|dc.description.abstract||Aluminum based thermites are a broad class of highly tunable solid energetic materials that can be formulated to perform as pyrotechnics, propellants, or explosives. Their functional versatility, energy density, reactivity in inert environments, and compatibility with silicon-based fabrication techniques has garnered lots of interest from researchers in the past 20 years for microelectromechanical systems (MEMS) with industrial, aerospace, and life-science applications. Thermites have traditionally come in the form of physically mixed powders of discrete metallic fuel and metal-oxide oxidizer particles which require intimate contact to ignite, but new fabrication techniques can produce monolithic composites of fuel and oxidizer phases; examples of these include laminate and core-shell structures. Since the fuel and oxidizer phases in a core-shell thermite exist within a single particle, individual particles can theoretically combust. Thus, a core-shell dispersion can theoretically remain reactive even in an inert fluid. As a result, there is growing interest in the use of core-shell thermites as energy carriers in liquids or gasses so they can be easily transported and ignited on-demand.
This thesis focusses on the design, characterization, and evaluation of a microfluidic system which uses electrokinetic techniques to transport dispersed core-shell thermite particles and ignite them in an inert, aqueous environment. The system consists of a single microchannel with a converging-diverging geometry and 30µm x 30µm throat for the collection of dispersed 1μm Al@CuO particles dispersed in 1[M] NaCl solution. Using electrodes at the inlet and outlet, a simple non-intrusive ignition system is incorporated by means of electrical discharge produced by capillary breakdown of the liquid at the throat.
The microchannel geometry is numerically modelled to determine the effects of applied voltage and solution conductivity on the electrical field, velocity, and temperature within the channel. The electric field is maximized at the transitions of the converging and diverging sections with the throat, producing magnitudes (6.4kV/cm for 500V across channel) which predict electrical breakdown for solutions with NaCl concentration exceeding 0.2 [mM]. Since phase change is not included in the model, temperature results can exceed the boiling point of water. It is also observed that the electroosmotic velocity prediction begins to fail once temperatures reach 100°C; this is deemed a limitation of the practical usefulness of the model.
A single particle suspended in the electrolytic solution is also modelled. The presence of a particle in the liquid increases the maximum electric field by 3 to 10 times. Because a large spread of conductivity was considered, several heating modes are predicted by the model: heating at the liquid-shell interface for solutions less conductive than CuO, heating within the shell for solutions more conductive than CuO, and heating in the liquid and the shell when the solution is significantly more conductive than CuO. These results suggest that a higher solution concentration results in increased heating in both the electrolyte and particle phases, so experimentally, a high conductivity 1 [M] NaCl solution was used.
The microchannel was experimentally characterized using IR imaging and high-speed microscopy techniques. When operated in slow-heating mode (2.5mA through channel), the liquid in the channel undergoes a long Joule heating stage and eventually results in the formation of a bubble at the throat. Once the bubble fills the cross-section of the channel current flow is halted due to the formation of an open circuit and the bubble cools and shrinks. If the bubble shrinks enough, a conductive path may be re-established around the bubble and Joule heating may begin again. Maximum measured temperature does not exceed 75°C when only the solution is considered. When 1μm Al@CuO particles are added to the channel in slow-heating mode, the Joule heating period is substantially shortened. Additionally, thermal hot spots are observed exceeding 140°C in the IR video and bubbles grow more rapidly when energetic particles are added.
In fast-heating mode (500V across channel), a Joule heating stage is not observed. Instead, capillary discharge occurs, where bubble formation is nearly instantaneous and closely followed by visible electrical discharge once power is applied. When 1μm Al@CuO particles are added to the channel in high-power/fast-heating mode, the difference is difficult to observe because the presence of particle combustion is masked by the bright electrical discharge.
A narrow-band 488nm filter is placed inline with the high-speed camera to distinguish between electrical and combustion emission sources. When the microchannel is operated without particles, no emission is visible in the filtered video, but when particles are introduced, bright spots are visible in the filtered video which correlate to hotspots observed in the IR video. This suggests that increased temperatures in particle tests are a result of thermite combustion. Morphological and chemical analysis is performed using SEM and EDS and both unreacted and reacted particles are observed in the channel post-test, further confirming presence of thermite combustion.
Finally, attempts to modulate the combustion size are made by purposeful collection of large particle aggregations at the channel throat. The capillary ignition technique used is found to be generally incompatible with this method because the bubbles formed by Joule heating destroy the particle clusters. A large-scale combustion event is observed when particles were clustered at the throat generating a flame which fills the entire channel and propagates at 1.9m/s, demonstrating that flame-size is indeed tunable with dispersed core-shell particles, but the event could not be reproduced due to limitations with the ignition system.||en