We present a method for targeted and maskless fabrication of single silicon vacancy (V_Si) defect arrays in silicon carbide (SiC) using focused ion beam. Firstly, we studied the photoluminescence (PL) spectrum and optically detected magnetic resonance (ODMR) of the generated defect spin ensemble, confirming that the synthesized centers were in the desired defect state. Then we investigated the fluorescence properties of single V_Si defects and our measurements indicate the presence of a photostable single photon source. Finally, we find that the Si⁺⁺ ion to V_Si defect conversion yield increases as the implanted dose decreases. The reliable production of V_Si defects in silicon carbide could pave the way for its applications in quantum photonics and quantum information processing. The resolution of implanted V_Si defects is limited to a few tens of nanometers, defined by the diameter of the ion beam.

In this paper, a fine particle matter (PM) sensor has been designed, built up, and measured. It consists of a LED, a photodiode, a glass air channel and collimated structure, as show in Fig.1. Compared with the conventional fine PM sensors [1-2], this sensor has higher sensitivity and better reliability via immunity to particle contamination.

A simulation model of the particle matter sensor is built based on the right angle Mie-scattering theory [3]. The simulation results are shown in Fig 2. In the simulation, the wavelength of LED is 650 nm, and particles diameter is 1 μm. The particles are guided into a “U” shape glass channel with a good light transparence at visible light region. And the light is collimated by an embedded lens in LED packaging, and injected into the glass channel with the minimum scattering from the glass channel. The particles are isolated by the glass channel so that they cannot contaminate the LED and the photodiode, which is a common failure mode in conventional PM sensors. Based on the simulation and fitting process, the detection limitation is down to around 0.2 μg/m3 when the power of the light source is 0.4mW and specific detectivity of the photodiode is 0.017uW.

The sensor used in characterization is assembled as shown in Fig. 3. A TSI 9306 six-jet atomizer is used to generate fine particles (diameter ~ 1 m) carried by dry-air gas flow. A reference PM detector (model: TSI 8520 DustTrak Aerosol Monitor) is connected at the outlet of the sensor to monitor the particle mass concentration in real-time. The glass channel with gold coating is used to guide the particle air. The experimental results are shown in Fig. 4. The blue dots and red line are the testing and fitting results, respectively. Through analysis, the relationship between particle mass concentration and PD voltage can match to linear trend. The output signal is up to approximate 1.5 μV when the PM concentration changes 1μg/m3, and the detection limitation is around 3.3ug/m3, which is much better than the value of 10μg/m3 of the conventional PM sensors.

In recent years, Er silicates and Yttrium (Y) and Ytterbium (Yb) co-doped Er silicates have attracted intensive investigations due to their efficient luminescence at 1.53μm. Compared with traditional Er-doped materials, the Er concentrations of silicates are about 1 to 2 orders of magnitude higher due to the stoichiometric nature of these Er compounds. However, it turned out that its large Er silicate waveguide transmission loss and high pumping power required inhibited the possibility of achieving high gain with device size scaled down. In order to solve the problems mentioned above, we use single crystal Er-Yb/Y silicate compound nanowires as the waveguide material to reduce the transmission loss due to fewer defects in single crystal nanowire. In addition, the quantum confinement effect existed in single crystal nanowire can improve the emission lifetime of the erbium, so that it will be much easier to achieve the erbium ion population inversion. In this talk, we will introduce our group recent work about Er/Yb/Y silicates nanowire waveguide materials. Firstly, we fabricated the Er silicate compound nanowire by a chemical vapor deposition method. The strong 1.53μm gain characteristics of ErxYb(Y)2-xSiO5 nanowire have been obtained. This property of Er silicate nanowires indicate that it is a promising material for achieving high gain nanowire optical waveguide amplifier and laser.

In this talk, I will introduce the recent advances of integrated microwave photonics in our group, including arbitrary waveform generator, microwave photonic filter, and energy efficient microheater in a chip. By incorporating a graphene on a slow-light silicon photonic crystal waveguide, we experimentally demonstrate an energy-efficient graphene microheater with a tuning efficiency of 1.07 nm/mW and power consumption per free spectral range (FSR) of 3.99 mW. The rise and decay times (10% to 90%) are only 750 ns and 525 ns, which, to the best of our knowledge, are the fastest reported response times for microheaters in silicon photonics.

Optical microcavities have attracted considerable attention and played an important role in on-chip classic as well as quantum information processing during the past decades. High quality optical microcavities with Q factor >10⁷ are mostly made from silica using techniques such as reflow, sophisticated etching to create wedged facet [1], etc. The transparent window for silica, however, is limited in the visible and near-infrared region. There are enormous vital applications in the mid-infrared region, where optical microcavities can be leveraged. Silicon is a transparent optical medium from near-infrared up-to 8mm in the mid-infrared region, and it has a very large nonlinear coefficient (e.g., its third-order nonlinearity susceptibility is 10³ times higher than that of silica) [2]. Therefore silicon is an excellent candidate for fabricating optical microcavities in this wavelength region. In order to improve the quality factor of silicon microcavities, much effort has been devoted to reduce the scattering loss caused by surface roughness, which could include dry oxidation, wet chemical etching and hydrogenation [3-5].

Here, we propose to fabricate silicon spherical microcavities by laser heating inverted silicon cones under atmospheric pressure and at room temperature. Different etching processes are used in experiment to produce inverted cones on silicon-on-insulator (SOI) substrate, and then a careful direct laser heating step with proper influence reshapes the cones to spherical microcavites as a result of surface tension. The exact shape and morphology of the prepared silicon spherical microcavities are characterized using SEM and AFM.

A flow chart showing our fabrication procedures is illustrated in Figure 1. An AZ5214 resist layer is first used for the lithography process on a SOI wafer with a 16um silicon device layer and a 2um buried oxide layer. The etching processes are carried out by inductively coupled plasma (ICP) with two steps each with unique etching parameters, first to define silicon pillars and the second to under etch the pillar to form inverted cones. Figure 2(a) shows a typical etched inverted silicon cone. By using an in-house imaging and laser heating setup, the invert silicon cones are located on the chip and illuminated by a focused green laser beam with a controlled exposure. According to the front and top view shown in Figure 2(b) and (c), the silicon spherical microcavity has a diameter about 16mm with slight imperfection near the top. Such imperfection is consistent from sample to sample. It is thus speculated to be related with the crystallization of the silicon during the rapid cooling down period when the laser exposure is off. Nevertheless, the top view of such structure suggests that it has a well-defined round-shaped equator. AFM is used to examine the surface roughness of our fabricated silicon microcavities. The surface tomography of a typical sample is shown in Figure 3. It shows that the root-mean-square (RMS) roughness is about 0.6 nm, a very encouraging number, which has been to shown to support very low loss propagation in a waveguide configuration. It is very likely that this kind of silicon micocavites can support high Q whisper-gallery modes confined to near the equator, away from the imperfection seen in the SEM picture. Optical testing of these microcavities are currently being actively conducted. We believe that our approach for producing low surface roughness silicon microcavities will have a major impact in CMOS integrated photonics in future.

We experimentally demonstrate femtosecond cylindrical vector beams (CVBs) and high-order optical vortex beams (OVBs) based on a mode-locked fiber laser (MLFL) by using a fused few-mode coupler. The fused SMF-FMF coupler inserted in the cavity not only acts as mode converter from fundamental mode to high-order fiber modes with a broadband width, but also directly delivers femtosecond vortex pulses out of the mode locked cavity. Mode coupling is analyzed in the coupling region between the fundamental mode in SMF and a high-order mode in FMF. Linearly polarized vortex modes can be obtained by combining different vector modes. Such ultrafast vortex beams are expected to fabricate a chiral nano-structure originated by angular momentum transfer of the optical vortex to a material. A shorter pulse of the MLFL is supposed to be implemented by optimizing the dispersion of the laser cavity.

This paper reports a high-speed packaging for Silicon Mach-Zehnder modulators (MZMs). The radio frequency (RF) input port of the package is commercially available GPPO type. High-frequency ceramic substrate is designed to shorten the length of gold wire to suppress the RF attenuation. Gratings and single-mode fibers using inclined-surface coupling are used to realize low-loss and large tolerance coupling. In measurement, clear eye diagrams with ER of 10.1 dB and 7.6 dB were obtained at 25 Gb/s and 40 Gb/s modulations.

Silicon photonics has become a hot research topic in the field of optoelectronics in recent years, since silicon–based optical device can achieve high density integration, low power consumption, low cost photonic devices, realize monolithic integration with electronic devices, and it is compatible with traditional CMOS processes [1]. Silicon electro-optic modulators which are key devices in optoelectronic information transmission and processing have got great developments in recent years. High speed OOK silicon modulators have been demonstrated [2-4]. The silicon modulator packaging has also been studied [5-6]. However, the operating speed of packaged silicon modulators still need to be improved.

Our butterfly package for silicon modulator is illustrated in Figure 1. Two main designs were applied for high speed packaging: the modulator chip was designed with reverse PN mode, whose PN junction capacitance is smaller and the modulator electrode was carefully designed using CPW travelling-wave electrodes; GCPW transmission lines with metallic holes were adopted to fabricate the high-frequency ceramic substrates to make the high-speed electrical signal effective loading on the electrode of the MZM chip. Since micro-strip-like line (MSL) mode in GCPW line might exist, which usually cause energy coupling of CPW mode and prompt a microwave resonant of transmission response in high frequency range [7], we made metallic holes on a transmission line to suppress the microwave resonant [8-9].

Grating coupler and single-mode fiber with inclined-surface are used in optical coupling. The advantages of this optical coupling method is larger misalignment tolerance and cheap cleaved fiber packaging. Nickel metal tube is installed close to the inclined-surface of single-mode fiber, the position and angle of the optical fiber is adjusted by special fixture clamping nickel metal tube and adjusting the special fixture. Assuming the optical fiber is parallel to the chip surface, there is a relation of θ=2*φ-90^o. We choose the φ=51 deg. as our grating incidence angle is 12 degree. 

Eye-diagram measurements were carried to demonstrate the high-speed operation of the packaged modules. Experimental results were presented in Fig. 4(a) and 4(b). Clear eye diagrams with ERs of 10.1 dB and 7.6 dB were obtained at 25 Gb/s and 40 Gb/s modulation speed. By improving the welding between the pin of GPPO connector and ceramic substrates, further reducing the length of gold wires from the ceramic substrates to the modulator pads, the speed of the packaged MZM module could be higher.

Silicon photonics for the mid-infrared (MIR) wavelength is an emerging domain and a promising solution to enable low-cost, high performance, compact sensing and detection capabilities and fully compatible with CMOS technologies. Silicon on insulator (SOI) has been the most dominant platform for telecom and Datacom domain (NIR). This platform has a major asset to extend their applications in the MIR wavelengths, due to the transparency of silicon and silica in this spectral range. In this paper, we present our recent results on optical passive devices for short MIR applications. The SOI platform is used and several devices are designed, fabricated and characterized at the wavelengths of 2μm. The passive structures like waveguides, MMI, ring resonators, (de)multiplexers and beam splitters were demonstrated with high performances.

Integrated optical light source on silicon is one of the key building blocks for optical interconnect technology. Great research efforts have been devoting worldwide to explore various approaches to integrate optical light source onto the silicon substrate. The achievements so far include the successful demonstration of III/V-on-Si hybrid lasers through III/V-gain material to silicon wafer bonding technology. However, for potential large-scale integration, leveraging on mature silicon complementary metal oxide semiconductor (CMOS) fabrication technology and infrastructure, more effective bonding scheme with high bonding yield is in great demand considering manufacturing needs. Here we will review IME’s efforts on the hybrid integration through high-throughput multiple dies-to-wafer (D2W) transfer bonding technology and its application for hybrid silicon laser demonstration. Such bonding technology is a key enabler towards the large-scale heterogeneous integration of optoelectronic integrated circuits (H-OEIC).

Inverse design has triggered great research interests in integrated photonic design. It can be  advantageous in discovering new photonic devices, because, unlike traditional device design (or forward design) where only a few parameters can be swept, a high-order multi-dimensional parameter space can be searched with an advance optimization algorithm for a global fitness that is often correlated with one or more performance metrics of the targeted device. Recently, such design concept has been applied to obtain ultra-compact photonic devices. They, however, suffer greatly from performance degradation due to large out-of-plane scattering from strong scattering processes. In this talk, we utilize the concept of inverse design and show that it is possible to design ultra-compact and highly performance silicon photonic devices that may have potential impact in future very large scale silicon photonic integrated circuits.