Search

Silicon nitride (SiN) formed via low pressure chemical vapor deposition (LPCVD) is an ideal material platform for on-chip nonlinear photonics owing to its low propagation loss and competitive nonlinear index. Despite this, LPCVD SiN is restricted in its scalability due to the film stress when high thicknesses, required for nonlinear dispersion engineering, are deposited. This stress in turn leads to film cracking and makes integrating such films in silicon foundries challenging. To overcome this limitation, we propose a bilayer waveguide scheme comprised of a thin LPCVD SiN layer underneath a low-stress and low-index PECVD SiN layer. We show group velocity dispersion tuning at 1550nm without concern for filmcracking while enabling low loss resonators with intrinsic quality factors above 1 million. Finally, we demonstrate a locked, normal dispersion Kerr frequency comb with our bilayer waveguide resonators spanning 120nm in the c-band with an on-chip pump power of 350mW.

The fabrication processes of silicon nitride photonic devices used in foundries require low temperature deposition, which typically leads to high propagation losses. Here, we show that propagation loss as low as 0.42 dB/cm can be achieved using foundry compatible processes by solely reducing waveguide surface roughness. By post-processing the fabricated devices using rapid thermal anneal (RTA) and furnace anneal, we achieve propagation losses down to 0.28 dB/cm and 0.06 dB/cm, respectively. These low losses are comparable to the conventional devices using high temperature, high-stress low-pressure chemical vapor deposition (LPCVD) films. We also tune the dispersion of the devices, and proved that these devices can be used for linear and nonlinear applications. Low threshold parametric oscillation, broadband frequency combs and narrow-linewidth laser are demonstrated. Our work demonstrates the feasibility of scalable photonic systems based on foundries.

Compact, high-speed electro-optic phase modulators play a vital role in various large-scale applications including phased arrays, quantum and neural networks, and optical communication links. Conventional phase modulators suffer from a fundamental tradeoff between device length and optical loss that limits their scaling capabilities. High-finesse ring resonators have been traditionally used as compact intensity modulators, but their use for phase modulation have been limited due to the high insertion loss associated with the phase change. Here, we show that high-finesse resonators can achieve a strong phase change with low insertion loss by simultaneous modulation of the real and imaginary parts of the refractive index, to the same extent i.e. $\frac{\Delta n}{\Delta k} \sim 1$. To implement this strategy, we utilize a hybrid platform that combines a low-loss SiN ring resonator with electro-absorptive graphene (Gr) and electro-refractive WSe$_2$. We achieve a phase modulation efficiency ($V_{\frac{\pi}{2}} \cdot L_{\frac{\pi}{2}}$) of 0.045 V $\cdot$ cm with an insertion loss (IL$_{\frac{\pi}{2}}$) of 4.7 dB for a phase change of $\frac{\pi}{2}$ radians, in a 25 $\mu$m long Gr-Al$_2$O$_3$-WSe$_2$ capacitor embedded on a SiN ring of 50 $\mu$m radius. We find that our Gr-Al$_2$O$_3$-WSe$_2$ capacitor can support an electro-optic bandwidth of 14.9 $\pm$ 0.1 GHz. We further show that the $V_{\frac{\pi}{2}} \cdot L_{\frac{\pi}{2}}$ of our SiN-2D platform is at least an order of magnitude lower than that of electro-optic phase modulators based on silicon, III-V on silicon, graphene on silicon and lithium niobate. This SiN-2D hybrid platform provides the impetus to design compact and high-speed reconfigurable circuits with graphene and transition metal dichalcogenide (TMD) monolayers that can enable large-scale photonic systems.

The lack of high power integrated lasers have been limiting silicon photonics. Despite much progress made in chip-scale laser integration, power remains below the level required for key applications. The main inhibiting factor for high power is the low energy efficiency at high pumping currents, dictated by the small size of the active device. Here we break this power limitation by demonstrating a platform that relies on the coupling of a broad-area multimode gain to a silicon-nitride feedback chip, which acts as an external cavity. The feedback provided by the silicon-nitride chip is routed through a ring resonator and a single-mode filter, thus causing the otherwise multimode-gain chip to concentrate its power into a single highly-coherent mode. Our device produces more than 150 mW of power and 400 kHz linewidth. We achieve these high-performance metrics while maintaining a small footprint of 3 mm^2.

Widely-tunable and narrow-linewidth integrated lasers across all visible wavelengths are necessary to enable on-chip technologies such as quantum photonics, optical trapping, and biophotonics. However, such lasers have not been realized due to the strong wavelength sensitivity and high loss of integrated optical devices at short wavelengths. Here we overcome these challenges and demonstrate a chip-scale visible lasers platform by using tightly-confined, micrometer-scale silicon nitride resonators and commercial Fabry-Perot laser diodes. We achieve tunable and narrow-linewidth lasing in the deep blue (450 nm), blue (488 nm), green (520 nm), red (660 nm) and near-IR (785 nm) with coarse tuning up to 12 nm, fine tuning up to 33.9 GHz, linewidth down to sub-kHz, side-mode suppression ratio > 35 dB, and fiber-coupled power up to 10 mW. These specifications of our chip-scale lasers are comparable to those of state-of-the-art commercial laser systems, making them stand out as powerful tools for the next generation of visible-light technologies.

Portable mid-infrared (mid-IR) spectroscopy and sensing applications require widely tunable, narrow linewidth, chip-scale, single-mode sources without sacrificing significant output power. However, no such lasers have been demonstrated beyond 3 $\mu$m due to the challenge of building tunable, high quality-factor (Q) on-chip cavities. We demonstrate a tunable, single-mode mid-IR laser at 3.4 $\mu$m using a high-Q silicon microring cavity with integrated heaters and a multi-mode Interband Cascade Laser (ICL). We show that the multiple longitudinal modes of an ICL collapse into a single frequency via self-injection locking with an output power of 0.4 mW and achieve an oxide-clad high confinement waveguide microresonator with a loaded Q of $2.8\times 10^5$. Using integrated microheaters, our laser exhibits a wide tuning range of 54 nm at 3.4 $\mu$m with 3 dB output power variation. We further measure an upper-bound effective linewidth of 9.1 MHz from the locked laser using a scanning Fabry-Perot interferometer. Our design of a single-mode laser based on a tunable high-Q microresonator can be expanded to quantum-cascade lasers at higher wavelengths and lead to the development of compact, portable, high-performance mid-IR sensors for spectroscopic and sensing applications.

We investigate intermodal forward Brillouin scattering in a solid-core PCF, demonstrating efficient power conversion between the HE11 and HE21 modes, with a maximum gain coefficient of 21.4/W/km. By exploring mechanical modes of different symmetries, we observe both polarization-dependent and polarization-independent intermodal Brillouin interaction. Finally, we discuss the role of squeeze film air damping and leakage mechanisms, ultimately critical to the engineering of PCF structures with enhanced interaction between high order optical modes through flexural mechanical modes., Comment: 8 pages, 5 figures; supplementary included with 9 pages, 8 figs

To date, no electro-optic platform enables devices with high bandwidth, small footprint, and low power consumption, while also enabling mass production. Here we demonstrate high-yield fabrication of high-speed graphene electro-absorption modulators using CVD-grown graphene. We minimize variation in device performance from graphene inhomogeneity over large area by engineering graphene-mode overlap and device capacitance to ensure high extinction ratio. We fabricate an 8 mm x 1 mm chip with 32 graphene electro-absorption modulators and measure 94% yield with bit error rate below the hard-decision forward error correction limit at 7 Gbits/s, amounting to a total aggregated data rate of 210 Gbits/s. Monte Carlo simulations show that data rates > 0.6 Tbits/s are within reach by further optimizing device cross-section, paving the way for graphene-based ultra-high data rate applications.

Electrophysiological time series were taken simultaneously in two locations in the hippocampus of a rat brain previously described as receiving innervation from the osmosensitive vasopressinergic neurons of the hypothalamus. A hyperosmotic saline solution injection was administered during the time of the experiment. We analyze the recorded time series using different methods. We detect a modification of the delta and theta oscillations just after the perturbation caused by the injection. We compare the quality and information that each one of the methods exhibit and we analyze the characteristics of the perturbation based on a hypothesis that the strength of the functional connections between the vasopressinergic hypothalamic magnocellular neurons and their target in the hippocampus is modified by the perturbation. We built a model of the hypothetic neural connections and numerically calculate the time series produced by the system when simulating the perturbation caused by the saline injection. The theoretical results resemble the experimental findings concerning the frequency and amplitude alterations of the delta and theta bands., Comment: 11 pages, 12 figures

The Orphan Stream is one of the most prominent tidal streams in the Galactic halo. Using data on red giants, RR Lyrae, and horizontal branch stars from Gaia and other surveys, we determine the proper motion of the Orphan Stream over a path of more than 90 degrees on the sky. We also provide updated tracks for the sky position, distance, and radial velocity of the stream. Our tracks in these latter dimensions mostly agree with previous results. However, there are significant corrections to the earlier distance and latitude tracks as the stream approaches the Galactic disk. Stream stars selected with three-dimensional kinematics display a very tight red giant sequence. Concordantly, we find that applying a proper motion cut removes the most metal-rich stars from earlier spectroscopic samples of stream stars, though a significant dispersion remains indicating a dwarf galaxy origin. The deceleration of the stream towards its leading end suggests a circular velocity of ~200 km/s at a galactocentric radius ~30 kpc, consistent with other independent evidence. However, the track of the stream departs significantly from an orbit; the spatial track does not point along the same direction as the velocity vector, and it exhibits a lateral wiggle that is unlikely to match any reasonable orbit. The low metallicity and small dispersion of the stream in the various coordinates point to a progenitor with a relatively low dynamical mass ~$10^8$ Msun., Comment: 14 pages, 10 figures, 1 table; submitted to MNRAS

The 3D velocities of M31 and M33 are important for understanding the evolution and cosmological context of the Local Group. Their most massive stars are detected by Gaia, and we use Data Release 2 (DR2) to determine the galaxy proper motions (PMs). We select galaxy members based on, e.g., parallax, PM, color-magnitude-diagram location, and local stellar density. The PM rotation of both galaxies is confidently detected, consistent with the known line-of-sight rotation curves: $V_{\rm rot} = -206\pm86$ km s$^{-1}$ (counter-clockwise) for M31, and $V_{\rm rot} = 80\pm52$ km s$^{-1}$ (clockwise) for M33. We measure the center-of-mass PM of each galaxy relative to surrounding background quasars in DR2. This yields that $({\mu}_{\alpha*},{\mu}_{\delta})$ equals $(65 \pm 18 , -57 \pm 15)$ $\mu$as yr$^{-1}$ for M31, and $(31 \pm 19 , -29 \pm 16)$ $\mu$as yr$^{-1}$ for M33. In addition to the listed random errors, each component has an additional residual systematic error of 16 $\mu$as yr$^{-1}$. These results are consistent at 0.8$\sigma$ and 1.0$\sigma$ with the (2 and 3 times higher-accuracy) measurements already available from Hubble Space Telescope (HST) optical imaging and VLBA water maser observations, respectively. This lends confidence that all these measurements are robust. The new results imply that the M31 orbit towards the Milky Way is somewhat less radial than previously inferred, $V_{\rm tan, DR2+HST} = 57^{+35}_{-31}$ km s$^{-1}$, and strengthen arguments that M33 may be on its first infall into M31. The results highlight the future potential of Gaia for PM studies beyond the Milky Way satellite system., Comment: 15 pages, 7 figures, ApJ, in press

Carrier recombination dynamics in strip silicon nano-waveguides is analyzed through time-resolved pump-and-probe experiments, revealing a complex recombination dynamics at densities ranging from ${10^{14}}$ to ${10^{17}}\,$cm$^{{-3}}$. Our results show that the carrier lifetime varies as recombination evolves, with faster decay rates at the initial stages (with lifetime of ${\sim 800}\,$ps), and much slower lifetimes at later stages (up to ${\sim 300}\,$ns). We also observe experimentally the effect of trapping, manifesting as a decay curve highly dependent on the initial carrier density. We further demonstrate that operating at high carrier density can lead to faster recombination rates. Finally, we present a theoretical discussion based on trap-assisted recombination statistics applied to nano-waveguides. Our results can impact the dynamics of several nonlinear nanophotonic devices in which free-carriers play a critical role, and open further opportunities to enhance the performance of all-optical silicon-based devices based on carrier recombination engineering.