This work investigates the fabrication of a linearly tunable capacitor using the standard CMOS (Complementary Metal-Oxide Semiconductor) process and a maskless post-process. This tunable capacitor is composed of a comb-drive actuator, a parallel capacitor and supported beams. The comb-drive actuator is employed to change the position of the movable electrode plate in the parallel capacitor, such that the overlap area between the two plates in the parallel capacitor is changed. The capacitance of the tunable capacitor is a linear variation. The benefits of the pos-process are compatible with the CMOS process and etching without mask. The post-process has two main steps. One is the use of phosphoric acid to remove the metal from sacrificial layers and etch holes. The other step employs RIE (reactive ion etching) to etch the passivation layer over the pads. The experimental results show a capacitance of 500 fF, and 50% tuning range at 20 V.
This work presents the micro-machined 2-D switch array for use on free-space optical interconnect platform integrated with a digital 1x8 de-multiplexer control circuit together. Moreover, this device employs electrostatic actuation for light beam directions control. The CMOS-MEMS array-based optical platform contains 10x10 circular micromirrors switching spots, the diameter of each mirror is about 50 μm and the overall chip size is around 2 mm by 2 mm. The commercialized simulation softwares were used to validate the micromirror design and elucidate the behavior of the micromirror before fabrication. The post-process simply employs HF based solution to etch silicon dioxide layer to release the suspended mirror structures. The micromirror array is actuated using an electrostatic force. The results reveal that the micromirror has a tilting angle of around 8° according to the triangular relation with a driving voltage of 18V at pull down state. Also described herein are the general principles of the light-beam switching method used, the detailed of device design, the post-CMOS fabrication process flow, the result of simulations and preliminary experimental results are discussed.
A capacitive micropressure sensor with readout circuits on a single chip is fabricated using commercial 0.35micrometers CMOS process technology and post-processing. The main break through feature of the chip is the positioning of its readout circuits under the pressure sensor, allowing the chip to be smaller. Post-processing included anisotropic dry etching and wet etching to remove the sacrificial layer, and the use of PECVD nitride to seal the etching holes of the pressure sensor. The sacrificial layer was the metal 3 layer of the standard 0.35 micrometers CMOS process. In addition, the readout circuit is divided into analog and digital parts, with the digital part being an alternate coupled RS flip- flop with four inverters that sharpened the output wave. Moreover, the analog part is employed switched capacitor methodology. The pressure sensor contained an 8 X 8 sensing cells array, and the total area of the pressure sensor chip is 2mmx2 mm. In addition to illustrating the design and fabrication of the capacitive pressure sensor, this investigation demonstrates the simulation and testing results of the readout circuit.
This investigation presenst a micromachined optical modulator with electrostatic actuation fabricated by the conventional CMOS process. The modulator is operated by interaction of fixed part, stationary gratings, and movable part, sliding gratings. The period of the gratings varies with the slide of the movable part, thereby allowing different diffraction patterns of the reflected light. In addition, 100% modulation in the first order can serve as an optical switch. All procedures following the CMOS process merely require a simple post-process. With maskless etching, the micromachined optical modulator is developed to obtain a high-aspect-ratio structure and high efficiency of modulation. Compare to the commercially available acoustic ones, the micromachined optical modulator proposed herein is smaller and weigh less.
This investigation presents a novel type of DOEs fabricated by the conventional CMOS process. A simple post-CMOS process is applied to form the relief pattern, which can be used directly for its optical properties or serve as a mold for the subsequent replication. By using the CMOS process, in addition to reducing the depth, alignment, dimension, and shape errors of the pattern, the scale is minimized by the advancing microfabrication as well. The performance of arbitrary DOEs can be directly related to the diffraction efficiency of the gratings. Therefore, in this investigation, the shape of the multi-level gratings is designed and the diffraction efficiency is calculated by the rigorous vector coupled-wave analysis. The largest constraint of the CMOS process for the multilevel gratings is that the depth of each layer is different and unchangeable. However, a suitable length for each level can be determined and, in doing so, the diffraction efficiency can reach 81%.
This investigation presents a concept of integrated device, which tracks the movement of eyeball in real time. Integrated optical components are applied to perform infrared oculography to find the position of the eyeball. First, the photodiodes emits infrared to human eyes via the output coupler, then the reflected light is collected by the input coupler and detected by the photodetectors. By analyzing the electrical signal, we could figure the position of the eyeball out. The basic principle is based on the differential of reflected index between sclera and iris. The light source and detectors are settled on the side of goggles worn and, in doing so, the eyesight could be wide without obstacles. Therefore, the optical guiding interconnection is an essential issue for the eyeball- tracking device. Since the fine line gratings are necessary for the optical coupler, the electron beam writing is used to meet the requirement. The gratings with 40 nm line width are achieved, and it is sufficient for optical coupling. On the other hand, the electrical signal related eyeball position is simulated. We could translate the electrical signal readout into the coordinate of the eyeball position. The whole eyeball-tracking device would be small volume, less weight, and portable.