This Is AuburnElectronic Theses and Dissertations

Microfluidic Circuit Designs for Nanoliter-Scale Flow Control and Highly Sensitive Quantitative Bioanalysis

Date

2017-12-04

Author

Negou Negou, Jean

Type of Degree

PhD Dissertation

Department

Chemistry and Biochemistry

Abstract

Within this dissertation, we have developed several new microfluidic circuit components that permit precise flow control at the nanoliter scale, and we have demonstrated their successful applications in either proof-of-concept experiments or in true bioanalytical settings. We first introduced key concepts of microfluidics (Chapter 1), the following two chapters are focused on improvements in the microfluidic sample chopper ($ \mu $Chopper). Secondly, we improved upon our $ \mu $Chopper’s performance by adding automation with on-chip pneumatic valving (Chapter 2). The device, in combination with lock-in analysis, was shown to significantly reduce 1/f noise, which is a challenging problem in small scale detection of low magnitude signals. This device allowed tightly phase-locked fluorescence detection with a bandwidth of only 0.04 Hz for droplets generated at 3.5 Hz, which resulted in a detection limit of only 12 pM fluorescein (3.1 x 10$ ^{-19} $ moles) or 9.3 nM insulin (1.9 x 10$ ^{-16} $ mol) using homogeneous immunoassays. The device was finally applied to the quantification of free fatty acid (FFA) uptake by 3T3-L1 adipocyte cells, and for the first time, single adipocyte FFA uptake was measured at 3.5 x 10$ ^{-15} $ mol cell$ ^{-1} $. Thirdly, we expanded the applicability of this high performance detection system by designing and testing a multi-channel $ \mu $Chopper. The device consisted of six sample input channels and one oil input, again with automated control using seven on-chip valves. The next-generation $ \mu $Chopper was shown to perform various modes of analyses such as constant calibration mode, mixed mode, multiplex mode, and standard addition mode. Each mode was characterized for future application using fluorescein to mimic biological assays. We present result showing that this next-generation $ \mu $Chopper can not only reduce 1/f noise, but also reduce human error. Furthermore, analysis times were significantly faster since all calibrations could be completed in a single run of the device. Fourthly, we implemented a new strategy for rapid, on-chip protein detection by combining a customized rotary mixer design with a novel, homogeneous (mix-and-read) protein assay (Chapter 3). Rapid mixing of a few nanoliters of each assay component (sample and probes) was achieved within 2.2 seconds. Aptamer or antibody-oligonucleotide based assays were used to measure protein quantities in these small-volume mixers. By combining the benefits of automated rotary mixer with homogeneous protein assays, we demonstrated that proteins could be detected at quantities as low as the attomole range. Finally, we outlined a frequency-dependent study on passive microfluidic flow control components, namely autoregulators (Chapter 5). Using an autoregulator design developed by others, we sought to characterize the component's frequency dependence to gauge its usability in more complex microfluidic circuits. Our approach first relied on simulation of the microfluidic circuit using electrical circuit analogies. The results confirm that the autoregulators were functional as nonlinear components, permitting feedback control of the output fluidic flow rate and giving constant output flow rates at a range of input pressures. We also confirmed for the first time that these components are functional at relatively high switching frequencies-up to 16 Hz-a result that matched well with circuit simulations. Finally, we applied this knowledge to design and demonstrate a new microfluidic circuit, a flow rate mirror. Results showed that an autoregulator could control the flow rate in a parallel fluidic path, essentially independent of the downstream flow resistance. Within this work, we also demonstrated two new methods for characterizing fluidic circuit components based on real-time microscopy (fluorescence and transmission modes) to allow continuous measurements of pressures and flow rates in these devices. Overall, we have demonstrated several new microfluidic circuit designs for nanoliter-scale flow control that allow sensitive bioanalytical measurements or provide novel flow control methodologies that should be useful for future studies in microfluidic systems.