Room temperature light emission from Ge self-assembled quantum dots (QDs) embedded in L3-type photonic crystal (PhC) nanocavity is successfully demonstrated under current injection through a lateral PIN diode structure. The Ge QDs are grown on silicon-on-insulator (SOI) wafer by solid-source molecular beam epitaxy (SS-MBE), and the PIN diode is fabricated by selective ion implantation around the PhC cavity. Under an injected current larger than 0.5 mA, strong resonant electroluminescence (EL) around 1.3–1.5 μm wavelength corresponding to the PhC cavity modes is observed. A sharp peak with a quality factor up to 260 is obtained in the EL spectrum. These results show a possible way to realize practical silicon-based light emitting devices.
© 2012 Optical Society of America
Silicon-based optical interconnection is now considered as one of the most attractive solutions to overcome the problems of bandwidth and power consumption in electronic integrated circuits [1,2]. So far, almost all of the photonic components have been realized with high performances on silicon-compatible platform, except the light emitting devices. This is due to the inherent property of indirect band gap of silicon. Numerous approaches have been proposed to solve this problem, including silicon nanocrystals , SiGe nanostructures [4, 5], erbium doping in silicon and SiN , silicon Raman lasers , and so on. Among these, Ge self-assembled quantum dots (QDs) [8, 9] is a promising solution due to its advantages of compatibility to CMOS technology and light emission in the telecommunication band.
However, the light emission efficiency of Ge QDs is still far away from the practical application level. By combining the Ge QDs with optical microcavities, the light emission can be dramatically enhanced due to the Purcell effect . Room temperature resonant photoluminescence (PL) has been successfully demonstrated with photonic crystal (PhC) cavities and microdisk resonators [11–13]. However, most of these works are based on impractical optical pumping. Only very recently, our group demonstrated current-injected light emission from microdisks . In this work, room temperature resonant electroluminescence (EL) corresponding to the whispering gallery mode was observed under current injection by a vertical PIN diode.
Compared with microdisks, PhC nanocavites are more promising due to their potential for higher Q-factor and smaller mode volume. On the other hand, it is difficult to inject and confine carriers into such small cavities. One of the most common ways is to use a vertical PIN junction with heavily doped area on top of the cavity . However, in this structure, most part of the current passes through the PhC area, instead of the cavity. Moreover, a large additional loss is introduced by the heavily doped region and metal contact just on top of the cavity. Although we have already realized room temperature EL from PhC cavity with this vertical PIN structure, the resonant peaks are not obvious and the Q-factor is very low . On the other hand, a lateral PIN junction provides an excellent solution for the injection and confinement of the carriers, while making very few changes to the cavity structure. Several groups have already used the lateral PIN diode structure in III–V material-based PhC cavities and realized ultralow threshold lasers [17, 18].
In this paper, we report the fabrication and characterization of lateral PIN junction PhC nanocavity light emitting diodes with Ge self-assembled QDs, in which the junction is fabricated by selective ion implantation into the PhC region. Strong resonant EL is successfully observed at room temperature.
2. Device structure and fabrication
Figure 1 shows the schematic diagram of our current-injected light emitting diode, together with the cross-section of the PhC cavity area. A lateral PIN diode is integrated on the PhC slab in order to inject carriers into the cavity.
We started the device fabrication from silicon-on-insulator (SOI) wafer with 160 nm-thick Si top layer and 2 μm-thick BOX layer. The Si layer was first thinned down to ∼50 nm through thermal oxidation and HF wet etching. The active layers were then grown on this wafer by solid-source molecular beam epitaxy (SS-MBE) at 700 °C: First 40 nm-thick Si buffer, then three layers of Ge self-assembled QDs separated by 20nm-thick Si spacers, and at last 40 nm Si cap layer.
The PIN junction was firstly fabricated by selective BF2 and As ions implantation. In order to realize uniform doping concentration along the depth direction and high surface concentration simultaneously, ion implantation with multiple energies and doses was performed. For BF2, we used implantation energies of 25KeV, 50KeV, and 75KeV, and doses of 2×1014 cm−2, 5×1014 cm−2, and 3×1014 cm−2, respectively. For As, implantation energies of 25KeV, 50KeV, and 100KeV, and doses of 2×1014 cm−2, 5×1014 cm−2, and 5×1014 cm−2 were used respectively. After that, a rapid thermal anneal of 10 seconds at 1000 °C was performed to activate the dopant. The PhC cavity was then fabricated by electron beam lithography and inductively-coupled reactive ion dry etching. At last, silver metal was evaporated and lifted-off to form the electrodes. Figure 2 shows the scanning electron microscope (SEM) image of the fabricated device, together with the zoomed area of the PhC cavity. The cavity is a common L3-type, with designed lattice constant of a = 420 nm and hole radius of r = 0.24a. Based on the SRIM simulation  of ion distribution and Hall effect measurement , we estimate the maximum p-type doping concentration to be about 0.8×1020 cm−3 and the maximum n-type doping concentration to be about 1.0×1020 cm−3. The width of the intrinsic region of the PIN junction is designed to be 900 nm in the doping layout, and shrunk to about 860 nm due to the lateral diffusion after activation anneal.
The fabricated devices were characterized by micro-photoluminescence at room temperature. The excitation was done either by a diode-pumped solid-state (DPSS) laser of 532 nm wavelength (for PL) or a DC current source (for EL). The light emission signal was collected by an objective lens (100×, NA = 0.50), and then dispersed by a monochromator with a 320 mm focus length and detected by a liquid-nitrogen-cooled InGaAs detector array.
3. Experimental results and discussion
The current-voltage (I–V) property of the fabricated device was first characterized by a semiconductor parameter analyzer. As shown in Fig. 3, a typical diode characteristic is obtained, verifying the electrical performance of our devices. The reverse leakage current is about 18 μA for a negative bias of −5 V. A rather high current density in the cavity region, about 19 kA/cm2 at an applied voltage of 1.5 V, is obtained from the I–V curve. A series resistance of about 975 Ω is extracted through fitting the I–V curve to the ideal diode equation.
Figure 4(a) shows the EL spectra of the device under different injected currents. As the current becomes larger than 0.5 mA, clear resonant peaks are seen to appear in the spectrum. These peaks can be well identified as the PhC cavity modes. In order to verify this, numerical simulation based on three-dimensional finite-difference time-domain (3D-FDTD)  was performed to calculate the cavity modes with geometry parameters measured from the SEM image. The wavelengths of the cavity modes are shown on the top panel of Fig. 4(a). The peak positions are seen to agree well with the simulation result. Through Lorentz fitting as shown in Fig. 4(b), we can obtain that the strongest peak in the EL spectrum (at 0.5 mA), corresponding to the fourth-order cavity mode, has a Q-factor of about 260. It is similar with that (∼265) of a reference cavity without electrical structure measured by PL, indicating that the heavily doped regions have few effects on the optical performances of the cavity. On the other hand, this Q-factor is lower than that (∼400) of our previous PL results from a similar cavity with free-standing structure. It is mainly due to the weak optical confinement of SiO2-cladding side and in-plane TE-TM coupling loss due to the vertical asymmetry . Moreover, compared with the simulation Q-factor (∼530) of the corresponding structure, it is reduced by a factor of 2, which might be attributed to the scattering loss induced by fabrication roughness and the absorption induced by the injected free-carriers. As the current increases, a broad peak around 1.22 μm is seen to evolve in the spectra. This peak is caused by enhanced emission from the PhC area around the cavity , since part of the current passes through the surrounding PhC area.
As one may notice, the resonant peak corresponding to the fundamental cavity mode (with longest wavelength) is not seen in the EL spectra. In order to understand the reason for this, we performed PL measurement for the device under optical pumping with two types of objective lens, one is with an NA of 0.50 and the other is 0.95. The PL spectra, together with the EL spectrum under 2 mA injected current, are shown in Fig. 5. The fundamental cavity mode around 1.54 μm is observed in the PL spectrum with NA = 0.95, but becomes very weak in the PL spectrum with NA = 0.50. The intensity reduction of this peak can be therefore attributed to the low collection efficiency of the collection optics. Compared with that of the fourth-order mode, the far-field pattern of the fundamental mode is much more dispersive . Moreover, due to the vertical asymmetry, the optical confinement of SiO2-cladding side is weaker than that of air-cladding side, and the upward radiation from the cavity becomes weaker than the downward radiation. In total, only very small portion of the light emission of the fundamental mode is collected by the optics with NA = 0.50 (with collection angle of 30°), so as the case in EL measurement. This problem can be solved by using free-standing structure and optimizing the far-field pattern by applying a double-period perturbation to the cavity .
As the injected current increased, the luminescence intensity and peak wavelengths were seen to vary. Figure 6(a) shows the dependence of the resonant wavelength, Q-factor of the fourth-order cavity mode on the injected current. It is reasonable that the peak wavelength shows red shift as the current increases. This is due to the thermo-optic effect induced by the injected current and the refractive index of silicon is increased by heating. A linear relationship between the changes of the peak wavelength Δλ and the refractive index Δn is obtained as Δλ = 0.29Δn through numerical simulation. The peak wavelength and the corresponding refractive index change against electrical power are then plotted in Fig. 6(b). By considering that the thermo-optic coefficient of silicon is 1.94×10−4 around 1.3 μm wavelength at room-temperature , the temperature increase is about 12 K/mW against electrical power. The nonlinear increase of the refractive index at high electrical power can be attributed to the increased thermo-optic coefficient of silicon at high temperature . On the other hand, the Q-factor reduction with the current increase is mainly due to the free-carrier absorption. At high injection level, a large amount of carriers cannot be recombined radiatively in the Ge QDs, which leads to absorption of the emitted light.
Figure 7 shows the current-dependent emission intensity of the fourth-order cavity mode. The mode intensity is obtained by integrating the resonant peak after subtracting the broad background. The mode intensity increases nonlinearly against injected current, which is similar to our previous PL and EL results [11,16]. The dependence is usually empirically characterized by a nonlinear function L ∝ Im, where L is the mode intensity and I is the current. Interestingly, it is seen that at low injection level (< 1.6 mA), the extracted nonlinear exponent m is 1.39 and increases to 1.93 at high injection level (> 1.6 mA). This increased nonlinearity is very similar to the case of Ge-based EL device , in which the increase of the population of the direct Γ valley due to the current heating is to be attributed. But here the reason is thought to be rather different due to the thermal quenching of light emission from Ge QDs at raised temperature . The possible explanation for this increased nonlinearity is the increased three-dimensional carrier localization inside the cavity [11, 28] at high injected current. Due to the existance of the BOX layer and surrounding air holes, the diffusion of the carriers is partially blocked outward from the cavity. The carrier concentrations in the cavity might increase nonlinearly with the current. At high current density, strong band bending will also occur around Ge/Si heterostructure. The overlap between electron and hole wave functions wii be, therefore, enhanced and then the forbidden rules will be relaxed further. Moreover, as the resonant peak shifts toward longer wavelength, the emitter-cavity detuning decreases and more Ge QDs can be coupled to the cavity mode, which also contributes to the enhancement of the emission. The increased emission efficiency at high current is thought to be a result of the combinantion effects of these factors. However, more detailed studies are necessary to clarify which factors dominates the dependence.
Silicon-based current-injected light emitting diodes were realized with Ge self-assembled quantum dots embedded inside L3-type photonic crystal nanocavities. By using a lateral PIN diode structure, the carriers could be injected into the cavity efficiently. Strong resonant electroluminescence corresponding to the PhC cavity modes were successfully observed at room temperature when the injected current was larger than 0.5 mA. A record Q-factor of 260 was obtained in this device. We believe that a much higher Q-factor and emission intensity can be achieved by optimizing the PhC cavity design and using free-standing structure. The resonant wavelength, Q-factor and mode intensity were found to be dependent on the injected current. These results provide us a reliable approach towards practical electrically-driven silicon-based light sources for optical interconnection application.
This work was partly supported by project for strategic advancement of research infrastructure for private universities, 2009–2013, and by Grant-in-Aid for Scientific Research (A) (Grant No. 21246003) from MEXT, Japan. The authors would like to thank Y. Hoshi, K. Sawano, and S. Taguchi for the discussion and help in the experiments.
References and links
1. G. T. ReedSilicon Photonics: The State of the Art (J. Wiley & Sons, 2008). [CrossRef]
2. L. Tsybeskov, D. J. Lockwood, and M. Ichikawa“Silicon Photonics: CMOS Going Optical,” Proc. IEEE 97, 1161–1165 (2009). [CrossRef]
3. L. Pavesi, L. Dal Negro, C. Mazzoleni, G. Franzo, and F. Priolo, “Optical gain in silicon nanocrystals,” Nature (London) 408, 440–444 (2000). [CrossRef]
4. D. K. Nayak, N. Usami, S. Fukatsu, and Y. Shiraki, “Band-edge photoluminescence of SiGe/strained-Si/SiGe type-II quantum wells on Si (100),” Appl. Phys. Lett. 63, 3509–3511 (1993). [CrossRef]
5. R. Apetz, L. Vescan, A. Hartmann, C. Dieker, and H. Luth, “Photoluminescence and electroluminescence of SiGe dots fabricated by island growth,” Appl. Phys. Lett. 66, 445–447 (1995). [CrossRef]
6. H.-S. Han, S.-Y. Seo, and J. H. Shin, “Optical gain at 1.54 μm in erbium-doped silicon nanocluster sensitized waveguide,” Appl. Phys. Lett. 79, 4568–4570 (2001). [CrossRef]
7. H. Rong, A. Liu, R. Jones, O. Cohen, D. Hak, R. Nicolaescu, A. Fang, and M. Paniccia, “An all-silicon Raman laser,” Nature (London) 433, 292–294 (2005). [CrossRef]
8. S. Fukatsu, H. Sunamura, Y. Shiraki, and S. Komiyama, “Phononless radiative recombination of indirect excitons in a Si/Ge type-II quantum dot,” Appl. Phys. Lett. 71, 258–260 (1997). [CrossRef]
9. T. Brunhes, P. Boucaud, S. Sauvage, F. Aniel, J.-M. Lourtioz, C. Hemandez, Y. Campidelli, O. Kermarrec, D. Bensahel, G. Faini, and I. Sagnes, “Electroluminescence of Ge/Si self-assembled quantum dots grown by chemical vapor deposition,” Appl. Phys. Lett. 77, 1822–1824 (2000). [CrossRef]
10. E. M. Purcell, “Spontaneous emission probabilities at radio frequencies,” Phys. Rev. 69, 681 (1946).
11. J. S. Xia, Y. Ikegami, Y. Shiraki, N. Usami, and Y. Nakata, “Strong resonant luminescence from Ge quantum dots in photonic crystal microcavity at room temperature,” Appl. Phys. Lett. 89, 201102 (2006). [CrossRef]
12. J. S. Xia, K. Nemoto, Y. Ikegami, Y. Shiraki, and N. Usami, “Silicon-based light emitters fabricated by embedding Ge self-assembled quantum dots in microdisks,” Appl. Phys. Lett. 91, 011104 (2007). [CrossRef]
13. M. E. Kurdi, X. Checoury, S. David, T. P. Ngo, N. Zerounian, O. Kermarrec, Y. Campidelli, and D. Bensahel, “Quality factor of Si-based photonic crystal L3 nanocavities probed with an internal source,” Opt. Express 16, 207–210 (2008).
15. H. G. Park, S. H. Kim, S. H. Kwon, Y. G. Ju, J. K. Yang, J. H. Baek, S. B. Kim, and Y. H. Lee, “Electrically driven single-cell photonic crystal laser,” Science 305, 1444–1447 (2004). [CrossRef] [PubMed]
16. T. Tsuboi, X. Xu, J. Xia, N. Usami, T. Maruizumi, and Y. Shiraki, “Room temperature electroluminescence from Ge quantum dots embedded in photonic crystal microcavities,” Appl. Phys. Express 5, 052101 (2012). [CrossRef]
17. B. Ellis, M. A. Mayer, G. Shambat, T. Sarmiento, J. Harris, E. E. Haller, and J. Vuckovic, “Ultralow-threshold electrically pumped quantum-dot photonic-crystal nanocavity laser,” Nat. Photon. 5, 297–300 (2011). [CrossRef]
18. S. Matsuo, K. Takeda, T. Sato, M. Notomi, A. Shinya, K. Nozaki, H. Taniyama, K. Hasebe, and T. Kakitsuka, “Room-temperature continuous-wave operation of lateral current injection wavelength-scale embedded active-region photonic-crystal laser,” Opt. Express 20, 3773–3780 (2012). [CrossRef] [PubMed]
19. J. ZieglerSRIM The Stopping and Range of Ions in Matter, Version 2008.03, http://www.srim.org.
20. A. Mokhberi, P. B. Griffin, J. D. Plummer, E. Paton, S. McCoy, and K. Elliot, “A comparative study of dopant activation in Boron, BF2, Arsenic, and Phosphorus implanted silicon,” IEEE Trans. Electron Dev. 49, 1183–1191 (2002). [CrossRef]
21. RSoft FullWAVE, RSoft Design Group, Inc., http://www.rsoftdesign.com.
22. Y. Tanaka, T. Asano, R. Hatsuta, and S. Noda, “Investigation of point-defect cavity formed in two-dimensional photonic crystal slab with one-sided dielectric cladding,” Appl. Phys. Lett. 88, 011112 (2006). [CrossRef]
23. S. Iwamoto, Y. Arakawa, and A. Gomyo, “Observation of enhanced photoluminescence from silicon photonic crystal nanocavity at room temperature,” Appl. Phys. Lett. 91, 211104 (2007). [CrossRef]
24. S. Nakayama, S. Ishida, S. Iwamoto, and Y. Arakawa, “Effect of cavity mode volume on the photoluminescence from silicon photonic crystal nanocavities,” Appl. Phys. Lett. 98, 171102 (2011). [CrossRef]
25. N. Tran, S. Combrie, P. Colman, A. D. Rossi, and T. Mei, “Vertical high emission in photonic crystal nanocavities by band-folding design,” Phys. Rev. B 82, 075120 (2010). [CrossRef]
26. B. J. Frey, D. B. Leviton, and T. J. Madison, “Temperature-dependent refractive index of silicon and germanium,” Proc. SPIE 6273, 62732J (2006). [CrossRef]
27. S. Cheng, J. Lu, G. Shambat, H. Yu, K. Saraswat, J. Vuckovic, and Y. Nishi, “Room temperature 1.6 μm electroluminescence from Ge light emitting diode on Si substrate,” Opt. Express 17, 10019–10024 (2009). [CrossRef] [PubMed]
28. M. El Kurdi, S. David, P. Boucaud, C. Kammerer, X. Li, V. Le Thanh, S. Sauvage, and J.-M. Lourtioz, “Strong 1.3-1.5 μm luminescence from Ge/Si self-assembled islands in highly confining microcavities on silicon on insulator,” J. Appl. Phys. 96, 997–1000 (2004). [CrossRef]