Treasures @ UT Dallas
Not a member yet
7697 research outputs found
Sort by
Superconductivity, Magnetism and Physical Properties of New Types of Low-dimensional Materials
No full textThis dissertation focuses on the physical property studies of a broad range of new low dimensional
compounds discovered in our research lab. These new materials structurally have two-dimensional
(2D) or quasi-2D features with Van der Waals interaction between layers, and display various
interesting physical properties such as superconductivity, magnetic order, anomalous Hall effect,
transport anomalies, and linear temperature dependent resistivity behavior.
Chapter 1 & 2 we introduce physics background and various experimental techniques involved in
this dissertation, including the different synthetic methods, X-ray diffraction, magnetic, electrical
transport studies.
Chapter 3 we develop a unique soft chemical intercalation route where both alkali metals (Li or
Na) and polar organic solvents could be cointercalated into the layered materials and significantly
change their electronic properties. Through this new method and using transition metal
dichalcogenide SnSe2 as prototype, we discovered a series of new superconductors with Tc up to
~8 K. These Tcs are significantly higher than those with only alkali metal doping or electrostatic
gating, suggesting these polar organic species are playing an important role for the enhanced
superconductivity in this system.
Chapter 4 we report several new layered compounds in the Nb-Fe-Te system. Two distinct α-
NbFeTe2 and β-NbFeTe2 are synthesized through different temperature route, with completely
different electrical and magnetic properties. In addition, some new phases are discovered with
different structures under different stoichiometry and growth conditions. Their structures and
physical properties are discussed in the dissertation.
Chapter 5 we present three new phases in the Ba-Cu-Pn (Pn = P, As) system. Through carefully
controlled growth conditions, we discover a new polymorphic β-BaCu2As2 phase as intergrowth
prototype of ThCr2Si2-type and CaBe2Ge2-type motifs, and two BaCu6Sn2Pn4-x (Pn = P, As) phase
with layered feature. The β-BaCu2As2 phase becomes superconducting under pressure with a
maximum Tc up to 7.2 K at 27.6 GPa, representing as the first Cu-based pnictide superconductor
under high pressure.
Chapter 6 we show structural diversity in the Ba-Ag-Pn (Pn = As, Sb, Bi) system where more than
six new phases are discovered with different stacking orders of ThCr2Si2 and CaBe2Ge2-type
structure motif. Most of these new phases display linear temperature dependent resistivity across
a large temperature range, and their exact physics origin are still under investigation. We have also
extended this study to the A-Ag-Sb (A = Alkali metal) system and find a few new compounds as
well.
Chapter 7 we report two new Pt-based chalcogenide compounds, one non-stoichiometric
Zr6.5Pt6Se19 and another layered BaPt4Se6. The Zr6.5Pt6Se19 is a new structure type (oC68) with
quasi-1D and quasi-2D structure feature, which is the first Pt-based ternary chalcogenide with
group 4 elements. The BaPt4Se6 adopts sesqui-selenide Pt2Se3 layer with mix-valence Pt oxidation
states, and has two distinct transport anomalies: a resistivity crossover, mimic to the metal-
insulator (M-I) transition at ~150 K, and a resistivity plateau at temperatures below 10 K.
Chapter 8 we report some several miscellaneous projects where crystal growth and bulk synthesis
of several materials are discussed while the full characterization of the materials are to be
investigated due to the time and instrument breakdown limitation at the end of this dissertation
work
Efficient Fair Learning With Subset Selection
No full textFairness in Machine learning has become crucial nowadays since most of the scenarios involve
data having a significant impact on people’s lives. The main purpose of this paper is to introduce
a fair efficient learning technique with subset selection. Secondly, machine learning
models are data-hungry. Training the state-of-the-art on large datasets requires a significant
amount of computation resources and time. Thirdly, most of the existing techniques
deal with ample changes either in model processing, data pre-processing, or post-processing
making it difficult to adopt in real-time applications. In this approach, we address these
issues by introducing a fair efficient machine-learning strategy with subset selection. The
approach follows the GLISTER strategy, a mixed continuous and discrete bilevel optimization
approach to perform data subset selection of the training data, by keeping the inner
optimizer a standard training algorithm and incorporating iterative processes to select a
subset in the outer optimizer. The strategy iteratively selects a subset to achieve our goal
of improving fairness without majorly sacrificing accuracy. The paper mainly focuses on
three significant fairness metrics - demographic parity difference, equalized odds difference,
and equal opportunity difference. Experiments are conducted on several real-time different
domain datasets and have seen a comparable and better performance against the other fair
learning and data subset selection techniques
Coalition Politics and Foreign Direct Investments: a Sub-national Study of India
No full textWhy do some states attract more Foreign Direct Investments (FDI) than others? Existing literature
on FDI is majorly focused on analyzing cross-national variation of FDI inflows and quantifying
the return on investments by means of economic and international trade causal mechanisms.
However, research so far has not adequately addressed the effect of domestic politics and inter-
governmental interactions on FDI inflows at a sub-national level. This dissertation addresses this
gap by exploring the effects of political structural differences between state and central coalition
governments on foreign direct investments in India. I argue that ideological and political
differences between state and central governments have a significant effect on FDI flows to states.
I hypothesize that higher the political differences, lesser the FDI inflows. Using a mixed methods
approach to analyzing FDI inflows between 2000-2019 to 28 states of India, I provide evidence
that that politically affiliated states attract relatively more FDI inflows in comparison to states
ruled by opposition parties or coalition partners. The findings highlight how several political and
regional factors cumulatively play a role in determining the level of FDI inflows as opposed to
these factors being considered independently and when aggregated at a national level
Bank's Industry Diversification and Debt Contracting
No full textI examine how industrial diversification of commercial loans influences banks’ lending terms
and borrower clientele. Using individual loan data from Dealscan, I derive an industry di-
versification measure in the spirit of the Herfindahl index. For loan terms, results suggest
that industry diversification is associated with lower spreads, less collateral, longer matu-
rities, and larger loans. Further, industry diversification is related to tighter covenants,
with a higher likelihood of performance-based covenants violations, a lower likelihood of
capital-based covenants violations, a higher likelihood of performance pricing provisions,
and a higher number of negative covenants. Analyzing banks’ loan portfolios reveal that
the average borrower risk of diversified banks portfolio is not statistically different from
non-diversified banks, there is greater diversity in the borrowers’ risk profiles. Additionally,
diversified banks’ portfolios have higher dispersion in the spreads, which is consistent with
diverse borrowers
Voltage Reconfigurable Bio-inspired Devices
No full textMoore’s law of transistor scaling predicts that due to shrinking transistor dimensions and other
improvements, the number density of transistors in an integrated circuit would double every two
years, resulting in a higher processing speed and performance. However, it is very challenging to
continue to reduce the physical dimensions of the transistors. Miniaturization of electronics is very
desirable as it will enable faster computational speed and better performance and advance the field
of quantum devices, biomedical sensors, and photonics. The future of miniaturized electronics may
utilize novel single molecular semiconductors. Molecular electronics may present several unique
advantages due to the broad structural variability and small size of molecules, enabling high
number density.
The intrinsic electrical properties of the single molecules can be studied by attaching them to the
electrodes separated by nanometer-scale gaps. Nanogaps facilitate the research to explore charge
transport mechanisms and other quantum effects at the nanoscale level. Electron beam lithography
(EBL) is a highly precise and reliable technique to construct arrayed nanogaps, and methods have
been developed to achieve sub-10 nm channel widths. However, many published approaches are
limited in throughput and universality due to multiple lithography steps or specialized techniques.
This work describes the fabrication of nanogap arrays with nanogap spacing from 5-40 nm
utilizing a simple, single EBL step, continuous dose, and direct feature write method. This method
involves optimizing the focusing using sputtered copper to reduce charging and latex nanoparticles
to facilitate coarse focus. In addition, cold lithography was utilized to preserve the nanoscale
features patterned in the resist. Overall, this protocol enables the straightforward production of
electrode nanogaps to facilitate the study of nanoscale materials and effects.
A bioinspired reconfigurable molecular electronic material, alloxazine-modified DNA was
investigated as a redox-active switch of hydrogen bonding. DNA is promising for nanoscale
applications due to its ability to self-assemble into arbitrary nanoscale shapes. An electrical switch
of DNA structure through modification of hydrogen bonds would bring rationally-controlled
functionality to these constructs. In this project, alloxazine DNA base surrogates were synthesized
and incorporated into DNA duplexes as a surrogate base functional as a redox-active switch of
hydrogen bonding. Thiolated duplexes were self-assembled onto multiplexed gold electrodes and
probed electrochemically. Cyclic voltammetry and square wave voltammetry revealed a redox
peak near −0.32 V vs. Ag/AgCl reference, corresponding to the reduction of the alloxazine moiety.
Alternating between alloxazine oxidizing and reducing conditions modulated the SWV peak in a
manner consistent with the formation and loss of hydrogen bonding, respectively, which disrupts
the base pair stacking and charge transport efficiency of the DNA. These results support the
assertion that alloxazine is a redox-active switch of hydrogen bonding, useful in controlling DNA
and bioinspired assemblies
Amorphous Silicon Carbide Neural Interface Devices
No full textMicroelectrode arrays (MEAs) are used in acquisition of neural activity for restoring
communication and controlling movement of paralyzed or prosthetic limbs. In addition,
neuroscientists need reliable recording probes, which can provide large numbers of unit activity
with high signal-to-noise ratio (SNR) over long periods. Compared to brain surface electrodes,
penetrating intracortical MEAs offer higher resolution of neural activity at the scale of single
neurons to enhance neural signal decoding. Of significant interest to neural engineers is the
development of MEAs with a high number of recording sites, reduced foreign body response,
longer shanks from 2 mm to 1 cm for shallow and deep recording, respectively. Laminar silicon-
based probes partially overcome the limitation of the high number of recording sites by increasing
the number of recording channels along the shank to allow recording from different depths along
one cortical column. Despite the success in increasing the number of recording channels, current
silicon-based intracortical MEAs, exhibit degraded performance 3-6 months after implantation. To
improve the longevity of MEAs, our group is investigating the use of a novel material, amorphous
silicon carbide (a-SiC), as a substrate for neural probe fabrication. This dielectric material has the
required mechanical and electrical properties (high stiffness and resistivity), chemical inertness,
and compatible with thin-film fabrication methods for MEA applications. Using a-SiC, we have
developed ultramicroelectrode arrays (UMEAs) with 16 to 32 electrode sites with scalable
shank dimensions ranging from 500 μm to 2 mm for shallow and up to 10 mm for deep recording.
We hypothesized that a-SiC based MEAs will provide electrochemical stability and enable reliable
neural recording and stimulation. Our ultimate goal was to develop high density (a-SiC)-based
UMEAs (increased number of electrode channels per unit area) and high spatial selectivity for
neural recording and stimulation while maintaining ultramicroelectrode dimensions. This
dissertation addresses the fabrication and development of these devices
Localization and Sensing Solutions for Smart IoT Systems
No full textAs more devices become interconnected, demands for Internet-of-Things-based (IoT-based)
ranging, localization, and sensing solutions are increasing. These solutions need to achieve
high accuracy, low complexity, robustness to environment conditions, and the ability to
be deployed on low-cost or common wireless infrastructure. Specifically for localization,
both active (device-to-device) and passive (device-to-target) applications are of interest.
Candidate technologies include ultrawideband (UWB), Bluetooth Low Energy (BLE), WiFi,
and millimeter-wave (mmWave) radar. Among these technologies, only BLE and WiFi
support are ubiquitous in modern smartphones, making them prime technologies for active
ranging and localization. On the other hand, mmWave frequency-modulated continuous-
wave (FMCW) radar solutions are attractive for passive localization and sensing due to
their ability to provide high-resolution information in multiple domains, including range,
direction of arrival (DoA), and velocity.
Towards BLE, WiFi, and mmWave radar localization and sensing solutions for IoT applica-
tions, we propose novel algorithms that leverage existing standards and devices. For BLE
ranging, we propose a data-driven support vector regression (SVR) approach that overcomes
challenges due to many closely spaced multipath components (MPCs), a single or low num-
ber of snapshots, and model imperfections that arise in practical scenarios. Evaluating our
SVR-based method on real-world measurements, we achieve decimeter-level accuracy with
single-antenna devices, whereas Multiple Signal Classification (MUSIC), a popular model-
based technique, requires multiple antennas to obtain comparable performance. Towards
WiFi-based ranging and localization, we build on the recently released IEEE 802.11az Next
Generation Positioning standard. To overcome resolution limitations imposed by the single-
channel bandwidth of operation in WiFi devices, we propose a phase-coherent multi-channel
(PCMC) approach that overcomes local oscillator phase offsets and time offsets per chan-
nel. Using 16 20 MHz channels and our PCMC technique, we demonstrate centimeter-level
ranging accuracy across multiple indoor environments. In terms of passive sensing using
mmWave FMCW radar, we develop novel deep learning algorithms targeted for in-vehicle
sensing applications. Our radar sensing methods overcome limitations in angular resolution,
ambient reflections, and multipath reflections that impair sensing accuracy. Evaluating our
sensing approach on participants unseen by the model during training and validation, we
achieve considerable accuracy for occupant seat-by-seat localization and classification of the
occupant as baby, child, or adult
Plasmonic-driven Nanoparticle Heating: From Photon-electron Transport to Localized Heating for Protein Inactivation
No full textPlasmonic nanoparticles (PNPs), with their excellent optical properties, chemical stability, and
biocompatibility, play essential roles across nanotechnology, biomedicine, energy, and chemistry.
When exposed to laser radiation, these nanoparticles efficiently absorb and convert photon energy
into heat—coined as plasmonic heating—rendering them ideal nanoscale heat sources. Leveraging
the broad range of laser excitations, PNPs induce a range of physical and biological responses.
This dissertation focuses on fundamental questions on the plasmonic heating for PNPs from the
dynamics of photon-electron-phonon transport during PNP heating, to the transition from localized
to bulk heating, with applications on photoacoustic imaging contrast, protein inactivation, and
theragnostic nanobubble generation.
Firstly, we investigated the photon-heat energy conversion with silica-coated gold nanoparticle
(AuNP). Our experimental study suggests 4-fold amplification in photoacoustic signals with <5
nm silica coating and picosecond laser excitation. Our theoretical model demonstrates a significant
enhancement of AuNP photon absorption with a silica-coating under picosecond pulsed laser
driven by competing effects of interfacial electron-phonon coupling and transient absorption,
resulting in amplified photoacoustic emission. This study provides new physical insights into the
amplified photoacoustic contrast with silica-gold core-shell nanoparticles.
Secondly, we investigated how an asymmetric thermal conductance at the particle-water interface
on the JANUS nanoparticle (JNP) influences the heating. We found that JNP-induced temperature
contrast, defined as the ratio of temperature increase in the surrounding water, shows a substantial
size, polar angle, and heating duration dependence. These results bring new insights into JNP
heating and provide new perspectives for potential applications.
Thirdly, we systematically analyzed how temperature evolves over time and space during the
transient heating process of NP arrays. We derived analytical solutions of temperature functions
for 2D, 3D, and spherical NP array heating and developed dimensionless parameters to
characterize the transition from localized to bulk heating. This study advances the understanding
of nanomaterials heating and guides the design of innovative approaches for NP heating.
Furthermore, we systematically investigated nanoparticle array heating-induced protein
inactivation/activation. We found that localized heating may lead to targeted protein denaturation;
however, nanoparticle heating does not lead to nanoscale selective TRPV1 channel activation. The
heating duration, NP concentration, and heating power are primary factors that define boundaries
for targeted protein denaturation. This work boosts our understanding of NP heating under realistic
physical constraints and provides guidance to customize biomedical platforms with NP heating.
Finally, we examined how the AuNP cluster affects plasmonic nanobubble (PNB) generation. In
this study, we fabricated an AuNP cluster by synthesizing Qβ-AuNP composites and
experimentally demonstrated a 5-fold enhancement of PNB generation with Qβ-AuNP
composites. Computational modeling showed that bulk heating in the Qβ-AuNP composites is a
key factor for enhanced PNB generation. This study proves the concept of enhanced PNB
generation with Qβ-AuNP composites and provides outlooks for applications in theragnostic
nanobubble generation.
In summary, PNP heating is versatile and induces various physical responses. Our studies advance
the fundamental understanding of the physical processes during PNP heating and provide new
insights into biomedical applications in photoacoustic imaging contrast, protein inactivation, and
theragnostic nanobubble generation
Distributed Integrated Sensing and Communications Systems
No full textThe continuous scaling up of the carrier frequencies and deployment of wireless communications has propelled the spectrum regulators to allow the use of the spectrum traditionally
reserved for radar (sensing) applications for commercial communications systems. As a
result, in recent years, integrated sensing and communications (ISAC) has emerged as a
promising technology to mitigate spectral congestion and efficiently utilize radio resources.
Prior research on ISAC was largely limited to colocated or centralized systems, where communications, radar, or both employed transmit/receive units placed close to each other.
However, next-generation wireless networks are envisaged to deploy a decentralized resource
management infrastructure and edge-device-centered network paradigms. Similarly, widely
distributed radars are gaining widespread usage because they offer the advantages of spatial
diversity, improved detection of stealth targets, and joint processing.
To this end, we investigate the design and performance evaluation of distributed ISAC systems. We consider a general system comprising a widely distributed multiple-input multiple-
output (MIMO) radar that operates in the same spectrum as a distributed MIMO communications network. A major challenge to co-design such an ISAC system is a unified performance metric for both communications and radar. We address this problem by proposing
a compounded weighted sum of mutual information as an objective to obtain optimized
waveforms, precoders, beamformers, and receive filters jointly for distributed sensing and
communications. Our design problem also includes various practical constraints such as link
budget, quality of service, and peak-to-average power ratios.
We then extend our formulation to in-band full-duplex (IBFD) distributed ISAC to enable
simultaneous uplink, downlink, and radar sensing transmissions. Conventional communications systems are based on either half-duplex (HD) or out-of-band full-duplex transmission for
low-complexity transceiver designs leading to reduced spectral efficiency. On the other hand,
IBFD enables concurrent transmission and reception in a single time/frequency channel to
potentially double the attainable spectral efficiency and throughput and reduce latency.
For distributed wireless systems, synchronization in time and space is a major concern.
The clocks at the radar and communications transmitters are synchronized both offline and
periodically. We use the feedback of the base station via pilot symbols to provide the radar
receivers with the clock times of uplink user equipment. Often, distributed beamforming
(DB) is employed to achieve the desired signal-to-noise and reduce power assumption by
providing a coherent beamforming gain. Further, MIMO radars may also employ DB in the
form of distributed coherent systems, wherein accurate phase synchronization is required
to obtain coherent processing gain. In this research, we show that such synchronization
techniques may also be incorporated in distributed ISAC.
Further, we develop techniques for multi-target localization in a distributed ISAC. In general,
echoes from multiple targets have different time-of-arrivals at different receivers. We solve
this association problem for the ISAC system via a mixed-integer programming framework.
Finally, we devise low-complexity design techniques, which utilize the block-coordinate descent and Barzilai-Borwein algorithms, for the aforementioned scenarios to obtain optimal
design parameters for distributed ISAC simultaneously
Innovative Biosensing Strategies: Electrochemical Profiling of vWFA2 for Early Sepsis Detection
No full textIn the dynamic intersection of technologically advancing medicine and engineering, rapid
diagnostic methods are imperative for prognostic evaluation of critically ill - upon multimer
assembly patients presenting with clinical complications from a dysregulated host response.
Evaluation of biomolecular markers in inflammatory conditions, specifically sepsis, is of peak
interest in research to develop a highly rapid, robust, and effective diagnostic option. The
complexity of sepsis, a life-threatening inflammatory condition arising from the systemic
response to an infectious agent, presents a grave challenge in detection and treatment.
Following vascular injury, endothelial cells, strategically placed at the blood-tissue interface,
respond to inflammatory stimuli by releasing circulating co-factors, including Von Willebrand
Factor (vWF). Von Willebrand Factor-Like 2 Domain (vWFA2), the A2 domain of the
heterogenous multimeric glycoprotein, plays a central role in primary hemostasis by promoting
platelet adhesion to the subendothelial matrix of damaged vessels and protecting FVIII from
proteolytic degradation. vWFA2 responds to shear stress by balancing bleeding and clotting
mechanisms. Reduced functionality of vWFA2 leads to severe hemorrhagic consequences
following defective formation of a platelet-rich thrombi and fibrin network. The immunethrombotic
role of vWFA2 in sepsis is understudied due to the emphasis on its narrowed role in
thrombotic events, but it is critically important to understand the developing biological
relationship between both physiological processes. Elevated levels of vWF are seen in septic
conditions, thus the marker is a vital for disease severity assessment. Current gold-standard
diagnostic methods include blood cultures for accurate microbial diagnosis but is faced with
compromised sensitivity, prolonged processing time, and a large sample volume requirement.
This gap paves the way for development of a prompt and accurate diagnostic protocol,
particularly one that targets inflammatory biomarkers.
The goal of this research is to develop a biosensor device for rapid detection of biomarkers in
septic and inflammatory states. The transformative potential of sensors in revolutionizing
detection guides refinement of sensitive assays that addresses the shortcomings of conventional
detection methods. This label-free biomolecular assay is fabricated on a flexible hybrid electrode
surface and leverages electrochemical impedance spectroscopy (EIS) to measure the capacitive
change in impedance, unveiling the binding effects of the target, vWFA2, to the capture probe.
The device aims for high sensitivity and specificity in the targeted assay development of a wide
dynamic range of 500-32,000 pg/mL. These methods offer a promising solution in point-of-care
diagnostic methods, offering a high sensitivity, rapid response time and minimal biofluid volume
requirement. The vision behind this research stems from the desire to develop this
electrochemical sensor to detect disease states rapidly and accurately, thus creating a pivotal
prognostic tool in sepsis treatment, and ultimately mitigating severe mortality and morbidity