Microfluidics Automation: Robotics

    I developed a tool that is capable of serially interfacing & manipulating a set of PDMS-based substrates. The tool is comprised of a set of micron- precision translation stages, a custom-designed injector/interface module, a vision system, and a set of solenoids for flow control. Custom software written in LabWindows CVI is used to coordinate the various components. While the initial scope of this tool was to automate the DNA flow-patterning process that is used to create custom assay substrates within our group, minor modification will allow it to perform any PDMS-based experiment that is normally done by hand.

DNA Microarray Patterning

    With the development of the DEAL technique, our lab abandoned electronic sensing in favor of colorimetric, ELISA-type assays. ELISA assays are commonly considered the gold standard for protein detection, but typically require relatively large amounts of reagent & analyte. The DEAL technique allowed us to transform standard DNA microarrays into antibody microarrays, where each unique DNA spot becomes sensitive to a specific protein. This enables us to run hundreds of ELISA-type protein detections in parallel and within a relatively small space, allowing for drastically reduced reagent and analyte volumes among other advantages. I worked extensively on methods to create second generation microarrays with more flexible spot size and morphology than commercially-available products, with the ultimage goal of further reducing reagent requirements and enabling new classes of experiments.

Microcontact Printing

    Microcontact printing utilizes PDMS as a physical stamp: extruded features are inked with molecules of interest and then brought into contact with the surface to be patterned. Single-stranded DNA is an ideal ink for such a process, as it initially adheres well to PDMS through hydrophobic interactions but then transfers efficiently to postively-charged substrates (aminated or poly-lysine slides) via electrostatic interactions. After an initial demonstration of the transfer process I was quickly able to reduce feature size by an order of magnitude (as compared to traditional microarrays) and incorporate arbitrary spot morphology. I also solved numerous technical issues associated with the process that are critical if such microarrays are to be used for subsequent experiments. Chief among these was aligning multiple stamps to a single substrate with micron precision such that each DNA spot/sequence can be located reliably relative to its neighbors; a standard Karl Suss MA6 was adapted to accomplish this goal. Moreover, I developed a technique to print extremely low-aspect-ratio features at any distance from one another without intermediary support structures, overcoming a common problem in the field known as "roof collapse". Microcontact printing ultimately proved to be incompatible with DEAL chemistry due to monomer contaminants left behind by the stamping process.

Microfluidic Flow Patterning

    In spite of a disappointing end to the microcontact printing project, the specific failure mode - substrate surface contamination - inspired a new approach to generating DNA microarrays. Indeed, the process of microfluidic flow patterning represents the exact inverse of microcontact printing: here a flat PDMS mold is placed onto a substrate with cavities/channels (rather than extrusions) at the spots that are to be patterned. A DNA solution is introduced into the features and allowed to evaporate, depositing DNA in the shape of the channel while relegating any contaminants to the spaces in between. With the introduction of 2-layer microfluidics with crossover channels, multiple discontinuous features could be patterned in a single step. This approach proved to be compatible with DEAL chemistry and serves as the basis for all of the high-denisty DEAL assays performed in the Heath Lab today, including blood analysis and single-cell secretion experiments.

Silicon Surface Functionalization

    A critical component of electrical sensing with Si nanowires is decorating the Si surface with capture agents that will spatially localize specific analytes of interest to the wire. Using fluorescence-based assays, I found that our existing strategies for covalent surface modification were relatively inefficient and yielded poor surface coverage of our sensors. I explored alternative methods of covalently functionalizing Si surfaces with higher efficiency and with spatial selectivity.

Protein Sensing with Silicon Nanowires

    Silicon is an extremely common, relatively cheap semiconductor material that is ubiquitous in today's electronics. The explosive growth of the computer industry over the past two decades spurred comprehensive study of the material, and the processing techniques used to create Si-based devices are very well characterized today. If Si could be configured as a chemically-gated Field Effect Transistor (FET), it would integrate seamlessly with today's electronics, providing a relatively cheap, compact, high-throughput method of exploring the biological molecules that underlie a wide variety of diseases. We chose to use nanowire structures for our FET sensors because the high ratio of surface area/volume inherent to nanowires renders them significantly more sensitive to local field effects. Under the guidance of postdoctoral scholars Peter Willis & Kristen Beverly, I investigated both DC & impedance detection methods on ca. 25nm wires defined by e-beam lithography. The project bore robust results for simple systems (e.g. biotin/streptavidin binding in low-salt buffers) but failed to yield convincing data with more complex analytes.