Encapsulating and Interfacing with Atomically Thin Nanomaterials for Integration in Transistors and Displays

Abstract

Atomically thin semiconducting nanomaterials – particularly, molybdenum disulfide (MoS2) and carbon nanotubes (CNTs) – have many inherent properties that make them promising building blocks for electronic devices. The nanometer-scale thickness of MoS2 and CNTs enables them both to maintain their excellent electronic properties even in extremely small devices, allowing the dense integration of many connected devices at the scale required for high-performance computing. Since their thickness is comparable or even smaller than many biological molecules, they are promising electrical transducers for developing biosensing applications. Monolayers of MoS2 and CNTs are quite strong and flexible, making them an attractive option for soft electronics such as wearable devices or flexible electronic displays. Yet, despite these considerable advantages, the atomic thinness and chemical inertness of these nanomaterials can also make them challenging to handle and interface with when integrating into electronic devices. Both nanomaterials face obstacles that must be overcome to facilitate their broader adoption, commercialization, and integration into future transistor and display technologies. This dissertation describes the development of several methods to facilitate the integration of MoS2 and CNTs into electronic devices and systems.Molybdenum disulfide, like other two-dimensional nanomaterials, is very sensitive to its ambient environment. While MoS2 does not permanently oxidize easily, it can adsorb oxygen or water molecules onto its surface, leading to threshold voltage instability in MoS2 transistors. When left unchecked, this type of instability can lead to signal drift, which is a critical issue for sensing applications. Yet, in practice, MoS2 has also proven to be challenging to deposit other conductive or insulating layers onto without causing disorder or damage at the interface. In this dissertation, encapsulation layers were discovered that effectively passivate the surface from molecular adsorption and stabilize the threshold voltage of MoS2 field-effect transistors (FETs) without introducing defects or damage. Specifically, aluminum oxide and SU-8 polymer encapsulation both proved effective – enabling uninterrupted operation of MoS2 transistors for days at a time, assuaging degradation concerns and providing insight into the transient mechanisms at play. Then, to enable a transition from dry to hydrated sensing environments, fabrication processes were developed to build MoS2 FETs on insulating quartz substrates. These liquid-gated devices showed near-ideal subthreshold switching and minimal leakage, which can enable future studies of Debye screening length effects in hydrated nanomaterial-based electronic biosensors. Taken together, these discoveries validate the robustness of these atomically thin semiconductors for long-term deployment in complex environments and can guide the design of future sensors. For CNTs, perhaps the most significant challenge is the difficulty in assembling large quantities of these nanometer-diameter wires into wide-area-coverage films at the density and scale needed for large devices, such as electronic displays. This has been greatly aided by advancements in producing polymer-wrapped CNTs suspended in liquid solutions, which enable rapid coating of large areas directly deposited from solution. Yet, these thin films of polymer-wrapped CNTs can be difficult to effectively contact, frequently exhibiting contact resistance an order of magnitude higher than individual CNTs. Even more promising are recent advancements in the development of nanomaterial-based printing techniques, which enable low-cost placement and patterning of CNT thin films integrated with other printable nanomaterials. While this technology has enabled impressive demonstrations of fully printed fundamental circuit components like resistors, capacitors, sensors, and even transistors, the development of fully printed microelectronic systems, such as electronic displays remains limited owing to the difficulty of integrating the many solution-processable conductive and insulating layers needed for such systems without intermixing. In this dissertation, methods were developed to map the uniformity of solution-processed CNT thin films, revealing challenges that emerge in fabrication, gating, and contacting as thin-film devices are scaled down to the length scale of individual nanotubes. A unique plasma-formed edge contact geometry was developed, which enables selective removal of polymer residue from the contact interface, providing enhanced uniformity and reliability in scaled devices. Finally, a process was developed to integrate these solution-processed CNT thin films into liquid crystal displays (LCDs) using a low-cost printing technology. This enabled the first demonstration of LCDs driven by a backplane of fully printed CNT thin-film transistors. These devices were fabricated entirely without vacuum-based processing and utilizing only recyclable carbon-based conducting, semiconducting, and insulating nanomaterials. Taken together, the combined advancements in contact geometry and print processing offers the potential to realize large-area CNT-based circuits with either vastly improved performance or dramatically reduced costs required to motivate a transition from mature display technologies to emerging micro-LED displays, low-cost flexible displays, and beyond.