Fluidic Microvalve Digital Processors for Automated Biochemical Analysis
The development of microfluidic sample processing and microvalve technology offers significant, thus far unmet opportunities for the miniaturization and large scale integration of automated laboratory systems. In this dissertation, the transistor-like nature of monolithic membrane valves for control of airflow and fluid flow is exploited to develop microfluidic processors for performing diverse bioassay procedures on a common programmable microchip format.
The transistor-nature of pneumatic microvalves is first exploited to fabricate devices that perform AND, OR, and NOT transistor-to-transistor logic operations. With this system, microvalves are used to control the actuation of other microvalves by regulating airflow. As a demonstration of computational universality, these operators are combined to perform more complex digital logic operations including binary addition. Integrated logical circuits such as demultiplexers and latching circuits are valuable because they reduce the power consumption and control equipment required for controlling large arrays of microvalves.
A digital microfluidic Automaton is demonstrated using 2-dimensional microvalve array technology. Digital transfer of fluids between microvalves enables precise and rapid metering of nanoliter scale sample volumes. Programs for reagent routing, mixing, rinsing, serial dilution, storage/retrieval and many other operations are demonstrated. Protocols for on-chip reagent mixing and serial dilution are optimized to achieve linearity over a 1000-fold dilution range. These optimized programs are combined to develop a rapid, quantitative assay for hydrogen peroxide, a biomarker of oxidative stress. A sub-micromolar limit of detection is demonstrated with an 8.5 min program runtime, thus establishing this platform as an effective tool for the miniaturization and automation of multi-step bioassays.
An extension of the Automaton platform to inhomogeneous immunoassays is presented. Capture antibody derivatized magnetic particles are utilized as a solid substrate. Effective procedures are developed for the transport, capture, rinsing, and delivery of reagents to magnetic particles in the digital microfluidic array. These procedures are used to demonstrate a model immunoassay for mouse IgG.
Automaton protocols are developed for processing and combinatorial mixing of a wide range of sample volumes. The ability to process large (µL scale) sample volumes enables the detection of low titer targets, and the modular coupling of the Automaton to a wide range of off-chip analytical detection instruments. The utility of these procedures is demonstrated for automated labeling of carboxylic acids for analysis with the Mars Organic Analyzer capillary electrophoresis instrument. Optimized programs result in peak efficiencies that are within 1% of those for manually labeled samples. In addition, protocols for µL scale serial dilution are presented, and an effective programming language is developed for these operations.
The prospects for this technology are also presented including 1) demultiplexed control of the Automaton, 2) fully autonomous sample processing for the Mars Organic Analyzer, 3) nucleic acid amplification and analysis, and 4) high-sensitivity protein biomarker detection. The technology developed in this dissertation enables miniaturized and automated analysis of metabolic, protein, and nucleic acid biomarkers using a common, programmable microchip platform.