Search

Graphene has gained much attention as a biosensing material since its discovery and characterization due to its highly sensitive electronic properties. Reported work on graphene as a biological sensor has focused on solution-gated graphene transistors (SGGFETs) that can measure the perturbed channel conductivity in response to environmental changes in the proximity of the graphene surface. Electrodes present a simpler method of biological detection, both from the operation and the fabrication standpoint. Investigation of graphene's electrochemical properties has reported higher electron transfer kinetics occurring at the edges than at the basal plane of the carbon allotrope. Yet, inconsistencies in sample preparation impede an accurate comparison of electrode performance. This thesis examines the fabrication and characterization of graphene microelectrode arrays made with a variety of graphitic materials that exhibit differences in the number of layers, domain size, defects and substrate. We examine, for the first time, the electrochemical properties of Van der Waals CVD graphene grown on sapphire substrates and electrode arrays made on epitaxial graphene grown on silicon carbide. We find no significant performance differences with mono-, bi- and multilayer graphene, but do observe microelectrode edge effects becoming more dominant in multilayer devices as they are scaled down. CVD graphene on sapphire, with domain sizes as small as 100-200 nm, show higher sensitivity and epitaxial electrodes display the lowest detection limit (1[MICRO SIGN]M) and fastest electron transfer kinetics, with the latter presumed to be effect of the high degree of corrugation in the material and consistent with reports that higher curvature leads to faster kinetics [1]. To further examine the effect of the edges, we patterned electrodes of the same area varying only the perimeter. For clean electrodes, the perimeter to area ratio had little effect on the electrode sensitivity. However, after exposure to a low-power 30-second ozone plasma, the electrode sensitivity and electron kinetics improved, increasing by almost by two-fold with increasing electrode length. This result is consistent with the graphene edges becoming more electroactive through functionalization and result implies that graphene electrode sensitivity can be increased by functionalization and optimization of the electrode geometry.

Recently, ZnO has garnered widespread attention in the semiconductor community for its large set of useful properties, which include a wide bandgap and its resulting optical transparency, a large exciton binding energy, a significant piezoelectric response, and good electrical conductivity. In many ways, it shares many properties with a widely used and technologically important semiconductor GaN, which is widely used for blue LEDs and lasers. However, ZnO cannot substitute for GaN in most optoelectronic applications, because it cannot be doped p-type. On the other hand, unlike many traditional, covalently bonded semiconductors like GaN, ZnO can be easily formed aqueous solutions at close to room temperature and pressure in the form of large crystals or a variety of nanostructures, making possible applications that are normally very difficult with traditional semiconductors. In this light, we aimed to take advantage of aqueous solution-based, ZnO growth techniques and incorporated ZnO structures novel optoelectronic and photonic structures. By controlling the morphology of ZnO, we studied the effects of nanowire-based \(ZnO/Cu_2O\) solar cells. Carrier collection was increased using a nanowire-based device architecture. The main result, however, was the time evolution of the performance of these devices due to the movement of ionized defects in the material. The effects of geometry on the ageing characteristics were studied, which showed that the carrier collection could be increased further with ageing in a nanowire \(Cu_2O\) solar cell. The aging behavior was substantially different between nanowire and planar solar cells, which implies that future design of nanostructured solar cells must long term aging effects. In addition to solar cells, we explored the possibilities of using aqueous solution growth of ZnO to fabricated whispering gallery mode optical cavities and waveguides for enhancing extraction from a single photon source. In both applications, we used templated growth of ZnO to fabricate geometrically (near) perfect rods and disks for these photonics applications. Finally, since epitaxy is important in the process of optimizing device performance and fabrication, we showed the ability to grow ZnO epitaxially on single crystalline plates of Au, expanding the options of epitaxial substrates to include a metal.

In this work we present a feasibility study of epitaxial growth of thin films of strontium cobaltite, SrCoO3-\delta. The properties of strontium doped lanthanum cobaltite, La1-xSrxCoO3, have been widely studied for dopant concentration x3 in polycrystalline samples can only be obtained under high pressures of oxygen (>10kbar) or by electrochemical oxidation. However, theoretical calculations predict a phase change with respect to strain in epitaxially grown samples, from ferromagnetic-metallic behaviour in the bulk state to insulating-ferroelectric-antiferromagnetic behaviour for strongly strained films. This provides strong motivation for epitaxial growth of SrCoO3-d films. In this work we will present a feasibility study by using the methods of high-pressure oxygen sputtering (typically 1.0-4.0mbar) on SrTiO3(001) and LaAlO3(001) substrates. As anticipated, the presence of oxygen vacancies is a severe problem, but also epitaxial stabilization of non-cubic phases, an unexpected issue, arises. These are found to grow in multiple orientations. Overall, the samples exhibit only weak or no ferromagnetism, even though bulk SrCoO3 is known to be a strong ferromagnet. Based on the results, we present an outline for suggested further research on this topic.

Spintronic systems, where the spin degree of freedom is exploited for storing and processing information, have gained tremendous interest in the recent years because of their potential in revolutionizing the current electronics industry. Low-dimensional magnetic structures, such as epitaxial magnetic films on insulators or semiconductors, are particularly interesting because they form the key components in many important spintronic applications including gigabit nonvolatile memory, magnetic tunnel junction and spin injector. With advanced ultra-high-vacuum growth techniques such as molecular beam epitaxy, it is possible to create low-dimensional magnetic structures and magnet/semiconductor hybrid structures with atomic precision. Here in this thesis, I will show several interesting magnetic nanostructures grown using this method. These structures range from thin films of magnetic alloys, to ultrathin transition metal based magnetic layers on GaN and W(110) surfaces, and antiferromagnetic Mn3N2 nanopyramids. To study the magnetic properties of these structures down to nanoscale, and to correlate magnetism with structure, scanning tunneling microscopy (STM) and spin-polarized scanning tunneling microscopy (SP-STM) are employed.As one example, in the Mn on GaN(0001) project, we have discovered a class of novel well-ordered striped superstructures with local sqrt3xsqrt3-R30 ordering. Combining STM and first-principles theory, we find that Mn atoms react with excessive surface Ga atoms and form a high-density two-dimensional MnxGa1-x structure. The surface electronic states are found to be dominated by the highly spin-polarized Mn d electrons. As a consequence, Mn atomic sites can be directly identified in STM images. For the narrowest stripes, calculations show a row-wise antiferromagnetic ground state, which is observed in real space at room temperature as an asymmetry in the density of states.As the second example, I have applied SP-STM to study the nanoscale magnetization of antiferromagnetic Mn3N2 films. Results have shown that the surface exhibits a spin pyramid structure, where the magnetism is strongly correlated with the surface topography. Using spin-polarized dI/dV mapping, different layers can be clearly distinguished due to their different chemical and spin properties. We will show that it is possible to separate the contributions from both the electronic and the magnetic structure by applying a small magnetic field. The field rotates the tip magnetization axis causing concomitant change in the magnetic sensitivity while keeping the electronic structure unchanged. The rotation of the tip magnetization also revealed a perpendicular spin-reorientation on one subset of the surface terraces.As shown through these projects, STM and SP-STM provide an unprecedented power to resolve magnetism and spin ordering with atomic precision. They also provide the ideal method for correlating magnetism with structure and electronic properties. By studying the interplay between structural defects such as anti-phase domain boundaries and magnetism, these projects not only unravel intriguing magnetic phenomena at reduced dimensions, but also provide methods for tailoring the material toward real-world applications.

As scaling continues in the semiconductor industry, silicon-on-insulator (SOI) wafers are increasingly becoming the substrate of choice, due to higher channel mobility, effective device isolation, reduced short channel effects, minimized parasitic capacitance, and therefore higher speed, compared to a regular silicon wafer. Current methods of SOI wafer production, however, will have difficulty achieving the desired silicon device layer and buried oxide insulator layer thicknesses and eliminating interface roughness as scaling proceeds. We propose "Floating Epitaxy SOI" as a novel method of SOI production utilizing an all in-situ growth process. Floating Epitaxy SOI involves Molecular Beam Epitaxy deposition of an epitaxial template oxide, oxidizing through the epitaxial template layer to establish the insulation layer, and silicon growth on top of the epitaxial template oxide layer (which is now "floating" on top of an amorphous oxide layer). The key to this process is the epitaxial oxide template layer, which must deposit on the silicon substrate as an atomically smooth film with a lattice parameter close to that of silicon and must be sufficiently stable in both an oxygen an in vacuum annealing to relatively high temperature to achieve Floating Epitaxy SOI. Although many researchers have examined epitaxial oxides on silicon, this study focuses on epitaxial films over large area substrates, while virtually all other studies report on growth on small substrate sizes. Also, the oxide stability limits on silicon in vacuum have not been thoroughly established by previous work, and are investigated here. The growth and thermal stability of this epitaxial oxide template layer are discussed, as well as brief results for through-oxidation "floating" of the template oxide layer and silicon growth experiments. BaO, SrO, CaO, Ba[subscript 1-x]Sr[subscript x]O, SrTiO₃, CaTiO₃, and Ca[subscript 1-x]Sr[subscript x]TiO₃ were successfully epitaxially deposited on Si(001) substrates. A 64:36 Ba:Sr ratio was used for the solid solution of Ba[subscript 1-x]Sr[subscript x]O, in order to achieve close lattice matching with silicon; a 50:50 Ca:Sr ratio was used initially for the Ca[subscript 1-x]Sr[subscript x]TiO₃ solid solution, an attempt to mediate SrTiO₃'s 2% lattice mismatch with silicon and CaTiO₃'s orthorhombic structure. Alloying SrTiO₃ with calcium to alter the lattice parameter has not been studied much to this point in thin films, and this is the first demonstration of Ca[subscript 1-x]Sr[subscript x]TiO₃ and CaTiO₃ thin films grown directly on silicon. Reflection High Energy Electron Diffraction patterns of both Ba[subscript 1-x]Sr[subscript x]O and Ca[subscript 1-x]Sr[subscript x]TiO₃ indicated high quality 2D epitaxial films. A thin (3 monolayer) film of Ba[subscript 1-x]Sr[subscript x]O is stable on silicon to 535°C in vacuum, while a 5 monolayer Ca[subscript 1-x]Sr[subscript x]TiO₃ film survives to 740°C in vacuum, but roughens from a 2D toward a 3D surface above ˜650°C. Of the epitaxial oxides studied, the solid solution Ca[subscript 1-x}Sr[subscript x]TiO₃ would be the best choice for Floating Epitaxy SOI, based on epitaxial growth quality and stability. High-resolution TEM indicates the presence of an amorphous interfacial layer at the SrTiO₃Si interface, as grown. X-ray diffraction confirms an epitaxial film, with a lattice parameter larger than that of bulk SrTiO₃, likely due to oxygen deficiency in the film. Annealing 17.5nm SrTiO₃Si(001) at 800°C in 5.5 Torr of oxygen for 30 minutes results in an equivalent oxide thickness of 10.3nm, sufficient for scaling to 2020. X-ray diffraction after annealing reveals a still-epitaxial SrTiO₃ film, with sharper 2θ and χ peaks and a lattice parameter closer to that of bulk SrTiO₃. These results validate the "floating" epitaxy approach: an epitaxial film remains on top of an amorphous insulator, after through-oxidation of the substrate. Direct deposition of epitaxial silicon on Ca[subscript 1-x]Sr[subscript x]TiO₃ and solid-phase epitaxy of silicon on a CaTiO₃ film are promising, but interface engineering or a surfactant may be required to achieve a high quality, single crystal silicon layer.

This thesis explores the effects of fluctuations and discreteness on the growth of physical systems where diffusion plays an important role. It focuses on three related problems, all dependent on diffusion in a fundamental way, but each with its own unique challenges. With diffusion-limited aggregation (DLA), the relationship between noisy and noise-free Laplacian growth is probed by averaging the results of noisy growth. By doing so in a channel geometry, we are able to compare to known solutions of the noise-free problem. We see that while the two are comparable, there are discrepancies which are not well understood. In molecular beam epitaxy (MBE), we create efficient computational algorithms, by replacing random walkers (diffusing atoms) with approximately equivalent processes. In one case, the atoms are replaced by a continuum field. Solving for the dynamics of the field yields---in an average sense---the dynamics of the atoms. In the other case, the atoms are treated as individual random-walking particles, but the details of the dynamics are changed to an (approximately) equivalent set of dynamics. This approach involves allowing adatoms to take long hops. We see approximately an order of magnitude speed up for simulating island dynamics, mound growth, and Ostwald ripening. Some ideas from the study of MBE are carried over to the study of front propagation in reaction-diffusion systems. Many of the analytic results about front propagation are derived from continuum models. It is unclear, however, that these results accurately describe the properties of a discrete system. It is reasonable to think that discrete systems will converge to the continuum results when sufficiently many particles are included. However, computational evidence of this is difficult to obtain, since the interesting properties tend to depend on a power law of the logarithm of the number of particles. Thus, the number of particles included in simulations must be exceedingly large. By combining continuum and discrete methods into the same model, we are able to probe the high particle density limit, while still retaining the effects of discreteness. Using this method, we have obtained new results on the roughness of pulled fronts, which disagrees with current theoretical work.

ZnO thin films were produced on GaN(0001) epilayers and ZnO(0001) substrates utilizing an iterative process requiring a structured low temperature (480°C) layer followed by a high temperature (800°C) densification step to create approximately 200 nm of contiguous film. This process is subsequently repeated to achieve thicker films with each iteration producing approximately 200 nm of dense film. Diethylzinc was used as the zinc source, UHP oxygen (O2) as the oxygen source, and UHP argon as both the carrier and diluent gas. Nitrous (N2O) and nitric oxide (NO2) were also used both as potential oxygen sources in the pure state as well as mixed with oxygen in the chamber and for nitrogen doping of the growing film. Major impurities of C, H, and N were incorporated into the films with the majority of the incorporation occurring during the low temperature step. Films grown using N2O + O2 contained an average of 5 x 1017 cm-3 atomic nitrogen while films using NO2 + O2 had an average nitrogen concentrations of 9 x 1019 cm-3. Needle microstructures were observed for low temperature layers using O2 and N2O + O2, while networked structures formed when using NO2 + O2. The surface of the densified films contained hexagonal pits that increased in number and depth with an increase in film thickness. Triple-axis XRD measurements indicated that the crystal structure of the films mimic the underlying substrates. A comparative analysis of undoped and N-doped films using capacitance voltage and photoluminescence measurements showed that the N-doped films were more insulating than the undoped films and the incorporation of nitrogen decreases the amount of excitonic peaks observed in the PL spectra. The 3.367 eV ionized donor bound exciton becomes dominant in N-doped films relative to the 3.361 eV donor bound exciton that dominates the undoped films. A preliminary inductively coupled plasma etching study determined that the smoothest sidewalls and surfaces were obtained using an ICP power of 600 W, a DC bias of 50 V, 5 mtorr total pressure, and 20 sccm of pure flowing BCl3 The etch rate under these conditions was 40 nm/min.

We present high resolution electron density maps of two types of systems investigated. The first type consists of two InAs films grown on GaSb (001) substrates (one grown using As4, the other one grown using As 2) and two GaSb films grown on InAs (001). The semiconductor samples were 9 monolayers thick and were grown using MBE. The three dimensional electron densities were obtained measuring the intensities along the Bragg rods and analyzing them with the newly developed Coherent Bragg Rod Analysis (COBRA) phase retrieval method. The study represents the first attempt to use a phase retrieval method in the characterization of a buried semiconductor heterostructure. The semiconductor was treated as a quaternary of form Ga mIn1-mAs nSb1-n. The m(z) and n( z) compositional fractions and the vertical lattice spacing profiles were extracted. The results tend to indicate that the transition regions have a quaternary composition and are relatively narrow (1.5 unit cells). In the case of the GaSb film, significant presence of As is observed. The lattice constant in the film is not constant and has a minimum ∼1.5 unit cells below the surface. The second type of system investigated consists of a thin film of PbTiO 3 on SrTiO3 (001), with and without a Pt electrode on top. The 50 A thick PbTiO3 film was grown by sputtering. The ED maps were investigated for peak position shifts. The O1 lattice seems to be shifted toward the film substrate interface in the case of the bare film. In the case of the Pt electrode region, its behavior is more complex. The O2 sublattice exhibits a shift toward the surface of the film, with a maximum of ∼0.25 A at 12 unit cells away from the substrate. The effect relaxes toward the surface. In the case of the bare film, the Ti peak width is slightly increased in the vicinity of both film interfaces. The presence of the Pt electrode seems to inhibit this behavior, fact that can be correlated with the inhibition of the presence of surface charge.