Moisture absorbed by the hygroscopic polymers like molding compound and die-attach vaporizes during reflow lead to a high vapor pressure inside the electrical components lead to failure in the electrical device, named as popcorn failure. Popcorn failure in plastic encapsulated microcircuits has been a critical issue for electronic device reliability. Researches have been conducted on investigating the failure mechanism. Among all the factors that contributed to the failure, vapor pressure is one of the primary sources of stress that causes crack of the molding compound and delamination between critical interfaces. Numerous publications demonstrate the vapor pressure evolution and contributing factors with mathematical models, simulations, and tests. However, direct measurement of vapor pressure is not yet reported. This thesis presents the design and fabrication verification of a pressure sensor to measure the vapor pressure evolution in moisture-containing polymers at reflow temperatures. The specifications and requirements are extracted from the failure mechanism reported in the literature. A touch-mode capacitive pressure sensor with in-situ doped poly-SiC is designed to measure vapor pressure from atmospheric pressure to 8 MPa under reflow temperature up to 300 °C. Simulation on touch mode capacitive readout is performed to verify the design parameters and provide an estimation of device performance. The fabrication process is designed and conducted with several methods for crucial steps along with different structure dimensions to investigate an optimal solution. A complete fabricated device is achieved with measured initial capacitance from 12.3 to 26.7 pF for different sizes of diaphragms. The deviation between the measured results and simulation results due to fabrication problems is analyzed. Possible causes and solutions of problems that occurred in fabrication and measurement, such as unexpected upward bending of the diaphragm, leakage current between the capacitor electrode plate, and uniformity of SiC layer fabrication, are discussed. Problem correction, device optimization, complete characterization, and experiments for vapor pressure measurement in molding compound remain as future work.

Silicon-based MEMS technology has been the standard for developing 2D and 3D
micro-structures for many years. There are 2 main classifications of MEMS manufacturing technologies, viz. bulk micro-machining, and surface micro-machining.
These techniques have its own set of drawbacks. Bulk micro-machining affects the structural integrity of the wafer because bulk silicon is being etched. Surface micro-machining is limited by the maximum thickness of the method of thin-film deposition (usually a few microns). The current silicon-based MEMS sensors can also fail when it comes to harsh environment sensing. Due to this, the sensors would require many supporting infrastructures such as radiation shield, cooling system and shock-proof packaging.
One way to circumvent this problem is by designing sensors using materials that would not require as much supporting infrastructure. Silicon carbide (SiC) has proved to be a viable candidate to be used in such harsh environments. This is
because it is found to be mechanically robust, chemically inert and with good wear resistance. However, bulk micro-machining with SiC is extremely challenging due to the high difficulty in etching. Many techniques have been tested to etch SiC, but they all have considerable drawbacks or low etch rates. Thus, to make high aspect ratio structures using SiC we will require a new technique where carbon nanotubes (CNT) can be used as a framework in the fabrication process for high aspect ratio surface micro-machining. CNTs can be grown to lengths of several micrometres to millimetres while their diameter is in the order of a few nanometres. Bundles of these tubes were found to have excellent conductance with high current densities, and their behaviour can be either metallic or semiconducting depending on their chiral vector. These unique properties make CNTs a very interesting material to be integrated into conventional MEMS technology. Although a single nanotube has excellent properties, bundles of nanotubes show a ‘foam-like’ property since they are held together by weak van der Waals’ forces. Thus, to realize mechanical structures with CNTs, we will have to coat the nanotube bundles with a filler material. This has previously been demonstrated to allow tuning of the mechanical properties of the composite. SiC would be an attractive filler candidate for harsh environment sensors. Due to the porous nature of nanotubes, it is possible to infiltrate the forest by deposition of a nanoscale coating. The deposition of the filler material is done by means of low pressure chemical vapor deposition (LPCVD) since a low pressure and deposition rate will enable the nanotube forest to get completely infiltrated and more uniformly coated. The goal of this research is to fabricate the first sensor using this technique, viz. a comb type capacitive accelerometer and test its performance and resilience to harsh environments by using SiC as coating. This technique will enable the user of a thicker layer for the proof mass and combs, resulting in a higher performance and potentially, resistance to harsh environments.