Thesis (PH.D.) -- Syracuse University, 2004. / "Publication number AAT 3132724."

<p> Efficient thermal management is becoming a critical issue for development of the next generation of electronics. As the size of electronic devices shrinks, the dissipated power density increases, demanding a better heat removal. The discovery of graphene&rsquo;s unique electrical and thermal properties stimulated interest of electronic industry to development of graphene based technologies. In this dissertation, I report the results of my investigation of thermal properties of graphene derivatives and their applications in thermal management. The dissertation consists of three parts. In the first part, I investigated thermal conductivity of graphene laminate films deposited on thermally insulating polyethylene terephthalate substrates. Graphene laminate is made of chemically derived graphene and few layer graphene flakes packed in overlapping structure. Two types of graphene laminate were studied: as deposited and compressed. The thermal conductivity of the laminate was found to be in the range from 40 <i>W/mK</i> to 90 <i>W/mK</i> at room temperature. It was established that the average size and the alignment of graphene flakes are parameters dominating the heat conduction. In the second part of this dissertation, I investigated thermal conductivity of chemically reduced freestanding graphene oxide films. It was found that the in-plane thermal conductivity of graphene oxide can be increased significantly using chemical reduction and temperature treatment. Finally, I studied the effect of defects on thermal conductivity of suspended graphene. The knowledge of the thermal conductivity dependence on the concentration of defects can shed light on the strength of the phonon - point defect scattering in two-dimensional materials. The defects were introduced to graphene in a controllable way using the low-energy electron beam irradiation. It was determined that as the defect density increases the thermal conductivity decreases down to about 400 <i> W/mK</i>, and then reveal saturation type behavior. The thermal conductivity dependence on the defect density was analyzed using the Boltzmann transport equation and molecular dynamics simulations. The obtained results are important for understanding phonon transport in two-dimensional systems and for practical applications of graphene in thermal management.</p><p>

<p> Over the last two decades, the emergence of quantum information science has uncovered many practical applications in areas such as communications, imaging, and sensing where harnessing quantum features of Nature provides tremendous benefits over existing methods exploiting classical physical phenomena. In this effort, one of the frontiers of research has been to identify and utilize quantum phenomena that are not susceptible to environmental and parasitic noise processes. Quantum photonics has been at the forefront of these studies because it allows room-temperature access to its inherently quantum-mechanical features, and allows leveraging the mature telecommunication industry. Accompanying the weak environmental influence, however, are also weak optical nonlinearities. Efficient nonlinear optical interactions are indispensible for many of the existing protocols for quantum optical computation and communication, e.g. high-fidelity entangling quantum logic gates rely on large nonlinear responses at the one- or few-photon-level. </p><p> While this has been addressed to a great extent by interfacing photons with single quantum emitters and cold atomic gases, scalability has remained elusive. In this work, we identify the macroscopic second-order nonlinear polarization as a robust platform to address this challenge, and utilize the recent advances in the burgeoning field of optical microcavities to enhance this nonlinear response. In particular, we show theoretically that by using the quantum Zeno effect, low-noise, single-photon-level optical nonlinearities can be realized in lithium niobate whispering-gallery-mode microcavities, and present experimental progress toward this goal. Using the measured strength of the second-order nonlinear response in lithium niobate, we modeled the nonlinear system in the strong coupling regime using the Schr&ouml;dinger picture framework and theoretically demonstrated that the single-photon-level operation can be observed for cavity lifetimes in excess of 500 ns for all the three waves in the interaction, provided a cavity of radius <i>R </i> &lt; 10 &mu;m is fabricated. </p><p> Experimentally, we showed that the absorption-limited quality (<i> Q</i>) factors for lithium niobate, <i>Q_intrinsic</i> &ap; 10⁸, can be achieved using diamond-turning methods for disk radii, <i>R</i> > 100 &mu;m, whereas for the smaller disks, additional rigorous polishing may be required. We also fabricated resonators as small as <i>R</i> &sim; 40 &mu;m via this method. In a millimeter-sized resonator, we experimentally demonstrated triply resonant sum-frequency generation, which allowed for an observation of the classical manifestation of the quantum Zeno effect, wherein line-splitting occurs due to the high efficiency intracavity frequency conversion. For the sub-100 &mu;m resonators, we present phase-matching calculations and dispersion-management techniques using analytical approximations and rigorous finite-element-method simulations. Experimentally, <i>Q </i>-factor measurements are shown, and we identify the specific short-comings of the fabrication procedure that may have led to the lower, surface-roughness-limited <i> Q</i>-factors. Finally, we identify pathways toward achieving the single-photon-level nonlinear optics using off-resonant nonlinear optics, which requires the simultaneous realization of phase-matching, large cavity lifetimes, and small mode volumes. We believe this would be feasible in the near future as more advanced fabrication and processing methods are developed for crystalline materials and novel nonlinear crystals are synthesized.</p>

Nowadays, for nanoelectronic devices, inter-atomic interactions and quantum effects are becoming increasingly important. For time dependent problem, such as high frequency electronics responses, or optical responses, the description of the system behaviour necessitates insights on the time dependent electron dynamics. In this dissertation, we proposed new effective modelling and numerical schemes to address the problem of time-dependent quantum simulations. An all-electron realspace real-time framework and TDDFT/ALDA type calculations are used for obtaining time dependent properties of molecules and nanostructures. Direct Hamiltonian diagonalizations are performed by using the innovative linear scaling eigenvalue solver FEAST. The spectral propagation schemes enable us to have much longer time steps, and it has been proven to be stable and highly scalable. A MPI parallel computing architecture is implemented, large monocles and nanostructures can be simulated in timely manner, which gives our model great advantage over traditional TDDFT calculation schemes. Optical absorption spectrum of small molecules are calculated and compared directly with the experimental values. Our results shows good agreement with experiments for a large selection of molecules. Finally, we apply our modelling and numerical schemes to study the (5,5) metallic Carbon Nanotubes, we successfully obtain the ―-and ← electrons plasmon which has been measured in experiments. Also, for the first time, we found the 1-D Luttinger liquid plasmon in 5 unit cell (5,5) CNT, whose plasmon velocity is consistent with other theoretical calculations.

<p> In this thesis, the conventional four-point measurement technique and the van der Pauw (vdP) method are systematically investigated in the presence of non-ideal conditions, namely, non-uniform carrier density distribution and absence of ohmic contacts, which are nonetheless commonly encountered in semiconductor characterizations. Upon understanding the challenges in the conventional methods, novel characterization techniques are developed to address these challenges. </p><p> A longitudinal magnetoresistance asymmetry method was developed to study the carrier density non-uniformity in two-dimensional samples. By analyzing the asymmetric longitudinal magnetoresistance under positive and negative <i> B</i>-fields, an analytical model based on a linear density gradient across the sample was deduced to quantitatively describe the asymmetry. Based on the theoretical model, a practical method was described which enabled one to experimentally measure the density gradient within a single sample. The method requires only measurements of longitudinal resistances <i>R_xx</i> and <i>R_yy</i> under both positive and negative <i>B</i>-fields, and equations have been provided to extract both the angle and the magnitude of density gradients from the measured resistances. The method was demonstrated in a GaAs quantum well wafer at cryogenic temperatures and <i>n</i>-GaAs bulk-doped wafer at room temperature. In both systems, the density gradient vectors extracted with our method matched well with the interpolated density gradient vectors estimated from actual density distribution maps as a base comparison set, suggesting that our method can be a universal extension of the vdP method to extract density gradients in various systems. The method also allows one to uncover the true local longitudinal resistivity &rho;<i>_xx</i> at the center of the sample, which the conventional vdP method cannot describe in the presence of non-uniform densities. The ability to find &rho;<i>_xx</i> makes it possible to study interesting physics in semiconductors such as interaction-induced quantum corrections to resistivity and valley filtering in multi-valley systems. </p><p> To extend the vdP method to cases where ohmic contacts are not available, a capacitive contact technique was introduced which sends current and senses voltage capacitively. A capacitive contact is formed between the buried conducting layer and the contact metal which is simply evaporated onto the sample. Systematic studies of four-point measurements with ohmic and/or capacitive contacts were conducted on a test sample and a Hall bar sample to demonstrate the effectiveness of the capacitive contact method. With a pre-defined capacitive scaling factor &gamma; and a measurement frequency band (<i>f_L</i> &sim; <i> f_H</i>), it was shown that capacitive contacts could extract the same four-point resistance as ohmic contacts, establishing the validity of the capacitive contact technique. </p><p> Built on the idea of capacitive coupling with capacitive contacts, a contactless electrical characterization probe was proposed. On the probe head, there are two types of metal gates: depletion gates to define a test region and separate the contacts, and capacitive contacts to conduct four-point measurements. To characterize a piece or a region on a wafer hosting a buried conducting layer, one brings the probe onto the sample, conducts the electrical measurements with the capacitive contacts, and removes the probe. The sample remains untouched and can be reused. The contactless probe should provide a fast and nondestructive way of semiconductor characterization.</p>