. At low temperature (25 K), panchromatic CL intensity is mapped simultaneously to the detection of the SEM signal (Fig. 2(b)). The images have a size of 256 × 256 pixels. The strongest integrated intensity appears in the top region of the NWs, where p-GaN is formed, just like tadpoles. From Fig. 2(c), the spatially averaged spectrum from the depicted region (Fig. 2(b)) exhibits three different emission contributions. A luminescence peak at 356 nm with a full width at half maximum (FWHM) of 8 nm is associated with the emission from free exciton (FE) recombination. The FE line has a shoulder, around 365 nm, which is ascribed to the bound exciton transition. The emission band from the InGaN QDisc is observed at around 381 nm. The discrepancy in the relative heights of the peaks between CL spectra in Fig. 2(c) and PL spectra in Fig. 1(d) is mainly due to the large difference in carrier generation rates and penetration depth for these two techniques.
For a detailed analysis of a single NW luminescence, the panchromatic CL mapping point spectra are extracted along the axial line of the NWs squared in Fig. 2(a). As shown in the inset of Fig. 3, 10 points from the top to bottom of the NW are recorded. The near-band-edge emission of GaN at 356 nm is observed almost extended over the whole NW. At the top part of the NW (points 1 and 2), namely the p-type region, it is surprised to observe that the luminescence is exclusively dominated by the emission with wavelength of 380 nm. This emission should be actually from the InGaN QDisc according to the previous PL and CL analysis. However, on the contrary, at point 3, where the InGaN QDisc is located as identified by TEM, there is unexpectedly no emission from 380 nm, while only emission with very weak intensity at 356 nm appears.
To identify and visualize the local microscopic origin of these two recombination channels, colored monochromatic CL pictures are shown in Figs. 4(a) and 4(e). The luminescence at 356 nm (with a 1 nm spectral window) exhibits an areal intensity distribution extended over the whole n-type region, and is well related to the emission of the GaN NW. But no particular emission from the p-type region at this wavelength is observed. In contrast, the emission line at 380 nm (with a 15 nm spectral window) appears exclusively at the top position of the GaN NW, which is the Mg-doped GaN region. Other NWs spreading on the silicon substrate do share the same phenomenon. One can naturally speculate that this peak is most likely due to Mg induced impurity bands in this Mg-doped part of the GaN NW. A 381 nm emission band attributed to the conduction band to Mg-related acceptor level (e-A) transition in doped GaN NWs at 20 K has indeed been reported [24–26]. To further demonstrate the origin of the emission at 380 nm observed in our samples, temperature dependent evolution of the monochromatic CL was followed up. Figure 4 sums up the CL spectra from an individual NW recorded at temperatures ranging from 25 K to 298 K, which clearly indicate that the integrated intensity decreases with increasing temperature, while the emission at 381 nm can however still be evidently observed at 298 K. Also according to the temperature dependent PL in Fig. 1(d) where the 381 nm emission is predominant even at room temperature, it excludes the possibility that it may originate from Mg induced impurity bands. To support this speculation, we characterize the CL features of some p-i-n GaN NWs without InGaN QDisc under the same growth condition. By extracting the point spectrum from the p-type region, the emission bands from the GaN NW as well as from the Mg induced impurity bands are indeed observed at 356 nm and 383 nm respectively, as shown in Fig. 5. However, as temperature increases, the intensity of the emission from the Mg induced impurity bands quickly decreases, and disappears at 150 K, while the band-edge emission remains. Combined with previous analysis, those results further confirm that the emission at 381 nm coming out from the p-i-n GaN NWs with InGaN QDisc is definitely not owing to the impurity bands.
From the perspective of recombination dynamics, the above phenomenon can be understood in the following way: when an electron beam bombs the NW, there will be excess carriers created, both electrons and holes, at each region of the whole nanowire. The excess carriers can either be recombined to emit light immediately, or transport to somewhere with lower potential, and then recombine there with oppositely charged excess carriers, to emit light corresponding to the energy of the corresponding excited states. In our case, as the QDisc holds one of the excited state, whether the excess carriers will recombine at the point of excitation, or at QDisc, depends on the ratio of two rates, i.e, the recombination rate right at the excitation site, and the transport rate from QDisc to that site, or vice versa [31–33]. As shown in the proposed energy band diagram in Fig. 6, it is very hard for the excess electrons in the n-type region to transport to the QDisc due to the sharp barrier in the conduction band, so the excess e lectrons will have an efficient recombination path with excess holes in the same region. Therefore, when the excitation occurs in the n-type region, what emits is mainly the luminescence of GaN itself. Whereas for the p-type region, there are a large number of holes in the valance band, while being excited, the excess holes will be accumulated near the QDisc, but they are difficult to transport over there because of the barrier. However, for the excess electrons, they are very easy to “fall down” to the QDisc. As such, the direct recombination rate is not competitive with the transfer rate of electrons from the QDisc to the p-type region as a spatially indirect recombination. That is similar to the case of a type-II heterostructure where the confined electrons and holes are separated in different locations. That explains why there is nearly no emission from the luminescence of GaN when the excitation is located in the p-type region. But the electrons in the QDisc will have an efficient recombination with the holes in p-type region, leading to the luminescence coming out from the p-type region but with the emission wavelength of the QDisc.
In summary, the micro-structure properties and emissions from an InGaN QDisc embedded in GaN NW p-n junction are investigated through panchromatic CL mappings at low temperature. Based on a detail study on point spectra along the axial line of the NW, we reveal that the luminescence from the p-type region is exclusively dominated by the wavelength around 380 nm, while in the InGaN QDisc region, there surprisingly is almost no emission from the InGaN QDisc can be observed. Due to the growth of the well confined QDisc, its emission wavelength happens to be the same as that from the Mg impurity in the p-type region. Temperature dependent evolution of the monochromatic CL and other p-i-n GaN NWs without InGaN QDisc are characterized to verify that the robust emission around 380 nm appeared in the p-GaN region of the nanowire is not due to Mg induced impurity bands, but rather being the recombination between electrons in the QDisc and holes in the p-type region. We believe that this identification would help to deepen understanding of the confused luminescence from NWs with thin QDisc embedded, which are aimed for fabricating electrically driven single photon emitters.
Science Challenge Project (No. TZ2016003-2); the National Key Research and Development Program of China (No. 2016YFB0400100); NSAF (No. U1630109); NSFC (No.61376060 and 61521004); the CAEP Microsystem and THz Science and Technology Foundation (No. CAEPMT201507); and the Open Fund of the State Key Laboratory on Integrated Optoelectronics.
References and links
1. C. Santori, D. Fattal, J. Vucković, G. S. Solomon, and Y. Yamamoto, “Indistinguishable photons from a single-photon device,” Nature 419(6907), 594–597 (2002). [PubMed]
2. V. Giovannetti, S. Lloyd, and L. Maccone, “Quantum-enhanced measurements: beating the standard quantum limit,” Science 306(5700), 1330–1336 (2004). [PubMed]
3. E. Knill, R. Laflamme, and G. J. Milburn, “A scheme for efficient quantum computation with linear optics,” Nature 409(6816), 46–52 (2001). [PubMed]
4. A. Bennett, D. Unitt, A. Shields, P. Atkinson, and D. Ritchie, “Influence of exciton dynamics on the interference of two photons from a microcavity single-photon source,” Opt. Express 13(20), 7772–7778 (2005). [PubMed]
5. A. Lochmann, E. Stock, J. A. To, W. Unrau, A. Toropov, A. Bakarov, V. Haisler, and D. Bimberg, “Electrically pumped microcavity based single photon source driven at 1 GHz,” Electron. Lett. 45, 566 (2009).
6. A. J. Shields, E. Limited, and C. S. Park, “Semiconductor quantum light sources,” Nat. Photonics 1, 215–223 (2007).
7. R. Arians, A. Gust, T. Kümmell, C. Kruse, S. Zaitsev, G. Bacher, and D. Hommel, “Electrically driven single quantum dot emitter operating at room temperature,” Appl. Phys. Lett. 93, 173506 (2008).
8. A. Jamil, J. Skiba-Szymanska, S. Kalliakos, A. Schwagmann, M. B. Ward, Y. Brody, D. J. P. Ellis, I. Farrer, J. P. Griffiths, G. A. C. Jones, D. A. Ritchie, and A. J. Shields, “On-chip generation and guiding of quantum light from a site-controlled quantum dot,” Appl. Phys. Lett. 104, 101108 (2014).
9. C. Bentham, I. E. Itskevich, R. J. Coles, B. Royall, E. Clarke, J. O. Hara, N. Prtljaga, A. M. Fox, M. S. Skolnick, and L. R. Wilson, “On-chip electrically controlled routing of photons from a single quantum dot,” Appl. Phys. Lett. 106, 221101 (2015).
10. Z. Yuan, B. E. Kardynal, R. M. Stevenson, A. J. Shields, C. J. Lobo, K. Cooper, N. S. Beattie, D. A. Ritchie, and M. Pepper, “Electrically driven single-photon source,” Science 295(5552), 102–105 (2002). [PubMed]
11. T. Heindel, C. Schneider, M. Lermer, S. H. Kwon, T. Braun, S. Reitzenstein, S. Höfling, M. Kamp, and A. Forchel, “Electrically driven quantum dot-micropillar single photon source with 34% overall efficiency,” Appl. Phys. Lett. 96, 011107 (2010).
12. S. Strauf, N. G. Stoltz, M. T. Rakher, L. a. Coldren, P. M. Petroff, and D. Bouwmeester, “High-frequency single-photon source with polarization control,” Nat. Photonics 1, 704–708 (2007).
13. M. Reischle, G. J. Beirne, W.-M. Schulz, M. Eichfelder, R. Rossbach, M. Jetter, and P. Michler, “Electrically pumped single-photon emission in the visible spectral range up to 80 K,” Opt. Express 16(17), 12771–12776 (2008). [PubMed]
14. S. Deshpande, T. Frost, A. Hazari, and P. Bhattacharya, “Electrically pumped single-photon emission at room temperature from a single InGaN/GaN quantum dot,” Appl. Phys. Lett. 105, 141109 (2014).
15. B. Monemar, “Fundamental energy gap of GaN from photoluminescence excitation spectra,” Phys. Rev. B 10, 676 (1974).
16. S. Deshpande, J. Heo, A. Das, and P. Bhattacharya, “Electrically driven polarized single-photon emission from an InGaN quantum dot in a GaN nanowire,” Nat. Commun. 4, 1675 (2013). [PubMed]
17. L. Zhang, C.-H. Teng, P.C. Ku, and H. Deng, “Electrically driven single-photon emission from site-controlled InGaN/GaN quantum dots” CLEO 2014 Postdeadline Pap. Dig. 1, JTh5B.6 (2014).
18. W. Guo, M. Zhang, A. Banerjee, and P. Bhattacharya, “Catalyst-free InGaN/GaN nanowire light emitting diodes grown on (001) silicon by molecular beam epitaxy,” Nano Lett. 10(9), 3355–3359 (2010). [PubMed]
19. P. Wang, X. Wang, T. Wang, C. S. Tan, B. Sheng, X. Sun, M. Li, X. Rong, X. Zheng, Z. Chen, X. Yang, F. Xu, Z. Qin, J. Zhang, X. Zhang, and B. Shen, “Lattice-Symmetry-Driven Epitaxy of Hierarchical GaN Nanotripods,” Adv. Funct. Mater. 27, 1604854 (2017).
20. M. Shahmohammadi, J. D. Ganière, H. Zhang, R. Ciechonski, G. Vescovi, O. Kryliouk, M. Tchernycheva, and G. Jacopin, “Excitonic Diffusion in InGaN/GaN Core-Shell Nanowires,” Nano Lett. 16(1), 243–249 (2016). [PubMed]
21. L. Yan, S. Jahangir, S. A. Wight, B. Nikoobakht, P. Bhattacharya, and J. M. Millunchick, “Structural and Optical Properties of Disc-in-Wire InGaN/GaN LEDs,” Nano Lett. 15(3), 1535–1539 (2015). [PubMed]
22. S. Chichibu, T. Azuhata, T. Sota, and S. Nakamura, “Recombination of Localized Excitons in InGaN Single- and Multiquantum-Well Structures,” Appl. Phys. Lett. 69, 4188 (1996).
23. Y. D. Jho, J. S. Yahng, E. Oh, and D. S. Kim, “Field-dependent carrier decay dynamics in strained InxGa1-x/GaN quantum wells,” Phys. Rev. B 66, 35334 (2002).
24. M. G. Kibria, F. A. Chowdhury, S. Zhao, M. L. Trudeau, H. Guo, and Z. Mi, “Defect-engineered GaN : Mg nanowire arrays for overall water splitting under violet light,” Appl. Phys. Lett. 106, 113105 (2015).
25. Y. S. Park, J. H. Na, R. A. Taylor, C. M. Park, K. H. Lee, and T. W. Kang, “The recombination mechanism of Mg-doped GaN nanorods grown by plasma-assisted molecular-beam epitaxy,” Nanotechnology 17, 913 (2006).
26. F. Furtmayr, M. Vielemeyer, M. Stutzmann, A. Laufer, B. K. Meyer, and M. Eickhoff, “Optical properties of Si- and Mg-doped gallium nitride nanowires grown by plasma-assisted molecular beam epitaxy,” J. Appl. Phys. 104, 074309 (2008).
27. M. A. Reshchikov and H. Morkoç, “Luminescence properties of defects in GaN,” J. Appl. Phys. 97, 061301 (2005).
28. C. Hsiao, H. Hsu, L. Chen, C. Wu, C. Chen, M. Chen, L. Tu, and K. Chen, “Photoluminescence spectroscopy of nearly defect-free InN microcrystals exhibiting nondegenerate semiconductor behaviors,” Appl. Phys. Lett. 91, 181912 (2007).
29. M. A. Reshchikov, G.-C. Yi, and B. W. Wessels, “Behavior of 2.8- and 3.2-eV photoluminescence bands in Mg-doped GaN at different temperatures and excitation densities,” Phys. Rev. B 59, 176–183 (1999).
30. U. Kaufmann, M. Kunzer, H. Obloh, M. Maier, C. Manz, A. Ramakrishnan, and B. Santic, “Origin of defect-related photoluminescence bands in doped and nominally undoped GaN,” Phys. Rev. B 59, 5561–5567 (1999).
31. J. Wallentin, P. Wickert, M. Ek, A. Gustafsson, L. R. Wallenberg, M. H. Magnusson, L. Samuelson, K. Deppert, and M. T. Borgstro, “Degenerate p-doping of InP nanowires for large area tunnel diodes Degenerate p-doping of InP nanowires for large area tunnel diodes,” Appl. Phys. Lett. 99, 253105 (2012).
32. N. M. Haegel, T. J. Mills, M. Talmadge, C. Scandrett, C. L. Frenzen, H. Yoon, C. M. Fetzer, and R. R. King, “Direct imaging of anisotropic minority-carrier diffusion in ordered GaInP Direct imaging of anisotropic minority-carrier diffusion in ordered GaInP,” J. Appl. Phys. 105, 023711 (2009).
33. B. J. M. Brenny, D. R. Abujetas, D. Van Dam, J. A. Sa, J. Go, and A. Polman, “Directional Emission from Leaky and Guided Modes in GaAs Nanowires Measured by Cathodoluminescence,” ACS Photonics 3, 677–684 (2016).