Three-dimensional (3D) integration of photonic devices is an ultimate route toward highly integrated photonic circuits. 3D photonic crystals (PCs) are a promising platform for 3D integration because of their complete photonic bandgaps (cPBGs) which enables full control of light. However, simultaneous integration of active and passive devices into a 3D PC has been hindered due to fabrication difficulties. Here, we simultaneously integrate a nanocavity laser and waveguides into a 3D PC with a cPBG at near-IR wavelengths using micromanipulation technology. The proposed plate-insertion stacking method allowed fabrication of 3D PCs with a large number of layers, enabling integration of the active and passive circuit components. Laser emission from the photo-excited nanocavity laser was observed from the output port of the waveguides, demonstrating successful guiding of the light from the nanocavity laser in the 3D PC. This work paves the way for 3D photonic circuits using 3D PCs with cPBGs.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Three-dimensional (3D) integration of photonic devices is an ultimate route to highly integrated photonic circuits. Recent technological advancements have stimulated the realization of 3D integrated photonic circuits. For example, laser inscription technology enabled the development of various 3D integrated waveguide circuits [1–3] by scanning a laser spot inside polymers or transparent materials. Interlayer couplings for stacking planar photonic circuits have also been studied extensively as another route toward the 3D integration of photonic circuits [4–8]. However, the waveguides used in these circuits confine light via total internal reflection, and thus difficulties arise, especially in the integration of sharp bends, due to significant radiation loss . In multilayered circuits, evanescent or grating couplers are used for interlayer couplings; however, these couplers usually need long waveguides (more than a few tens of micrometers) [4–8] or additional mirrors  for efficient couplings. These limitations could prevent miniaturization of these 3D circuits.
Photonic crystals (PCs) [10,11], which are optical media with periodic refractive indices in space, can manipulate photons by utilizing photonic bandgap (PBG) effects. As such, PCs have become an important platform for the miniaturization of integrated photonic circuits [12–14]. Waveguides with sharp bends or cavities with small mode volumes (of a few cubic wavelengths) have been demonstrated by carefully designing defects in a two-dimensional (2D) PC where the refractive index is periodic in a plane [14–16]. Since the confinement of light in small volumes can enhance light–matter interactions such as spontaneous emission , small active devices such as nanolasers [18–20] can also be realized in PCs. Although numerous efforts have been devoted to studies on 2D PCs, the lack of PBG along the out-of-plane direction can pose difficulties in stacking and interlayer coupling of 2D PC circuits. Thus, 3D PCs with 3D periodic refractive indices and complete PBGs (cPBGs) in which light propagation can be controlled for any wave vectors and polarizations could be promising platforms not only for emission control [21,22] but also for 3D photonic circuit applications.
Many efforts have been devoted to the development of photonic devices using 3D PCs. Passive components such as waveguides or cavities [23–26] have been realized in semiconductor-based 3D PCs at telecommunication wavelengths. Active devices such as nanocavity lasers  and electrically driven light-emitting diodes  have also been demonstrated experimentally at near-IR wavelengths by embedding active media in 3D PCs, such as III–V compound semiconductor quantum wells and dots. However, the simultaneous integration of active and passive devices in a single 3D PC has not been demonstrated experimentally.
In this Letter, we experimentally demonstrate a 3D PC integrating both waveguides and a nanocavity laser. Light from the nanocavity laser was guided through orthogonally connected straight waveguides, and then emitted outside the 3D PC. We designed the integrated photonic circuit components in the so-called woodpile structure [25–27], which has a cPBG in near-IR wavelengths. The 3D PC circuit was fabricated with a layer-stacking method using micromanipulation [27,29,30], and our stacking method, called the plate-insertion stacking method, was developed to realize the 3D PC circuit. The plate-insertion stacking method enabled stacking of up to 61 layers, which is more than with previously reported methods, facilitating simultaneous integration of several photonic components in a single 3D PC. In low-temperature micro-photoluminescence (μ-PL) measurements, we observed laser emission from the output port of the waveguides, demonstrating successful guiding of laser light from the nanocavity. Such 3D PC circuits composed of passive waveguides and light sources could be fundamental building blocks for future 3D integrated photonic circuits.
Figure 1(a) shows the schematic of the studied 3D PC circuit, and its cutaway image Fig. 1(b) shows the integrated circuit components. The GaAs 3D PC had the woodpile structure, which was composed of a periodic stack of four different line-and-space patterns. The patterns for one period along the stacking direction (Layers I to IV) are shown in gray on the right-hand side of Fig. 1(c). In each pattern, rods of width and thickness were periodically arranged with period . The rods lay along the direction in layers I and III and along the direction in layers II and IV. The rod positions in layers III and IV were shifted by a half-period with respect to the rod positions in layers I and II, respectively. The 3D PC was constructed by stacking a total of 61 plates in a trench, as shown in Fig. 1(a). Detailed structures of the trench and plate are shown in Supplement 1. Inside the 3D PC, a point-defect nanocavity with a cuboid shape and an in-plane size of [yellow shading in Fig. 1(b)] was embedded at the center of the 13th layer [the number of layers was counted from the back to the front in Fig. 1(a)]. For optical gain, the cavity layer contained InAs quantum dots (QDs) at a high areal density (). A waveguide along the direction [blue shading labeled “-waveguide” in Fig. 1(b)] was introduced between the 16th and 52nd layers. This waveguide was connected to another waveguide [red shading labeled “-waveguide” in Fig. 1(b)] extending toward the positive direction in between the 48th and 50th layers. The light from the nanocavity was expected to be output from the top of the structure, indicated by a white arrow in Fig. 1(b), after sequential guiding through the - and -waveguides.
The details of the - and -waveguides are shown in the upper-left images in Figs. 1(c) and 1(d), respectively, together with the plane view in the lower-left images. The waveguides were formed by introducing defects into the layers of the woodpile structure, as shown by the blue and red shading for the - and -waveguides, respectively. The defects are defined in the right images in Figs. 1(c) and 1(d), which show an plane view of the layers embedding the defects for the - and -waveguides, respectively. The defects for the -waveguide were short rods of width , which were embedded only in layers III and IV, as shown in Fig. 1(c), and were bonded with neighboring rods at the center of the pattern. For the -waveguide, the defects were single rods  of width , which were perpendicularly crossed with the rods in layers II and IV, as shown in Fig. 1(d). The waveguides were connected at the center of the 49th layer and were extended, as shown by the arrows in Fig. 1(e). The lengths of these extensions were for the -waveguide and for the -waveguide. Such extensions are useful for increasing the transmission of waveguide bends in 3D PCs .
We numerically investigated the optical characteristics for these circuit components. Dispersion curves calculated for the waveguides using the plane wave expansion method are shown in Fig. 2(a) for the -waveguide and in Fig. 2(b) for the -waveguide. In both Figs. 2(a) and 2(b), the frequency and wave vector are normalized by the unit cell length along the propagation direction; that is, for the -waveguide and for the -waveguide. Among several waveguide modes supported within the cPBG for the -waveguide, we focused our attention on a blue curve, labeled Mode Z in Fig. 2(a), which is a non-degenerate waveguide mode. For the -waveguide, we focused on the red curve, labeled Mode Y in Fig. 2(b), which is a broadband single waveguide mode. The possibility of guiding light from the nanocavity laser though the waveguides was investigated via a 3D finite-difference time-domain (FDTD) simulation. Among the multiple modes supported in the nanocavity, the mode at the frequency of [green line in Fig. 2 labeled Mode C] was coupled with Mode Z (see Supplement 1 for spatial field distributions), and then with Mode Y due to the waveguide extensions shown in Fig. 1(e). This allowed light to be guided through the waveguides from the nanocavity. To evaluate the guiding efficiency, we defined the transmissivity from the nanocavity to the output port as the ratio of the energy flux reaching the output port () to the total energy flux emitted from the nanocavity (), leading to a high transmissivity of for this guiding. In addition to efficient guiding, a high quality factor () was also required for lasing the nanocavity mode. The 3D FDTD calculations showed a value for Mode C of , which was sufficiently high for lasing the nanocavity mode, as reported in Ref. . These results indicated that the 3D PC circuit could successfully guide the laser light from the nanocavity to the output port.
We fabricated the 3D PC circuit using a micromanipulation method [27,29,30]. In this technique, planar structures fabricated on square plates were stacked one by one. A previous stacking method  using standing posts for aligning the plates improved the quality factors of 3D PC nanocavities by accurately stacking the plates, which enabled lasing of 3D PC nanocavity modes . However, due to difficulties in the fabrication of long posts with straight edges, the number of stackable layers remained small , which significantly limited the volume of the 3D PC and thereby the scale of the integrated circuits. Thus, we developed a stacking method called the plate-insertion stacking method, which was suitable for stacking a larger number of layers compared to the post-based stacking method  and maintained high stacking accuracy.
Figure 3(a) shows the scheme for stacking the plates in the plate-insertion method. Plates with the patterns of the woodpile structure were inserted into a trench, and then stacked along the edge of the trench. Parts of the GaAs slab in the trench were suspended in air, as shown in the cross-sectional image in Fig. 3(a). The plate design was matched with the space under the GaAs slab of the trench, which allowed accurate stacking (see Supplement 1 for details). The planes of the stacked plates were perpendicular to the substrate surface. The manipulation was performed using a gold-covered pointed glass controlled by a three-axis piezoelectric stage under scanning electron microscope (SEM) observation. Straightness along the edge of the trench was critical for reducing stacking errors such as displacements and rotations between plates. In comparison with the posts in the post-based stacking method , the trench can be uniformly elongated by simply changing the electron-beam lithography pattern. Thus, the plate-insertion stacking method can be used to stack a large number of layers while maintaining high stacking accuracy.
We succeeded in fabricating the woodpile structure with 61 layers, as shown in the SEM image in Fig. 3(b), by using the plate-insertion stacking method. The magnified pattern shown in Fig. 3(c) clarifies that four periodic patterns in the woodpile structure were formed. The number of stacked layers was 1.8 times greater than that of the 3D PC fabricated by the previous stacking method , which highlights the advantage of this stacking method in increasing the number of layers. Displacements between the stacked plates were estimated to be within 150 nm by SEM observation. This value was greater than that for the previous stacking method  due to the use of larger plates, for which rotation can be observed as a larger displacement at the edge of the plates.
To optically characterize our 3D PC circuit, we conducted low-temperature μ-PL measurements (see Supplement 1 for details). Figure 4(a) shows a PL spectrum measured at 10 K for the averaged input power . Within the broad background emission (gray region) originating in spontaneous emissions from the ensemble QDs in the cavity layer (mainly in the outer frame of the plate), several sharp peaks were clearly observed. We investigated the input power dependence for the peak at 1101 nm wavelength. The integrated intensity and linewidth for this peak are shown in Fig. 4(b) by black and red dots, respectively, as a function of the time-averaged input power. As the input power increased, the mode linewidth narrowed, and the integrated modal intensity was nonlinearly increased, which suggests lasing of a cavity mode. The broadening of the linewidths at a high input power is likely due to the carrier plasma effect. Using a rate-equation model without nonradiative recombination processes, we fitted the integrated intensity as shown by the green curve in Fig. 4(b). The -factor, which describes the coupling efficiency of spontaneous emissions from the QDs to the nanocavity mode, was estimated as for this cavity mode. For this moderately high , no clear kink was observed in the input-output curve, which has been reported in a 2D PC nanolaser with ensemble QDs .
We measured the spatial distribution of the PL intensity for this lasing mode. Figures 4(c) and 4(d) respectively show a spatial PL image measured with and without the bandpass filter (BPF), which is used for extracting the lasing mode. The 10 nm bandwidth of the BPF was centered at 1101 nm. The horizontal and vertical axes in the figures correspond to the and axes defined in Fig. 1(b), respectively. The origin point is positioned immediately above the point-defect cavity. The PL signals were measured through a polarizer set to polarization. We observed a high intensity around the cavity layer without the BPF, as shown in Fig. 4(c). This is because of the broad background emission from the QDs in the outer frame of the cavity layer. While the background emission is largely suppressed by the BPF in Fig. 4(d), a bright spot can be clearly observed around the position (0, 5.3 μm). This position is close to the position of the output port (0, 5.4 μm) within the spatial resolution of 380 nm (see Supplement 1). For comparison, we calculated the component of energy flux at the plane 0.42 μm above the top surface of the 3D PC for a guided cavity mode using the 3D FDTD method, as shown in Fig. 4(e). The optical spot was observed at the position of the output port, which was consistent with the experimental result. This agreement indicates that the laser light emitted from the nanocavity emerges at the output port as the optical spot, demonstrating successful guiding of laser light from the nanocavity via the waveguides in the 3D PC. Comparing the wavelength of the guided mode with the calculations, it is blueshifted by about 100 nm in the experiment. This could be mainly caused by fabrication imperfections (e.g., narrowed rod widths). The polarization dependence for the guided mode, PL images for other spectral peaks in Fig. 4(a), and comparison with the sample without waveguides are shown and discussed in Supplement 1.
In conclusion, we realized simultaneous integration of a nanocavity laser and waveguides in a 3D PC with a cPBG at near-IR wavelengths and succeeded in guiding light from the nanocavity laser via the waveguides. The developed plate-insertion stacking method enabled fabrication of a woodpile 3D PC with 61 layers, which has approximately twice the layers of that realized by previous micromanipulation techniques . The enlarged 3D PC size was the key for realizing the 3D PC circuit. In low-temperature micro-PL measurements, a laser spot was observed near the output port of the 3D PC circuit, demonstrating the successful guiding of laser light from the nanocavity via the waveguides in the 3D PC. Further development of this technique could realize room-temperature operation of 3D PC circuits in the future. We believe that this work is an important step toward integrated 3D photonic circuit applications using 3D PCs with cPBGs.
Japan Society for the Promotion of Science (JSPS) (15H05700, 17H02796); Ministry of Education, Culture, Sports, Science and Technology (MEXT) (15H05868); Materials Education Program for the Future Leaders in Research, Industry, and Technology (MERIT).
We would like to thank M. Nishioka and S. Ishida for their technical support and fruitful discussions.
See Supplement 1 for supporting content.
1. N. Lindernmann, G. Balthasar, D. Hillerkuss, R. Schmogrow, M. Jordan, J. Leuthold, W. Freude, and C. Koos, Opt. Express 20, 17667 (2012). [CrossRef]
2. S. Gross and M. J. Withford, Nanophotonics 4, 332 (2015). [CrossRef]
3. S. J. B. Yoo, B. Guan, and R. P. Scott, Microsyst. Nanoeng. 2, 16030 (2016). [CrossRef]
4. N. Nishiyama, J. Kang, Y. Kuno, K. Itoh, Y. Atsumi, T. Amemiya, and S. Arai, IEICE Trans. Electron. E101-C, 501 (2018). [CrossRef]
5. K. Shang, S. Pathak, B. Guan, G. Liu, and S. J. B. Yoo, Opt. Express 23, 21334 (2015). [CrossRef]
6. S. Zhu and G.-Q. Lo, J. Lightwave Technol. 34, 386 (2016). [CrossRef]
7. W. D. Sacher, J. C. Mikkelsen, P. Dumais, J. Jiang, D. Goodwill, X. Luo, Y. Huang, Y. Yang, A. Bois, P. G.-Q. Lo, E. Bernier, and J. K. S. Poon, Opt. Express 25, 30862 (2017). [CrossRef]
8. J. Chiles, S. Buckley, N. Nader, S. W. Nam, R. P. Mirin, and J. M. Shainline, APL Photon. 2, 116101 (2017). [CrossRef]
9. R. L. Espinola, R. U. Ahmad, F. Pizzuto, M. J. Steel, and R. M. Osgood Jr., Opt. Express 8, 517 (2001). [CrossRef]
10. E. Yablonovitch, Phys. Rev. Lett. 58, 2059 (1987). [CrossRef]
11. S. John, Phys. Rev. Lett. 58, 2486 (1987). [CrossRef]
12. A. Chutinan and S. John, Phys. Rev. B 72, 161316 (2005). [CrossRef]
13. S. Noda, N. Yamamoto, M. Imada, H. Kobayashi, and M. Okano, J. Lightwave Technol. 17, 1948 (1999). [CrossRef]
14. J. D. Joannopoulos, P. R. Villeneuve, and S. Fan, Nature 386, 143 (1997). [CrossRef]
15. S. Noda, J. Lightwave Technol. 24, 4554 (2006). [CrossRef]
16. M. Notomi, Rep. Prog. Phys. 73, 096501 (2010). [CrossRef]
17. S. Noda, M. Fujita, and T. Asano, Nat. Photonics 1, 449 (2007). [CrossRef]
18. O. Painter, R. K. Lee, A. Scherer, A. Yariv, J. D. O’Brien, P. D. Dapkus, and I. Kim, Science 284, 1819 (1999). [CrossRef]
19. S. Noda, Science 314, 260 (2006). [CrossRef]
20. Y. Arakawa, S. Iwamoto, M. Nomura, A. Tandaechanurat, and Y. Ota, IEEE J. Sel. Top. Quantum Electron. 18, 1818 (2012). [CrossRef]
21. S.-Y. Lin, J. G. Fleming, and I.-E. Kady, Appl. Phys. Lett. 83, 593 (2003). [CrossRef]
22. G. Subramania, Y.-J. Lee, A. J. Fischer, T. S. Luk, C. J. Brinker, and D. Dunphy, Appl. Phys. Lett. 95, 151101 (2009). [CrossRef]
23. M. Qi, E. Lidorikis, P. T. Rakich, S. G. Johnson, J. D. Joannopoulos, E. P. Ippen, and H. I. Smith, Nature 429, 538 (2004). [CrossRef]
24. S. A. Rinne, F. G. Santamaría, and P. V. Braun, Nat. Photonics 2, 52 (2008). [CrossRef]
25. I. Staude, G. von Freymann, S. Essig, K. Busch, and M. Wegener, Opt. Lett. 36, 67 (2011). [CrossRef]
26. K. Ishizaki, M. Koumura, K. Suzuki, K. Gondaira, and S. Noda, Nat. Photonics 7, 133 (2013). [CrossRef]
27. A. Tandechanurat, S. Ishida, D. Guimard, M. Nomura, S. Iwamoto, and Y. Arakawa, Nat. Photonics 5, 91 (2011). [CrossRef]
28. E. C. Nelson, N. L. Dias, K. P. Bassett, S. N. Dunham, V. Verma, M. Miyake, P. Wiltzius, J. A. Rogers, J. J. Colema, X. Li, and P. V. Braun, Nat. Mater. 10, 676 (2011). [CrossRef]
29. K. Aoki, D. Guimard, M. Nishioka, M. Nomura, S. Iwamoto, and Y. Arakawa, Nat. Photonics 2, 688 (2008). [CrossRef]
30. S. Takahashi, T. Tajiri, K. Watanabe, Y. Ota, S. Iwamoto, and Y. Arakawa, Electron. Lett. 54, 305 (2018). [CrossRef]
31. S. Kawashima, L. H. Lee, M. Okano, M. Imada, and S. Noda, Opt. Express 13, 9774 (2005). [CrossRef]
32. Y. Ota, M. Kakuda, K. Watanabe, S. Iwamoto, and Y. Arakawa, Opt. Express 25, 19981 (2017). [CrossRef]