The broader impacts/commercial potential of this Small Business Innovation Research (SBIR) Phase I project is that it will enable a dramatic leap in the technology of electrooptic modulators. Electrooptic modulators have widespread applications in fiber-optic telecommunications. The modulators are also key components for optically-steered phased-arrayed antennas, electromagnetic-attack-resistant and radio-over-fiber radio frequency photonic links, light detection and ranging, optical sensing, metrology, storage, sampling and communications, all-optical signal processing. The most important future application is in optical interconnects for supercomputers and data centers. The high-performance devices developed under this program will eventually be found in data communication to transmit data between racks of data centers and supercomputer facilities and between microprocessors, graphic and memory chips.
This Small Business Innovation Research (SBIR) Phase I project aims at developing high-performance optical modulators using a novel lithium niobate (LiNbO3) on silicon waveguide technology. LiNbO3 has been long regarded as the most attractive material for electrooptic modulation for high-performance optical communication systems. However, the weakly confined LiNbO3 waveguides formed by diffusion or implantation of dopants do not lend themselves to high-level chip integration. The goal of this proposal is to merge two complementary photonic technologies (i.e., silicon and LiNbO3 photonics) and make a hybrid platform that offers the advantages of both technologies, while it avoids their respective disadvantages. Based on the technology, LiNbO3-on-Si waveguides, microring resonators and electrooptic modulators (Mach-Zehnder interferometer and microresonator types) will be commercialized. The Mach-Zehnder devices are expected to operate at voltages several times less than commercial devices and several times more compact.
Data-center communication is shifting to 100Gig and beyond. For these communication speeds, compact external modulators the can fit into small form factor optical transceivers are needed. Also coherent systems in telecommunication need compact optical modulators circuits that can be made in low cost. Partow technology have developed a novel integrated optical platform that can be used to make these modulators. In phase I, our target was to improve the performance of the devices to a level that can be used for these applications. For this purpose we targeted to achieve an optical modulator based on thin film lithium niobate that can achieve a modulation length product of 4v.cm and a bandwidth of 10GHz and a waveguide insertion loss of 1dB/cm. Ion implantation and room-temperature wafer bonding was used to transfer 400-nm-thick Y-cut LN onto thermal SiO2 cladding layers grown on Si substrates. The thin LN films were covered with index-matched Ge23Sb7S70 Chalcogenide (ChG) glass layers deposited by electron-beam evaporation and dry-etched to form the rib-loaded region to achieve single-mode waveguides. Metal electrodes were deposited and patterned to achieve modulators. Micro-ring resonators (MR), Mach-Zehnder modulators (MZM), and micro-ring modulators (MRM) were designed and fabricated. The devices were characterized in the telecom band. Grating couplers with apodized gratings followed by parabolic tapers were also demonstrated for the first time on LN-on-Si and used to couple light in and out of the MRs and MRMs. End-butt fiber coupling was used for the MZM chips. Propagation loss of 1.2 dB/cm and a quality factor, Q, of 1.35 × 105 at 1560 nm wavelength were measured from the transmitted power spectrum of under-coupled micro-ring resonators with 200 μm radius. The critically-coupled MRMs have a loaded Q of 8.6 × 104 at a 150μm radius, extinction ratios of 13 dB, and modulate at 0.4 GHz/V. The 6-mm long, push-pull configured MZMs, with 5.5 μm electrode gaps, have a Vπ.L of 3.6 V.cm, and 20 dB extinction ratio at 1550 nm at low modulation frequencies. Preliminary S21 measurements on a 7-µm electrode-gap MZM, indicate the potential for broadband modulation upon electrode optimization. The optical bandwidth is more than 8 GHz for the modulator and is limited by the available 8-GHz photodetector. A reduced electrode gap around 4 μm, is predicted to yield Vπ.L below 1.5 V.cm. Once this is complemented by our ongoing broadband simulations and designs, accounting for the frequency variations present in the RF response, the MZM can be optimized for broadband performance.
