High-density photonic integrated circuits (PICs) are expected to replace their current electronic counterparts in the future. The most crucial prerequisite for realizing successful PICs is to develop a low-loss coupling technique between active and passive photonic components based on various nanoscale materials and devices. Here we propose and demonstrate an on-chip integration technique in which a high-refractive-index layer constitutes the coplanar structural backbone across the entire PIC chip. To prove the concept, patterns of a two-dimensional photonic crystal (PhC) band-edge laser and grating couplers are engraved into the backbone layer, and colloidal quantum dots (CQDs) for optical gain are selectively deposited in the PhC area by a conventional lift-off process. Using optical excitation, we observe that the CQD–PhC structure emits coherent single-mode laser light, which is subsequently coupled to and propagates through an adjacent slab waveguide in the well-defined directions corresponding to the selected band-edge point, finally emerging through the grating coupler. Our study demonstrates a simple but highly suggestive PIC platform that automatically guarantees high coupling efficiencies between the micro- and nanophotonic devices to be integrated (through high degrees of modal matching in the vertical direction) and will therefore advance the development of high-density PIC technologies.
© 2017 Optical Society of America
Lasers based on the resonance principles of photonic crystals (PhCs) have been studied actively because of their small sizes and low power consumption [1–7], the virtues for compact light sources to be used in future high-density photonic integrated circuits (PICs). In particular, hybrid PhC lasers, composed of a passive PhC backbone structure and an extrinsic optical gain material, have recently attracted much attention. They offer not only a cost-effective way of preparing compact lasers but also the freedom in choosing gain material from a variety of available ones, including (but not limited to) colloidal quantum dots (CQDs) [8–10], halide perovskites [11,12], transition metal dichalcogenides , organic dyes [14,15], or even biomaterials [16,17]. The authors’ group also demonstrated PhC band-edge lasers by overcoating a layer of either CdSe–CdS–ZnS core-shell-shell CQDs  or CH3NH3PbI3 organometallic perovskite material  on top of a two-dimensional (2D) Si3N4 PhC backbone structure.
Interestingly, the hybrid PhC lasers still offer a unique structural advantage that other types of regular or PhC laser structures cannot. By limiting the spatial area for a PhC laser (in terms of both PhC pattern and optical gain material), one can easily distinguish the zones for active and passive devices. Furthermore, a high optical coupling efficiency between them is automatically guaranteed because they share the same coplanar backbone waveguide layer. This not only facilitates the integration of a PhC laser and a passive waveguide, but also makes its extension to a large-scale PIC conceivable. This scheme is distinguished from the previously reported ones that require enormous efforts in the alignment between different nanoscale/submicron structures [20,21]. In this study, we demonstrate a simple conceptual PIC, in which a hybrid PhC band-edge laser, a slab waveguide, and output couplers are all integrated on a single chip. Upon optical excitation, coherent single-mode laser output is emitted with well-defined in-plane emission directions, and subsequently coupled into the adjacent passive waveguide layer with a guaranteed high coupling efficiency. The coupled laser light propagates through the slab waveguide and is eventually out-coupled into free space through remote grating couplers. To vividly demonstrate directional in-plane emission from a PhC band-edge laser and also the controllability on it, here we intentionally employ a simple slab waveguide structure with no lateral confinement feature. Coupled laser light then propagates freely within the waveguide section but along the directions designated to the band-edge mode. Detailed coupling characteristics to a full 2D transverse waveguide will be addressed in our next work.
2. Results and discussion
Figure 1(a) shows a schematic of our device structure. Constructed in a single passive waveguide layer are a PhC pattern (for the formation of the band-edge laser), a waveguide section (with no specific waveguide pattern in the lateral direction), and a grating coupler. We choose silicon nitride (Si3N4; n = 2.01) and fused silica (SiO2; n = 1.46) as the backbone waveguide layer and substrate, respectively, because of the high refractive index contrast and very low absorption in the visible to near-infrared wavelength range. Thus the Si3N4 thin-film layer constitutes an asymmetric slab waveguide, cladded by air above and the silica substrate below. The thickness of the Si3N4 slab waveguide is designed to support only the fundamental guided mode in the vertical direction. Engraved into the Si3N4 backbone layer are a 2D PhC air-hole array for a PhC band-edge laser and grating boxes for output couplers. Densely packed CQDs are then selectively deposited directly on top of the PhC pattern to provide optical gain for stimulated emission in red (λ ~615 nm). A band-edge laser fabricated as such offers a continuous coupling interface to the adjacent passive waveguide section, which is composed of the Si3N4 layer itself (with neither structural pattern nor CQDs atop).
Figure 1(b) shows the refractive index profiles of the two sections and the corresponding transverse electric (TE) guided mode profiles, which visually demonstrate a high degree of mode match. For the refractive index of the CQD-capped PhC laser section, we take the geometrical average of the refractive indices of the two materials, CQDs and Si3N4. Simple mode overlap integration indicates that the butt coupling efficiency  between the two sections can be as high as Γ ≈95% (see Appendices for details). For confirmation, we also perform finite-difference time-domain (FDTD) simulations using commercial software (FDTD Solution, Lumerical Solutions). Figure 1(c) shows the simulated guided mode profile (Ey) near the joint between the PhC and the waveguide sections, which is obtained by launching from the left the fundamental guided mode supported by the PhC laser structure. The mode is guided well except for a small perturbation at the joint; consequently, most of the incident power (~0.955) is transmitted through the joint. A similar FDTD simulation is performed for the grating out-coupler. As shown in Fig. 1(d), a significant amount of the electromagnetic energy flux of the waveguide mode is coupled out into free space within the coupler section, which is composed of a second-order grating.
Figure 2(a) shows the photonic band structure of our 2D PhC band-edge laser structure for TE polarization, which is calculated using the 2D plane-wave expansion method. The PhC model structure is composed of a square-lattice array of air-holes that is etched into the Si3N4 PhC backbone layer (t = 140 nm) and overcoated with CQDs. The refractive index of the densely packed CdSe–CdS–ZnS core-shell-shell CQDs was obtained from spectroscopic ellipsometry measurements of an independently prepared film (see Appendices for details; the chemical synthesis of CdSe–CdS–ZnS CQDs can be found in [23,24]). It is well known that the photon group velocity at a band-edge mode vanishes (υg = dω/dk = 0) and that the optical gain is then effectively enhanced, enabling lasing action . We use the X- or M-point band-edge modes below the light lines—marked by circles in Fig. 2(a)—because they can afford in-plane lasing action with well-defined emission directions [26,27], which is suitable for on-chip integration with other photonic components.
Figures 2(b) and 2(c) are simulated mode spectra for PhC lattice constants of a = 185 and 255 nm, respectively, where the air-hole radius is fixed at r = 0.26a. The two lattice constants correspond to the conditions under which the X- and M-point band-edge modes, respectively, coincide with the red emission band of the CQDs used in this study (λc ~615 nm; ΔλFWHM ~32 nm). The simulations are performed using the three-dimensional FDTD method with periodic boundary conditions. Each case exhibits two major distinct modes (M1 and M2 are narrowly spaced nondegenerate modes), which correspond to two neighboring band-edge modes, one belonging to the air band and the other to the dielectric band. The insets show simulated mode profiles; each image displays the in-plane electric field intensity (Et2 = Ex2 + Ey2) at the center plane of the Si3N4 backbone layer. The two sets of X-point band-edge modes are expected to have in-plane momenta along the square Bravais lattice vectors, which is exactly what we see from the mode profiles in the insets of Fig. 2(b). The difference between them is the relative positions of the nodes and antinodes: the X1/X2 modes (belonging to the dielectric band) have antinodes in the Si3N4 dielectric region, whereas those of the X3/X4 modes (belonging to the air band) run through the air-holes. Therefore, the X3/X4 modes have greater overlap with the CQDs and thus are likely to lase; according to our evaluation, the CQD overlap factors are 37.1% for X3/X4 and 5.9% for X1/X2. Similar arguments can be made for the M-point band-edges modes (except that their in-plane momenta are along the diagonal directions of the square lattice); the M3/M4 modes, which belong to the air band, have a much higher modal overlap with the CQDs (25.7%) than the dielectric band-edge modes do (6.6% for M1 and 8.4% for M2), and thus have a better chance of lasing.
To make real devices, we first prepared a clean fused silica substrate. A Si3N4 thin film was deposited on the substrate by plasma-enhanced chemical vapor deposition to a nominal thickness of 140 nm. An array of simple PIC patterns was then defined using electron-beam lithography. Each PIC pattern contained a PhC laser in the center and eight grating couplers surrounding the PhC laser but away from it, as shown in Fig. 3(a). The electron-beam resist patterns were then transferred to the underlying Si3N4 layer by reactive-ion etching. The PhC laser pattern (50 μm in width) was selectively covered with a CQD layer (80 nm in thickness) by lifting off the CQD layer after spin-coating it from solution (2.1 wt% in cyclohexane; 1000 rpm). The eight grating coupler boxes (each 10 μm × 30 μm in size) face the eight side edges of the central octagon, with a slab waveguide region in between (50 μm in length). We did not define any lateral confinement structure in this slab waveguide so that we could attempt to confirm directional emission and propagation of the band-edge laser beams by observing selective lighting of the grating boxes located in the corresponding directions. Figures 3(b) and 3c are scanning electron microscope (SEM) images of the PhC laser pattern (a = 255 nm) and grating regions (Λ = 310 nm; fill factor = ~0.65), respectively, before selective CQD coating, and Figs. 3(d) and 3(e) are tilted SEM images of the 2D PhC region (after CQD lift-off) and a grating coupler box, respectively. Figure 3(d) and its inset clearly demonstrate that CQDs not only cover the PhC-patterned area but also fill the air-holes uniformly.
We measured the microphotoluminescence of the fabricated PIC devices. Each device was optically activated using a frequency-doubled Nd:YAG laser (λex = 532 nm) at room temperature. We operated the laser in a pulsed mode at a low duty cycle (400 ps pulse width and 1 kHz repetition rate) to minimize the possibility of thermal damage. The excitation beam was focused downward onto the PhC band-edge laser with a spot size slightly larger than the octagonal PhC pattern using a 5 × objective lens (N.A. = 0.12). The PhC laser emission was captured from the top using the same objective lens and fed into a spectrometer (DW700, DongWoo Optron) for analysis. Figure 4(a) shows emission spectra from the a = 185 nm sample recorded at a few different excitation powers. Single-mode laser emission at λ ≈624 nm is clearly seen with a threshold pulse energy density of ~1 mJ/cm2 [Fig. 4(b)]. Figures 4(c) and 4(d) are sample images taken with a charge-coupled device (CCD) camera while the device was optically excited below and above the laser threshold, respectively. A notch filter was installed in front of the CCD camera to eliminate scattered and stray light from the 532 nm excitation beam. Below the laser threshold, the device exhibits faint but uniform emission only from the CQD-coated PhC region, which is characteristic of spontaneous emission. In contrast, however, the emission image changes dramatically above the threshold; it becomes much brighter and exhibits vivid speckle patterns indicative of lasing. More importantly, there is additional emission from the four grating boxes located on the x and y axes. This is strong evidence that lasing occurs at the X-point band-edge modes, which is consistent with the fact that this PhC structure (a = 185 nm) is designed to operate at these modes [Fig. 2(b)]. The experimental CCD image shown in Fig. 4(d) is consistent with a numerical simulation result based on the 2D FDTD method (see Appendices for details).
The characteristics of the X-point band-edge lasing can be assessed further by examining its polarization dependence, which can be done by inserting linear polarizers into our microphotoluminescence setup. Figures 4(e) and 4(f) are emission images taken when the polarizer angle is set to θ = 0° and 90°, respectively. When θ = 0° (90°) [or the electric field is along the x (y) axis], only the band-edge mode oscillating in the y (x) direction is visible. This observation implies that we could discriminate the two degenerate X-point band-edge modes (presumably X3/X4) whose oscillation directions are orthogonal to each other. The emission spectra in Figs. 4(g)–4(j) were obtained under the same conditions as in Figs. 4(c)–4(f), respectively. One can clearly see not only the difference between the emission resulting from excitation below and above the threshold (in terms of the emission spectral widths) but also the contributions of the two degenerate band-edge modes to the total emission intensity.
Similar arguments can be made for the a = 255 nm sample, which is designed for M-point band-edge lasing. Figure 5(a) shows emission spectra captured at various excitation levels. Again, sharp single-mode laser emission (λ ≈616 nm) is obtained over a wide range of excitation pulse energy densities. The laser threshold (~2 mJ/cm2) is about twice that of the X-point band-edge laser, as shown in Fig. 5(b), which we attribute to a lower Q factor and a smaller modal overlap with the gain volume for the M-point band-edge mode. The emission images of the sample in Figs. 5(c)–5(f) are counterparts of those in Figs. 4(c)–4(f) for the X-point band-edge laser. Whereas the excited CQD–PhC area glows dimly below the threshold [Fig. 5(c)], intense emission occurs with distinct speckle patterns not only from the CQD–PhC region but also from the grating couplers located at the four corners (in the diagonal directions) [Fig. 5(d)]. This is a direct indication that lasing indeed occurs at the M-point band-edge modes, which is also supported by the FDTD simulation result for these modes. The polarization dependence of the laser emission patterns shown in Figs. 5(e) and 5(f) reveals that the M-point lasing is composed of two orthogonally degenerate modes. We assume that lasing occurs at the M3/M4 modes because our CQD–PhC hybrid laser platform prefers the band-edge modes in the air band to those in the dielectric band (owing to a higher modal overlap factor with the CQD gain material). The emission spectra shown in Figs. 5(g)–5(j) confirm the lasing action and contributions of the two degenerate band-edge modes.
In conclusion, we designed and fabricated a hybrid PhC band-edge laser device and demonstrated its integration with other passive photonic components (waveguides and grating couplers) on a single Si3N4 thin film platform. When optically excited, the CQD–PhC devices lased, emitting intense coherent light in the lateral directions predetermined by the particular band-edge mode employed, and were eventually out-coupled through grating couplers. Our scheme for the integration of active and passive photonic devices on a single chip guarantees a high coupling efficiency and thus should be an important milestone toward future high-density PICs with a variety of sophisticated functionalities.
Butt coupling efficiency
Coupling efficiency at an abrupt interface between two waveguides can be roughly estimated by the overlap integral:
FDTD simulation for the interface of CQD-PhC and Si3N4 waveguides
FDTD simulations are performed using a commercial software (FDTD solution, Lumerical Solutions). The TE-polarized fundamental mode supported by the CQD-PhC waveguide is obtained beforehand, which is then followed by the launch of a mode source in pulsed form (pulse width = 4000 fs; wavelength = 615 nm) at the far side of the CQD-PhC waveguide section. We let the launched mode stabilize through the CQD-Si3N4 PhC waveguide first, encounter the waveguide interface, and finally propagate through the Si3N4 waveguide. The pulse duration is long enough so that the launched mode source reaches its constant intensity, as shown in the snapshot of Fig. 1(c). Then transmitted powers before and after the interface are compared to obtain the coupling efficiency; two power monitors are located at x = ±12.5 μm on both sides of the interface.
FDTD simulation on the grating coupler
The TE-polarized fundamental mode source (pulse width 4000 fs, wavelength = 615 nm) is launched from the far side of the Si3N4 waveguide and let propagate through the grating coupler region; a resultant snapshot is shown in Fig. 1(d). Optical power coupled out through the grating is monitored as the structural parameters (such as period and fill factor) are scanned. When the grating period and fill factor are Λ = 310 nm and FF = 0.65, the maximal out-coupling efficiency of 66% is obtained with 39% and 27% coupled out to the upper and lower claddings, respectively.
Spectroscopic ellipsometer measurements on the CQD film
The complex refractive index dispersions of densely packed CdSe/CdS/ZnS core-shell-shell CQD material is obtained from spectroscopic ellipsometer measurement on an independently prepared CQD film, which are summarized in Fig. 6.
FDTD simulations of photonic crystal band-edge modes
To characterize the photonic band-edge modes, 3D FDTD simulations are performed for the Bloch boundary points: (kx = π/a, ky = 0) for the X symmetry point and (kx = π/a, ky = π/a) for the M symmetry point. Perfect matching layers are applied to the top and bottom of the simulation field. Multiple electric dipole sources are distributed randomly (in terms of positions and orientations) within the waveguide layer, which are then operated for a few fs pulse duration. Data are collected and analyzed for 5000 fs after the pulsed launch of dipole sources. Electric field magnitudes in Fig. 2 and Fig. 7 are integrated over the time between 1000 fs and 5000 fs.
FDTD simulations on photonic crystal band-edge mode propagations
In order to investigate the overall modal behavior (including both mode generation and propagation), 2D FDTD simulations are performed for an extended area that includes the regions of not only the square lattice photonic crystal but also a part of the planar waveguide. To facilitate the simulations, the refractive index contrast between the two materials is increased to 2 (instead of ~0.23) while the width of the photonic crystal region is reduced to 3.5 μm (instead of 50 μm). For each band-edge mode, 40 electric dipole sources with random distributions and orientations are allowed for a few fs pulse duration. Electric field intensity is then integrated over the time interval between 2500 fs and 3000 fs after the launch of multiple dipole sources. Figure 8 vividly shows not only the stable photonic crystal band-edge modes in the center but also their subsequent propagation in the appropriate directions.
Microphotoluminescence measurements and imaging are performed using the setup of which configuration is schematically illustrated in Fig. 9.
National Research Foundation of Korea (NRF) (2014R1A2A1A11051576).
References and links
2. S. Nojima, “Optical-gain enhancement in two-dimensional active photonic crystals,” J. Appl. Phys. 90(2), 545–551 (2001). [CrossRef]
3. J. D. Joannopoulos, P. R. Villeneuve, and S. Fan, “Photonic crystals: Putting a new twist on light,” Nature 386(6621), 143–149 (1997). [CrossRef]
4. O. Painter, R. K. Lee, A. Scherer, A. Yariv, J. D. O’Brien, P. D. Dapkus, and I. Kim, “Two-dimensional photonic band-gap defect mode laser,” Science 284(5421), 1819–1821 (1999). [CrossRef] [PubMed]
5. D. Ohnishi, T. Okano, M. Imada, and S. Noda, “Room temperature continuous wave operation of a surface-emitting two-dimensional photonic crystal diode laser,” Opt. Express 12(8), 1562–1568 (2004). [CrossRef] [PubMed]
6. M. Nomura, S. Iwamoto, K. Watanabe, N. Kumagai, Y. Nakata, S. Ishida, and Y. Arakawa, “Room temperature continuous-wave lasing in photonic crystal nanocavity,” Opt. Express 14(13), 6308–6315 (2006). [CrossRef] [PubMed]
7. 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(5689), 1444–1447 (2004). [CrossRef] [PubMed]
8. F. Fan, O. Voznyy, R. P. Sabatini, K. T. Bicanic, M. M. Adachi, J. R. McBride, K. R. Reid, Y.-S. Park, X. Li, A. Jain, R. Quintero-Bermudez, M. Saravanapavanantham, M. Liu, M. Korkusinski, P. Hawrylak, V. I. Klimov, S. J. Rosenthal, S. Hoogland, and E. H. Sargent, “Continuous-wave lasing in colloidal quantum dot solids enabled by facet-selective epitaxy,” Nature 544(7648), 75–79 (2017). [CrossRef] [PubMed]
9. C. Dang, J. Lee, C. Breen, J. S. Steckel, S. Coe-Sullivan, and A. Nurmikko, “Red, green and blue lasing enabled by single-exciton gain in colloidal quantum dot films,” Nat. Nanotechnol. 7(5), 335–339 (2012). [CrossRef] [PubMed]
10. C. Dang, J. Lee, K. Roh, H. Kim, S. Ahn, H. Jeon, C. Breen, J. S. Steckel, S. Coe-Sullivan, and A. Nurmikko, “Highly efficient, spatially coherent distributed feedback lasers from dense colloidal quantum dot films,” Appl. Phys. Lett. 103(17), 171104 (2013). [CrossRef]
11. S. Chen, K. Roh, J. Lee, W. K. Chong, Y. Lu, N. Mathews, T. C. Sum, and A. Nurmikko, “A photonic crystal laser from solution based organo-lead iodide perovskite thin films,” ACS Nano 10(4), 3959–3967 (2016). [CrossRef] [PubMed]
12. M. Saliba, S. M. Wood, J. B. Patel, P. K. Nayak, J. Huang, J. A. Alexander-Webber, B. Wenger, S. D. Stranks, M. T. Hörantner, J. T.-W. Wang, R. J. Nicholas, L. M. Herz, M. B. Johnston, S. M. Morris, H. J. Snaith, and M. K. Riede, “Structured organic–inorganic perovskite toward a distributed feedback laser,” Adv. Mater. 28(5), 923–929 (2016). [CrossRef] [PubMed]
13. S. Wu, S. Buckley, J. R. Schaibley, L. Feng, J. Yan, D. G. Mandrus, F. Hatami, W. Yao, J. Vučković, A. Majumdar, and X. Xu, “Monolayer semiconductor nanocavity lasers with ultralow thresholds,” Nature 520(7545), 69–72 (2015). [CrossRef] [PubMed]
14. S. Toffanin, S. Kim, S. Cavallini, M. Natali, V. Benfenati, J. J. Amsden, D. L. Kaplan, R. Zamboni, M. Muccini, and F. G. Omenetto, “Low-threshold blue lasing from silk fibroin thin films,” Appl. Phys. Lett. 101(9), 091110 (2012). [CrossRef]
16. M. C. Gather and S. H. Yun, “Single-cell biological lasers,” Nat. Photonics 5(7), 406–410 (2011). [CrossRef]
18. H. Chang, K. Min, M. Lee, M. Kang, Y. Park, K.-S. Cho, Y.-G. Roh, S.-W. Hwang, and H. Jeon, “Colloidal quantum dot lasers built on a passive two-dimensional photonic crystal backbone,” Nanoscale 8(12), 6571–6576 (2016). [CrossRef] [PubMed]
19. H. Cha, S. Bae, M. Lee, and H. Jeon, “Two-dimensional photonic crystal bandedge laser with hybrid perovskite thin film for optical gain,” Appl. Phys. Lett. 108(18), 181104 (2016). [CrossRef]
20. H.-G. Park, C. J. Barrelet, Y. Wu, B. Tian, F. Qian, and C. M. Lieber, “A wavelength-selective photonic-crystal waveguide coupled to a nanowire light source,” Nat. Photonics 2(10), 622–626 (2008). [CrossRef]
21. M. Tchernycheva, A. Messanvi, A. de Luna Bugallo, G. Jacopin, P. Lavenus, L. Rigutti, H. Zhang, Y. Halioua, F. H. Julien, J. Eymery, and C. Durand, “Integrated photonic platform based on InGaN/GaN nanowire emitters and detectors,” Nano Lett. 14(6), 3515–3520 (2014). [CrossRef] [PubMed]
22. L. Pavesi and G. Guillot, Optical interconnects: The silicon approach (Springer, 2007)
23. J. Lim, S. Jun, E. Jang, H. Baik, H. Kim, and J. Cho, “Preparation of highly luminescent nanocrystals and their application to light‐emitting diodes,” Adv. Mater. 19(15), 1927–1932 (2007). [CrossRef]
24. K.-S. Cho, E. K. Lee, W.-J. Joo, E. Jang, T.-H. Kim, S. J. Lee, S.-J. Kwon, J. Y. Han, B.-K. Kim, B. L. Choi, and J. M. Kim, “High-performance crosslinked colloidal quantum-dot light-emitting diodes,” Nat. Photonics 3(6), 341–345 (2009). [CrossRef]
25. J. P. Dowling, M. Scalora, M. J. Bloemer, and C. M. Bowden, “The photonic band edge laser: A new approach to gain enhancement,” J. Appl. Phys. 75(4), 1896–1899 (1994). [CrossRef]
26. M. Imada, A. Chutinan, S. Noda, and M. Mochizuki, “Multidirectionally distributed feedback photonic crystal lasers,” Phys. Rev. B 65(19), 195306 (2002). [CrossRef]
27. R. M. Cazo, C. L. Barbosa, H. T. Hattori, and V. M. Schneider, “Steady‐state analysis of a directional square lattice band‐edge photonic crystal lasers,” Microw. Opt. Technol. Lett. 46(3), 210–214 (2005). [CrossRef]