UC Berkeley

The ability to precisely control and characterize the electronic structure of low-dimensional materials is essential for the realization of device logic beyond the limits of Moore’s Law. In addition, such materials can act as versatile testbeds for the study of fundamental condensed matter physics, including magnetism and superlattice physics. Due to the sensitivity of their electronic properties to atomic-scale structure, these reduced-dimensional materials must be fabricated and designed with atomic-precision. This dissertation focuses on the on-surface growth and electronic structure characterization of two-dimensional (2D) covalent organic frameworks (COFs) and one-dimensional (1D) graphene nanoribbons (GNRs) studied using low temperature (LT) scanning tunneling microscopy (STM) and spectroscopy (STS).

The first part of this dissertation details the study of boroxine-linked COFs grown and characterized on both Au(111) and single-layer hexagonal boronitride (hBN) on Cu(111). On Au(111), we image the LDOS of unoccupied COF states of this prototypical class of COFs for the first time. These results show that the reduction of COF pore-size is associated with a confinement-based destabilization of unoccupied COF orbitals. For COFs studied on hBN/Cu(111), the addition of the insulating hBN layer significantly reduces the COF-substrate hybridization. This leads to a much more detailed spectral profile for STS conducted on boroxine-linked COFs, which we determine reflects the COF’s Kagome lattice.

The second part of this dissertation focuses on several studies, each associated with growth and electronic structure characterization of a specific target GNR structure. The first section centers around a series of studies on boron-doped GNRs. We determine that boron substituents form two new in-gap orbital states that exhibit parity-dependent hybridization with the underlying Au(111) substrate. The next group of studies focuses on chevron-based GNR heterojunctions (HJs), including edge-doped carbazole and fluorenone systems, and wide/narrow junctions realized via hierarchical growth strategies. We were able to observe dopant-induced shifts in the GNR electronic structure and to uncover a subtle charge-transfer process in differentially-doped GNR HJs that ultimately determines the magnitude of HJ orbital energy offsets. In the next series of experiments, we demonstrate the ability to precisely control GNR topology in HJs of 7- and 9-atom-wide armchair GNRs (7-AGNRs and 9-AGNRs). Based on the specialized reaction geometry of a “linker” molecular precursor, we are able to realize topological junction states as single junctions, coupled double-junctions, and junction superlattices. The resulting electronic structures are observed to be well explained using basic 1D tight binding concepts and suggest a general route to engineering GNR electronic properties based on precise implantation of topological boundary states. Here we also discuss an intuitive realization of 1D GNR topology in terms of real-space charge position, with a simple mapping onto Clar Theory. The final study is a spinoff of the above-described topological superlattice, in which we embed a symmetric zero-mode superlattice into a semiconducting GNR backbone in order to realize metallicity in GNRs for the first time. Our results demonstrate general criteria for designing GNR metallicity and reveal strategies for engineering frontier bandwidth through control of sublattice polarization of zero-mode wavefunctions.

The results presented in this dissertation reveal the broad range of tunable electronic structures accessible via atomically-precise on-surface synthetic strategies. In so doing, we are able to uncover general design principles for engineering low-dimensional materials at the smallest possible length scales.