Integrated microfluidic systems enable manipulation of fluids at the submillimeter length-scale, which offers particular advantages for various chemical, biological and biomedical testing applications. Conventional technologies based upon micromachining approaches are inherently planar in nature, therefore alternative approaches for the construction of microfluidic systems via additive manufacturing or three-dimensional (3D) printing gave drawn great interest in the field of microfluidic device engineering. This dissertation presents the Multijet 3D printing approach - a layer-by-layer, multi-material micro-scale deposition method - towards the development of geo- metrically and functionally-unique submillimeter microfluidic structures, resulting in new classes of three dimensionally-complex integrated microfluidic platforms to address challenges currently faced by conventional microfluidics in relevant biomedical and diagnostic applications.
Homogenous mixing of co-laminar fluids inside conventional microchannels, especially under low Reynolds Number conditions (Re < 1) continues to be a primary issue for the efficient on-chip mixing of reagent and analytes. In the first part of this dissertation, various 3D μ-mixer prototypes capable of inducing fluidic motion and chaotic advection in three-dimensions to promote mixing quality enhancement over a range of Re’s (0.1 < Re < 10) have been engineered. Theoretical simulations and experimental characterization demonstrate that greater mixing per-unit microchannel length is achieved utilizing intra-channel fabricated 3D μ-mixers (~148% increase in mixing quality is achieved at Re = 0.1 using a 4mm long 3D μ-mixer) as compared to smooth-walled 3D printed microchannels.
Microfluidic concentration gradient generators (μ-CGG) are incapable of generating symmetric gradients of more than two fluids simultaneously. In the second part of this dissertation, two different 3D microchannel network designs which accomplish 3D fluidic routing impossible to achieve using planar fabrication methods in order to generate symmetric three-fluid gradients are demonstrated. Fabricated prototypes output discrete μ-drug cocktails containing certain potentially experimentally-useful concentrations of each input fluidic species (~100%, 76%, 50%, 32%, 11%, 10% & 0% and ~100%, 70%, 50%, 33%, 30% & 0%, respectively). Furthermore, incorporating 3D μ-mixer structures to improve the simulated accuracy of the generated gradient up to ~88% is demonstrated. The biomedical application of both fabricated 3D μ-CGG prototypes is demonstrated through single, pair-wise and three-antibiotic susceptibility testing experiments involving multiple clinically-relevant antibiotic drugs against antibiotic-resistant Escherichia coli bacteria.
In the third part of this dissertation, entirely-3D printed modular human-powered microfluidic actuators are proposed to serve as easy-to-operate, portable and completely electrical power-free sources of fluidic actuation. A new 3D fluidic one-way valve concept employing a dynamic bracing mechanism is presented, demonstrating enhancement in diodicity from ~95.4 to ~1117.4 and significant reduction in back-flow in the system. As result, fabricated prototypes demonstrate experimental fluid flow rates from ~600 to ~3000 μL/min, without the use of electricity. Furthermore, the on-chip integration of human-powered actuators into larger 3D printed microfluidic networks is demonstrated in the form of single finger-actuated pressure source to enable an integrated two-fluid pulsatile mixer prototype.
The fourth and final part of this dissertation proposes the concept of an entirely-3D printed, hand- held microfluidic device to enable human-powered fluid sample collection and entirely on-chip enrichment and multiplexed pathogenic detection. A custom drop-casting procedure is developed and experimentally-optimized to pre-treat fabricated prototypes with nutrients and bacteria- specific colorimetric reagents for on-chip biological enrichment and practical optical pathogen detection. The quantitative limit-of-detection of the employed proof-of-concept colorimetric detection scheme is experimentally determined to be ~1x10^6 cells/mL in 6 hours, a result validated by the demonstration of collection, enrichment and detection of E. coli bacteria in model drinking water using the actual fabricated prototype device.
