This paper reports on large field-of-regard, high-efficiency, and large aperture active optical phased arrays (OPAs) for optical beam steering in LIDAR systems. The fabricated 5 mm-long silicon photonic OPA with a 1.3 μm waveguide pitch achieved adjacent waveguide crosstalk below −12dB. A relatively large and uniform emission aperture has been achieved with a low-contrast silicon nitride assisted grating (~20 dB/cm) whose emission profile can be further optimized using an apodized design. The fabricated silicon-photonic OPA demonstrated > 40° lateral beam steering with no sidelobes in a ± 33° field-of-regard and 3.3° longitudinal beam steering via wavelength tuning by 20 nm centered at 1550 nm. We have fully integrated the silicon photonic OPA device with electronic controls and successfully demonstrated 2-dimensional coherent optical beam steering of pre-planned far-field patterns. Future improvements include placement of a distributed Bragg reflector (DBR) underneath the grating emitter in order to achieve nearly a factor of two improvement in emission efficiency.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Light detection and ranging (LIDAR) provides for relatively long-range and high-precision 3D imaging, essential for future autonomous vehicles. Compared to RADAR, LIDAR can resolve fine features that can distinguish humans from other objects. Compared to camera systems, LIDAR provides additional ranging and motion detection capabilities in 3D and functions under day and night conditions. Notably, recent advances in software technologies applied to conventional 2D camera imaging for 3D interpretations fail to work consistently under low light conditions. By contrast, LIDAR offers direct data in the form of accurate measurements of distance and velocity of the surrounding objects. However, to-date, most of the commercially available LIDAR systems utilize mechanical beam-steering, which in turn makes the entire system bulky, slow, and unreliable.
Optical phased array (OPA) technology offers a promising non-mechanical optical beam steering at chip-scale. Recent advances in silicon photonics available from a number of foundry manufacturing services have enabled large-scale integrated OPAs to be fabricated in large volume at low cost. Thanks to high refractive index contrast and mature CMOS-compatible fabrication processes, the silicon photonic platform offers strong light confinement, enabling small element pitch while keeping optical loss manageable. For such OPAs, waveguide arrays with subwavelength element pitch and low optical crosstalk are necessary to sidelobe-limited angle steering and to high emission power efficiency contained within a diffraction-limited far-field spot . Current demonstrations of integrated OPAs typically have element pitches larger than the operating wavelength and a relatively small (< mm2) active aperture [2–6], which limits the field-of-regard and limits the operation range of the LIDAR system based on the integrated OPAs. Very recently, >4 mm2 aperture size has been demonstrated with element pitch of 1.65 μm  and a 1.3 μm pitch OPA has been demonstrated with 0.65 mm2 aperture size .
This paper reports on large field-of-regard (theoretically 70°), high-efficiency, and large aperture active OPAs for optical beam steering in LIDAR systems. We design and fabricate high efficiency OPAs with waveguide pitch down to 1.3 μm and active emission area larger than 1 mm2 using a standard silicon photonic fabrication process. Figure 1 illustrates the schematic of our fabricated active OPA device as a key component for 2D coherent optical beam steering. We discuss in detail design optimization of each component and of the entire system towards achieving large field-of-regard, high efficiency, and large emitting aperture active OPAs. The simulation and measurement results demonstrate low waveguide crosstalk (< −12 dB over 5 mm interaction length). Such sub-wavelength waveguide pitch enables a sidelobe-limited steerable angle of ± 35° and a sidelobe-free lateral steering of ± 5°. While non-uniform waveguide array design  can support a half-wavelength waveguide pitch at low crosstalk to achieve a single-lobe ± 90° beam steering at 1.55 μm wavelength, it is not attempted here, since our primary goal is to achieve both longitudinal and lateral beam steering simultaneously combining OPA and diffraction grating effects. In order to achieve a relatively low grating emission rate across the 5 mm waveguide length, we adopted a low-contrast silicon nitride (SiN) assisted grating design. The experimental results using a 24-element silicon photonic OPA demonstrate 3.3° axial beam steering with 20 nm wavelength tuning and > 40° lateral beam steering with no sidelobes in a ± 33° field-of-regard. We have further fully integrated a 120 element OPA device with electronic controls and successfully demonstrated 2D coherent optical beam steering.
2. System design and component optimization
2.1 Element pitch and optical crosstalk reduction
Large field-of-regard single-lobe OPA beam-steering requires waveguide pitches to be below the operating wavelength. Figure 2(a) shows the calculated sidelobe-free steering range of the OPA at 1550nm wavelength for waveguide pitches. Ideally, the element pitch needs to be less than or equal to half-λ for hemispherical steering without side lobes. Unwanted side lobes will limit the maximum steering angle and undesirable power loss in the main beam. Figure 2(b) shows the simulated far-field patterns from a 24 element OPA with element pitch of 2 μm and 1.3 μm. In an integrated OPA system, the element pitch is typically limited by the optical crosstalk from the adjacent element.
To suppress adjacent channel optical crosstalk, we fully etched silicon waveguides, which reduces the optical coupling from the adjacent waveguides. Figure 3(a) illustrates a schematic of two fully etched silicon waveguide next to each other. We use a finite element tool (COMSOL RF module) to simulate the optical coupling from an adjacent waveguide shown in Fig. 3(a). The coupling length L, is estimated as follows:
Where neven is the refractive index of the symmetric mode of the super waveguide and nodd is the refractive index of the asymmetric mode of the super waveguide.
Our FEM simulation suggests a minimum element pitch of 1.3 µm for <10% coupling over 5mm interaction length for a waveguide width of 500 nm, as shown in Fig. 3(b). We choose the silicon layer thickness to be 500 nm in the splitter and phase modulator region to minimize the optical coupling in that area and also for future integration with low power InP on Si phase modulators . The silicon layer thickness in the waveguide grating region is chosen to be 300 nm for single mode operation.
We have designed directional coupler test structures with different length to measure the optical crosstalk from the adjacent waveguide. Figure 4(a) shows the top-view scanning electron microscope (SEM) picture of the fabricated coupling test structures and Fig. 4(b) reveal the exact dimension in the coupling region with 1.3 µm pitch. We measured <-12 dB crosstalk for 1.3 µm pitch (Fig. 4(c)) and a <-20 dB crosstalk for 1.5 µm pitch (Fig. 4(d)) over 5mm.
To further enlarge the steering range, spare non-uniform array  and beta mismatched waveguide array  have been investigated. However, demonstrated OPAs have either low overall optical efficiency or limited to 1D beam steering with end-firing waveguide outputs.
2.2 System efficiency improvement
This section discusses our methods for system power efficiency improvements. In particular, we have (1) reduced the component excess loss and (2) enhanced the grating radiation efficiency by inserting a DBR under the grating OPA layer for reflecting the downward emission of the grating.
The tunable laser emission into the single mode silicon photonic waveguide is split by binary-tree waveguides using 1 × 2 multimode interference (MMI) splitters where equal-pathlength and equal-power splitting are achieved with relatively low excess loss. Figure 5(a) shows the top-view SEM pictures of a fabricated single stage MMI splitter. The MMI design employed finite-difference-time-domain (FDTD) simulations (Lumerical) to minimize the excess loss. Figure 5(b) plots the simulated and measured MMI excess loss as a function of MMI length, LMMI. We obtained a minimum excess loss of 0.08 ± 0.05 dB per MMI stage with LMMI = 16 μm.
As noted in Section 2.1, we have adopted different thickness at the MMI, phase shifter and grating waveguide regions. Therefore, a low-loss taper is necessary at the silicon thickness transition region going from the 500 nm thickness waveguide to the 300 nm thickness waveguide. Previous work has achieved 0.6 dB per transition loss with an etched taper . Here, we utilized the local oxidation of silicon (LOCOS) process  to form a smooth vertical taper. Figures 6(a) and 6(b) show the top-view and 75°-tilted-view SEM of the fabricated vertical tapers on top of the grating waveguide. The minimum feature size of the taper is ~100 nm. Figure 6(c) reveals the detailed structure of the smooth vertical taper at the transition tip before the waveguide etching. We measured the vertical taper transition loss using cascaded transition structures and measured loss was less than 0.1 dB per transition.
Another critical loss reduction requires grating emission efficiency improvement. In a standard diffraction grating waveguide design, waveguide gratings emit evenly in both upward and downward directions. With sophisticated dual layer grating design with λ/4 shift in both vertical and lateral directions, unidirectional emission characteristic can be achieved with > 90% efficiency . In this work, we designed and deposited 2 pairs of a-Si/SiO2 DBRs placed under the grating waveguide layer to reflect upward the undesirable downward emission. As the designed DBR has a broad reflection band, it is less sensitive to the layered thickness variation compared to the demonstrated method . Figure 7(a) shows the cross-sectional SEM picture of deposited 2 pairs of DBR on a silicon substrate for reflectance measurement. The designed thickness from top to bottom are: 108 nm/ 269 nm/ 108 nm/ 269 nm and measured thickness are: 104 nm/ 288 nm/ 121 nm/ 270 nm. We characterized normal incidence DBR reflectance using a Fourier-transform infrared spectroscopy (FTIR) normalized to gold-coated reference sample (97.8% @ 1550 nm). Figure 7(b) plots the measured reflectance of different samples in 1500 to 1600 nm wavelength range. We extracted a >82% reflectance for 1 pair of DBR and a >94% reflectance for 2 pairs of DBR.
We further optimized the grating integrated with DBRs using a FDTD simulation tool (Lumerical). Figure 8(a) shows the cross-sectional schematic of the simulation setup. By sweeping the top (Ttop) and bottom (Tbottom) oxide cladding thickness of the waveguide grating, we can obtain a maximum upward emission efficiency ~95% (Fig. 8(b)) with Ttop = 700 nm and Tbottom = 800 nm while only 46% can be achieved without the DBR (Fig. 8(c)). Test structures for the grating upward efficiency measurement and comparison are currently in progress.
2.3 Towards large emitting aperture
Large emitting aperture areas are desirable to form low-divergence far-field optical beam when transmitting the laser signal and to efficiently collect signal reflected from a remote object. To achieve a large emitting aperture over a 5 mm length, we designed a weakly emitting structure using silicon nitride (SiN) assisted waveguide gratings. We deposited ~80nm SiN on top of the silicon waveguide and partially etched 60 nm to form a weakly emitting grating. Figure 9(a) shows the schematic and top-view and cross-sectional-view of the SiN assisted grating. We measured an emission rate of 23 dB/cm (~95% power emitted over 5mm length) from a test structure (Fig. 9(b)). The emission intensity profile of the grating is measured by scanning across the grating. Figure 9(c) shows the captured near-field infrared (IR) images at different sections of the grating.
To further enlarge the emitting aperture, we have designed and initially fabricated a beta matched apodized grating by tuning the SiN grating teeth-width and duty-cycle while keeping the waveguide propagation constant (beta) to be invariant. Detailed design, fabrication, and experimental demonstration can be found in .
3. Fabrication and component characterization
We fabricated our devices using a standard silicon photonic fabrication process. Figure 10(a) shows a fabricated 120-element OPA with a waveguide pitch of 2.0 μm. A vertical taper from the 500 nm to the 300 nm silicon waveguide region is first formed using the LOCOS process. Then the waveguides are fully etched to minimize the optical coupling and to eliminate phase perturbations induced by etching depth variations. The measured waveguide propagation loss is 2.6 dB/cm, which is comparable to the reported values from the foundry . We used an MMI-tree based splitter to equally distribute the power into each element (Fig. 10(b)). A SiN assisted grating is then formed as described in section 2.3 (Fig. 10(d)). We deposited another 200 nm SiN layer to form a low-loss fiber coupler and optical power transmission from Si to SiN is achieved using vertical evanescent couplers . The total coupling loss from a 5 μm diameter lensed fiber to 300 nm silicon region is less than 3 dB. 120 nm of Ti/Au metal is deposited to form the thermal-optical phase tuner for a proof-of-concept demonstration (Fig. 10(c)). In the phase tuner session, the pitch is fanned out to 38 μm to reduce thermal crosstalk from adjacent channels.
We characterized the thermal optical phase tuner tuning efficiency using an asymmetric Mach-Zehnder interferometer (AMZI) structure. Figure 11(a) plots the transmitted optical intensity of the AMZI at 1550.8 nm as a function of heating power. We extracted a heating power required for 2π phase shift of 22 mW. The thermal optical tuners are staggered to reduce the thermal crosstalk between each element and to ease heat dissipation. We extracted the thermal crosstalk from the redshift of the AMZI fringes from a heat source place 38 μm away from the AMZI arm (Fig. 11(b)). The measured thermal crosstalk from the adjacent channels is 5%.
To scale up the element number and OPA size, low power phase tuner like the p-n diode based silicon electro-optical (EO) modulator (~μW)  or hybrid silicon EO modulator (~nW) with III-V materials on top  should be used at the expenses of extra loss and more complicated fabrication process.
4. Beam steering characterization
4.1 Far-field setup and phase error correction
Figure 12(a) shows a two-lens far-field measurement setup used for characterizing our OPA beam steering device. The measurement setup has an NA of 0.55, corresponding to a field-of-regard of ± 33°. We mount our 24 element OPA chips on a metal chuck with TEC to control the chip temperature. All the thermal tuners are wire-bonded to a PCB and controlled by digital-to-analog converters (DACs). The laser signal is fed into the SiN edge coupler using a lensed fiber. Figure 12(b) shows the measured far-field IR image without applying voltages to the thermal tuners. We apply a gradient descent based algorithm  to correct fabrication induced phase errors. After 10 iterations, we were able to produce sharp spots at desired far-field locations, as shown in Fig. 12(c). We estimated an overall optical efficiency to be −13.7dB and it can be decomposed into the fiber-to-chip coupling loss (2.5 dB), the MMI splitting network excess loss (0.6 dB), the waveguide propagation loss (5.2 dB) and the grating emission loss (3.4 dB for downward emission, 2 dB for imperfection in co-phasing). We are expecting a 3.2 dB improvement from the grating by inserting a bottom DBR and a ~4dB improvement from the propagation loss can be achieved with advanced lithography .
4.2 Axial and lateral direction beam steering
We demonstrated axial and lateral beam steering via wavelength tuning and phase tuning, respectively. Figure 13(a) shows the measured far-field spot at different laser wavelengths. We measured an axial direction tuning efficiency of 0.15 o/nm. Figure 13(b) shows the measured far-field spot at 3 different sets of thermal tuner voltages. For the 1.3μm pitch OPA, we demonstrated >40° steering angle without any sidelobe within the field-of-regard. Theoretically, we should be able to achieve sidelobe limited maximum steerable angle of ± 35°. The measured side-lobe-suppression (~7 dB) is relatively low compared to the simulated value (~13 dB). We attributed the relatively low side-mode-suppression partly to our unoptimized gradient descent algorithm and partly to the imperfect working phased shifters (2 open circuit after the fab and an extra 1 open circuit after the wire-bonding) in this particular OPA. From our simulation of the 24 channel OPA with random phase error on each channel, we were only able to achieve ~7dB side-mode-suppression even after 100 times iteration. With advanced optimization algorithm , higher side-mode-suppression could be achieved.
5. System assembly and 2D beam steering demo
Following device testing and characterization, we fully packaged our OPA device into a computer controlled 2D optical beam steering system. Figure 14(a) shows a photo of a fully packaged coherent 2D optical beam steerer with electronic controls. Figure 14(b) shows a zoom-in of the packaged 120 channel OPA chip on a PCB. Following careful system calibration, we demonstrated 2D optical beam steering through tuning OPA phase for lateral beam steering and tuning wavelength for longitudinal beam steering. Figure 14(c) shows a composite photo of over 100 far field beam patterns trained in forming the Lockheed Martin logo.
6. Future LIDAR system integration
For applications that require far greater operating ranges than automotive applications (e.g. >>1 km), aperture and power scaling is required. Such scaling, while maintaining high emission area fill factor (>95%), requires 3D integration of optical layers as well as electronics for semiconductor optical amplifier drivers and phase shifters. Extensive work has been carried out to demonstrate 3D integration aspects and is described in greater detail in . Figure 15 illustrates some of these efforts. From left to right: two-layer optical chip with splitters and phase shifters in one layer and emitter grating in second layer; sub-micron 45° mirror pairs used to couple light between layers; ultrafast laser inscription (ULI) used to write 3D optical waveguides of arbitrary 3D shape for stitching parts together optically . With each unit cell handling an emitting optical power >10 mW (assuming −10 dB efficiency with 100 mW input limited by the two photon absorption induced free carrier loss ), > 1W LIDAR system can be formed with a 10 × 10 or larger array.
In summary, we have designed, fabricated and demonstrated a large field-of-regard, high efficiency and large emitting aperture active OPA. Fabricated 24 element OPA devices have 5 mm long emitting area and waveguide pitch down to 1.3 μm with adjacent waveguide crosstalk < −12 dB. We demonstrated a 3.3° axial beam steering upon 20nm wavelength tuning and > 40° lateral beam steering with no sidelobe in ± 33° field-of-regard. We fully packaged a 120 element OPA chips into an electronic-optical system and demonstrated 2D optical beam steering. Such devices can be further scaled up to a 1 cm × 1 cm footprint and used as unit cells in larger assemblies. With 3D integration schemes that use vertical U-shaped couplers  and with further development of close-packed InP-on-Si low power phase modulator, it appears feasible to tile such unit cells to form large coherent apertures (e.g. 10 cm × 10 cm) for long-range LIDAR systems and similar applications.
Defense Advanced Research Projects Agency (DARPA) (HR0011-16-C-0106).
Fabrication of the devices utilized the facilities at the Marvell Nanofabrication Laboratory (Berkeley, CA) and at the Center for Nano-Micro Manufacturing (Davis, CA). We acknowledge helpful discussions from Profs. John Bowers, Larry Coldren and Jonathan Klamkin from the University of California, Santa Barbara. The views, opinions and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.
1. W. L. Stutzman and G. A. Thiele, in Antenna Theory and Design, 3 ed. (Wiley, 2013).
2. J. C. Hulme, J. K. Doylend, M. J. R. Heck, J. D. Peters, M. L. Davenport, J. T. Bovington, L. A. Coldren, and J. E. Bowers, “Fully integrated hybrid silicon two dimensional beam scanner,” Opt. Express 23(5), 5861–5874 (2015). [CrossRef] [PubMed]
3. C. V. Poulton, A. Yaacobi, D. B. Cole, M. J. Byrd, M. Raval, D. Vermeulen, and M. R. Watts, “Coherent solid-state LIDAR with silicon photonic optical phased arrays,” Opt. Lett. 42(20), 4091–4094 (2017). [CrossRef] [PubMed]
5. K. Van Acoleyen, W. Bogaerts, J. Jágerská, N. Le Thomas, R. Houdré, and R. Baets, “Off-chip beam steering with a one-dimensional optical phased array on silicon-on-insulator,” Opt. Lett. 34(9), 1477–1479 (2009). [CrossRef] [PubMed]
6. S. Chung, H. Abediasl, and H. Hashemi, “A monolithically integrated large-scale optical phased array in silicon-on-insulator CMOS,” IEEE J. Solid-State Circuits 53(1), 275–296 (2018). [CrossRef]
7. C. V. Poulton, M. J. Byrd, E. Timurdogan, P. Russo, D. Vermeulen, and M. R. Watts, “Optical Phased Arrays for Integrated Beam Steering,” in 2018 IEEE 15th International Conference on Group IV Photonics (GFP), 2018), 1–2. [CrossRef]
8. S. A. Miller, C. T. Phare, Y.-C. Chang, X. Ji, O. Jimenez, A. Mohanty, S. Roberts, M. C. Shin, B. Stern, zadka, and M. Lipson, “512-element actively steered silicon phased array for low-power LIDAR,” in Conference on Lasers and Electro-Optics, OSA Technical Digest (online) (Optical Society of America, 2018), JTh5C.2. [CrossRef]
9. C. T. Phare, M. C. Shin, J. Sharma, S. Ahasan, H. Krishnaswamy, and M. Lipson, “Silicon optical phased array with grating lobe-free beam formation over 180 degree field of view,” in Conference on Lasers and Electro-Optics, OSA Technical Digest (online) (Optical Society of America, 2018), SM3I.2. [CrossRef]
10. H. W. Chen, Y. H. Kuo, and J. E. Bowers, “A hybrid silicon-AlGaInAs phase modulator,” IEEE Photonics Technol. Lett. 20(23), 1920–1922 (2008). [CrossRef]
11. D. N. Hutchison, J. Sun, J. K. Doylend, R. Kumar, J. Heck, W. Kim, C. T. Phare, A. Feshali, and H. Rong, “High-resolution aliasing-free optical beam steering,” Optica 3(8), 887–890 (2016). [CrossRef]
12. W. Song, R. Gatdula, S. Abbaslou, M. Lu, A. Stein, W. Y. C. Lai, J. Provine, R. F. W. Pease, D. N. Christodoulides, and W. Jiang, “High-density waveguide superlattices with low crosstalk,” Nat. Commun. 6(1), 7027 (2015). [CrossRef] [PubMed]
13. M. A. Tran, T. Komljenovic, J. C. Hulme, M. J. Kennedy, D. J. Blumenthal, and J. E. Bowers, “Integrated optical driver for interferometric optical gyroscopes,” Opt. Express 25(4), 3826–3840 (2017). [CrossRef] [PubMed]
14. R. C. Jaeger, Introduction to microelectronic fabrication (Addison-Wesley Longman Publishing Co., Inc., 1987), p. 232.
16. K. Shang, C. Qin, Y. Zhang, G. Liu, X. Xiao, S. Feng, and S. J. B. Yoo, “Uniform emission, constant wavevector silicon grating surface emitter for beam steering with ultra-sharp instantaneous field-of-view,” Opt. Express 25(17), 19655–19661 (2017). [CrossRef] [PubMed]
17. E. Timurdogan, Z. Su, C. V. Poulton, M. J. Byrd, S. Xin, R.-J. Shiue, B. R. Moss, E. S. Hosseini, and M. R. Watts, “AIM process design kit (AIMPDKv2.0): silicon photonics passive and active component libraries on a 300mm wafer,” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2018), M3F.1. [CrossRef]
18. K. Shang, S. Pathak, B. Guan, G. Liu, and S. J. B. Yoo, “Low-loss compact multilayer silicon nitride platform for 3D photonic integrated circuits,” Opt. Express 23(16), 21334–21342 (2015). [CrossRef] [PubMed]
19. W. Xie, T. Komljenovic, J. Huang, and J. Bowers, “Dense III-V/Si Phase Shifters for Optical Phased Arrays,” in 2018 Asia Communications and Photonics Conference (ACP), 2018), 1–3.
21. T. Horikawa, D. Shimura, H. Okayama, S. Jeong, H. Takahashi, J. Ushida, Y. Sobu, A. Shiina, M. Tokushima, K. Kinoshita, and T. Mogami, “A 300-mm silicon photonics platform for large-scale device integration,” IEEE J. Sel. Top. Quantum Electron. 24(4), 1–15 (2018). [CrossRef]
22. Y. Zhang, Y. C. Ling, Y. Zhang, K. Shang, and S. J. B. Yoo, “High-density wafer-scale 3-D silicon-photonic integrated circuits,” IEEE J. Sel. Top. Quantum Electron. 24, 1–10 (2018).
23. S. J. B. Yoo, B. Guan, and R. P. Scott, “Heterogeneous 2D/3D photonic integrated microsystems,” Microsyst. Nanoeng. 2(1), 16030 (2016). [CrossRef]
24. R. Adams, M. Spasojevic, M. Chagnon, M. Malekiha, J. Li, D. V. Plant, and L. R. Chen, “Wavelength conversion of 28 GBaud 16-QAM signals based on four-wave mixing in a silicon nanowire,” Opt. Express 22(4), 4083–4090 (2014). [CrossRef] [PubMed]