Nano-Bio Hybrid Electronic Sensors for Chemical Detection and Disease Diagnostics

The need to detect low concentrations of chemical or biological targets is ubiquitous in environmental monitoring and biomedical applications. The goal of this work was to address challenges in this arena by combining nanomaterials grown via scalable techniques with chemical receptors optimized for the detection problem at hand. Advances were made in the CVD growth of graphene, carbon nanotubes and molybdenum disulfide. Field effect transistors using these materials as the channel were fabricated using methods designed to avoid contamination of the nanomaterial surfaces. These devices were used to read out electronic signatures of binding events of molecular targets in both vapor and solution phases. Single-stranded DNA functionalized graphene and carbon nanotubes were shown to be versatile receptors for a wide variety of volatile molecular targets, with characteristic responses that depended on the DNA sequence and the identity of the target molecule, observable down to part-per-billion concentrations. This technology was applied to increasingly difficult detection challenges, culminating in a study of blood plasma samples from patients with ovarian cancer. By working with large arrays of devices and studying the devices' responses to pooled plasma samples and plasma samples from 24 individuals, sufficient data was collected to identify statistically robust patterns that allow samples to be classified as coming from individuals who are healthy or have either benign or malignant ovarian tumors. Solution-phase detection experiments focused on the design of surface linkers and specific receptors for medically relevant molecular targets. A non-covalent linker was used to attach a known glucose receptor to carbon nanotubes and the resulting hybrid was shown to be sensitive to glucose at the low concentrations found in saliva, opening up a potential pathway to glucose monitoring without the need for drawing blood. In separate experiments, molybdenum disulfide transistors were functionalized with a reengineered variant of a μ-opiod receptor, a cell membrane protein that binds opiods and regulates pain and reward signaling in the body. The resulting devices were shown to bind opiods with affinities that agree with measurements in the native state. This result could enable not only an advanced opiod sensor but moreover could be generalized into a solid-state drug testing platform, allowing the interactions of novel pharmaceuticals and their target proteins to be read out electronically. Such a system could have high throughput due to the quick measurement, scalable device fabrication and high sensitivity of the molybdenum disulfide transistor. Degree Type Dissertation Degree Name Doctor of Philosophy (PhD) Graduate Group Physics & Astronomy First Advisor A. T. Charlie Johnson This dissertation is available at ScholarlyCommons: http://repository.upenn.edu/edissertations/1079