12023 research outputs found
Sort by
First-Principles Calculations of Magnetotransport and Electron-Phonon Interactions in Semiconductors and Topological Materials
Get PDFUnderstanding and predicting electron transport in novel materials is crucial to develop practical applications and accelerate materials discovery. Electron-phonon (e-ph) interactions are a key source of electron scattering and therefore play a dominant role in limiting electron transport under applied external fields. These interactions and the resulting phonon-limited charge transport can be calculated very accurately using ab-initio methods based on the semiclassical Boltzmann transport equation (BTE), where electron and phonon properties are obtained using density functional theory (DFT) and density functional perturbation theory (DFPT) techniques. Despite these advances, first-principles calculations of magnetotransport are still in their infancy, primarily due to technical challenges associated with solving the BTE in the presence of a magnetic field. Additionally, calculations of electrical charge transport and magnetotransport in topological materials are lacking because of various technical challenges, including computational cost and the absence of a unified formalism combining electron scattering and band topology in the BTE. In this work, we develop a framework that incorporates these effects into the BTE to compute charge transport, magnetotransport and topological transport regimes in several classes of conventional and quantum materials. Our magnetotransport calculations achieve excellent agreement with experiments, and we uncover an interplay of strong e-ph interactions and magnetic fields in graphene through a microsopic analysis of steady-state electron distributions. As a first step toward including band topology, we compute e-ph interactions and charge transport in the Dirac semimetal Na₃Bi and find that specific two-dimensional phonons control charge transport near room temperature. These lattice vibrations induce a dynamic phase transition to a Weyl semimetal, providing a platform for ultrafast control of dynamical phases in Na₃Bi. Expanding into more advanced phenomena, we incorporate the electron Berry curvature in the BTE formalism and study topological transport effects such as the chiral anomaly and nonlinear Hall effect (NLHE). Our calculations provide an accurate quantitative framework and demonstrate the importance of e-ph interactions in accurately describing topological transport in quantum materials. Lastly, we compute e-ph interactions in a novel correlated metal, RuO₂ which has been widely studied for its unconventional magnetism. We uncover various interesting properties such as phonon softening, strong e-ph band renormalization and a high superconducting Tc upon application of strain in RuO₂. Finally, we show a method to significantly accelerate all these calculations by compressing the matrices representing e-ph interactions. In summary, this work expands the scope of first-principles transport calculations to include magnetic fields and band topology. This enables future studies of electron dynamics in broad classes of novel quantum materials
Location, Location, Location: Insights from Spatially-Resolved Observations of Marine Seep Carbonate Ecosystems and Carbonaceous Chondrite Surfaces
Get PDFSpatially heterogeneous, multi-component systems are prevalent topics of study in geobiology and planetary science. However, previous studies of these systems often represent limited measurements that abstract or separate the sample from its localized context, thereby obscuring or precluding insights into the drivers ultimately shaping these systems. This challenge motivates the work presented in this thesis, where we provide an extensive and spatially-resolved examination of two complex, heterogeneous systems in geobiology and planetary science: marine seep carbonates and carbonaceous chondrite surfaces, respectively. In marine seep systems worldwide, seep carbonates are a mineral byproduct of a microbial metabolism (the anaerobic oxidation of methane, or AOM) and can continue hosting metabolically active microbial communities, including methane-oxidizing microbes. However, much of our understanding of these endolithic microbial communities stems from bulk, centimeter-scale evaluations of microbial identity and/or metabolic activity across a limited number of samples. As such, the range of structural and environmental conditions that ultimately shape the degree and extent of microbial activity in seep carbonates, including AOM, remains relatively under-constrained. To address this gap, Chapters 1-3 investigate carbonate-hosted microbial communities at a methane seep site in Santa Monica. In Chapter 1, we explore carbonate ‘nodules’ from methane seep sediments at and below the sulfate-methane transition zone (SMTZ), analyzing their mineral composition, internal structures, and hosted microbial communities compared to their host sediment communities and porewater chemistry. We also discuss key implications of the connectivity of seep sediments to nodules over geologic timescales and the preservation of microbial ‘thumbprints’. Chapter 2 describes rare tripartite associations between two groups of anaerobic methanotrophic archaea (ANME-1 and ANME-2) and a bacterial partner within seep carbonate crusts and other substrates at the seafloor, with implications towards understudied diversity in the syntrophic interactions governing AOM beyond seep carbonates. Chapter 3 examines the impact of seep carbonate internal structure on endolithic communities from various carbonate crusts, revealing similarities and differences between surface and interior communities that may reflect the importance of pore networks in maintaining favorable local environments. In Chapter 4, we pivot to an extensive analysis of spectra from carbonaceous chondrite surfaces. Carbonaceous chondrites (CCs) are a group of meteorites that represent the oldest materials in the solar system, whose mineralogy preserves a record of early alteration processes thought to be shared with certain asteroids. However, most studies connecting specific CCs to specific asteroids have relied on spectroscopic measurements of bulk powder CCs, which are spatially unresolved and destroy textures, thereby hindering tying shared spectral features to particular phases, petrologic contexts, and alteration histories. As such, Chapter 4 presents an analysis of CCs measured using microimaging hyperspectral visible-and-shortwave-infrared (VSWIR) spectroscopy, where we capture chondrite surfaces features at high spatial resolution. We also compare CC spectral features with asteroids using the Expanded Bus-DeMeo taxonomy, which provides a systematic framework to examine and identify shared drivers of spectral diversity within this spectral range, including Fe-bearing minerals from both original and terrestrial alteration processes. Together, these studies emphasize the importance of spatially-resolved sampling across disciplines, specifically in geobiology and planetary science, thereby capturing and highlighting the heterogenous nature of key systems in these fields and bettering our understanding of the factors shaping them
Smart Masks for in situ Exhaled Breath Condensate Harvesting and Analysis
Get PDFWith the growing focus on personalized breath health management and early detection of chronic pulmonary diseases, there is an urgent demand for noninvasive wearable technologies capable of continuous breath molecular monitoring during daily activities. Existing respiratory monitoring systems remain limited to physical signal tracking and lack the capability for real-time biochemical analysis of exhaled biomarkers. To address this critical gap, we developed EBCare, a fully integrated smart mask platform for automated in situ analysis of exhaled breath condensate (EBC) biomarkers. The system combines tandem passive cooling strategies (hydrogel evaporation and radiative metamaterials) with bioinspired microfluidics to enable sustainable breath condensation and efficient sample transport under real-world conditions. A multiplexed electrochemical sensor array functionalized with nanoengineered interfaces achieves selective detection of key inflammatory markers (nitrite, pH) and metabolic indicators (ammonia, alcohol), while an embedded wireless module facilitates continuous data transmission. System validation through controlled breathing experiments and field trials demonstrates reliable operation across diverse environments (10-35°C, 30-80% humidity). Clinical evaluations involving healthy subjects, COPD/asthma patients, and post-COVID cohorts reveal EBCare's ability to dynamically track airway inflammation patterns and metabolic shifts during daily tasks. This wearable EBC analysis platform bridges the gap between laboratory-based breath testing and real-world respiratory monitoring, offering a scalable solution for home-based management of chronic respiratory conditions and post-infection recovery tracking. The modular design and automated operation framework further support future expansion to monitor airborne pathogens and systemic metabolic disease biomarkers through exhaled breath
New Structural and Electronic Degrees of Freedom in Epitaxial Square-Net Materials
Get PDFMaterials belonging to the "square-net" (SN) family of crystal structures share the structural motif of highly conducting, 2D square-planar sheets sandwiched between complex spacer layers. Materials in this class have attracted attention for their diverse array of electronic properties such as topological, magnetic, charge/spin density wave (CDW/SDW) and superconducting ground states. Familiar examples include the cuprate and pnictide superconductors, the rare-earth tellurides and the Dirac semimetals such as ZrSiS. In this thesis we exploit the abilities of molecular beam epitaxy to synthesize and study ultra-thin films of the SN compounds and uncover several unexpected behaviors.
The first compound we explore is DyTe₂, a member of the telluride family of SN materials known for their charge density wave ground states. We begin by describing the methods to fabricate epitaxial films using MBE. The high crystalline quality allows for characterization of subtle superlattice modulations with X-ray diffraction. Combinations of this experimental data with theoretical calculations reveal the origin of this superlattice to be an ordering of Te vacancies driven by Fermi-surface nesting.
We then turn to the related compound LaSb₂. This material is thought to undergo a CDW transition that can be suppressed under pressure and replaced by a superconducting ground state. To our surprise, thin films of LaSb2 adopt a crystal structure distinct from that of the bulk crystals. We characterize this new structure comprehensively and find that concomitant with this new structure is an enhancement of superconducting Tc relative to the bulk.
Finally, we exploit this enhanced T꜀ to observe magnetic field-induced superconductivity in ultra-thin LaSb2 doped with magnetic Ce dopants. This is the result of the unique robustness of the material to application of a parallel magnetic field. The combination of strong spin orbit coupling and reduced dimensionality allows the magnetic field to polarize paramagnetic spins, thereby reducing their deleterious impact on T꜀, before the field itself destroys superconductivity. This allows a superconducting ground state to be induced from an otherwise normal metal ground state at T = 0.
The results of this thesis highlight the unique degrees of freedom that can be accessed via epitaxial growth of single crystalline films of quantum materials.</p
A Path Towards Wearable Affective General Intelligence
No full textArtificial intelligence continues to support our daily decision-making tasks yet remains disconnected from our dynamic emotions driving these behaviors. Wearable technologies can supplement interactions with continuous emotion biofeedback, but existing models struggle to generalize across emerging biomarkers, platforms, and affective expressions. Here, we introduce a meta-analysis into embedding concurrent fragmented biosignals across 15 medical platforms, spanning five bodily locations, within a single profile that enables efficient and generalizable downstream affective analysis. We achieved this through a Lie manifold neural architecture that simultaneously reconstructs over 118,000 missing biometric details in 205 biomarkers and accurately forecasts 100 affective states across cohorts, questionnaires, and activities. We validated this framework across five datasets to propose a new skin-conformal, soft bioelectronic, affective computing platform that demonstrates closed-loop emotion modulation within thermal, audio, and visual interventions delivered through virtual, holographic, and conversational agents. Our framework establishes a new foundational bidirectional architecture for scalable, interpretable, and emotionally intelligent human-computer interactions
Acoustic Radiation in Hypersonic Turbulent Boundary Layers: Deciphering Linear Dynamics
No full textThis thesis is primarily concerned with hypersonic turbulent boundary layers and the unique features – present in them. This problem is studied in three levels of varying fidelity – by means of linear resolvent analysis, a blended resolvent estimation approach, and direct decomposition of a temporally-resolved dataset. This thesis then explores three complementary research directions: (i) quantification of how streamwise development influences acoustic radiation across various parameter regimes, (ii) development of a forcing model that enables acoustic radiation estimation using only near-wall measurements, and (iii) evaluation of these findings through comparison with data-driven analysis techniques.
First, the resolvent analysis is performed on a turbulent hypersonic streamwise developing mean profile. It is shown that these (acoustically radiating) streamwise developing resolvent modes may be effectively modeled using resolvent modes around an assumed-parallel mean profile. Then this model is used to investigate the impact of streamwise development on acoustic radiation for varying bulk parameters.
Second, the modeling of acoustic radiation from near-wall information is tackled. To achieve this, resolvent based estimation (RBE) is leveraged along with a small number of near-wall measurements. It is shown that RBE alone is insufficient to accurately predict the freestream power spectral density. Resolvent analysis around a streamwise developing mean profile is then analyzed by performing a Helmholtz decomposition, where it is shown that the solenoidal part of the resolvent forcing is primarily responsible for the linear amplification. This observation is used to develop an approximate forcing CSD method, which filters out any dilatational forcing, to supplement RBE. Using the approximate forcing with RBE shows significantly improved estimation of the freestream PSD.
Finally, spectral proper orthogonal decomposition (SPOD) is applied to a 3-D temporally-resolved dataset resulting from a direct numerical simulation of a hypersonic streamwise developing turbulent boundary layer. It is shown that the SPOD of the fluctuations around a streamwise developing mean extracts modes with a constant streamwise wavenumber and shows high-rank behavior. By further transforming the data in the streamwise direction, an SPOD of the fluctuations around a 1-D mean profile uncovers low-rank behavior and similar structures are seen between the resolvent and SPOD modes.</p
From Melting Dynamics to Medical Diagnostics: Studies in Geochemical Kinetics
Get PDFThis thesis investigates geochemical kinetics across different subfields, from isotope metallomics in the human body to melting dynamics in igneous petrology. Chapters II and III explore the chemical complexities of rapid mineral melting in igneous systems. An experimental-computational approach is used, with the experiments providing data that help calibrate the numerical model. This integrated strategy contributes to a comprehensive understanding of the kinetics of melting that could not be captured by either method alone. Chapter II outlines the experimental work, which includes both equilibrium and kinetic melting experiments performed on the ubiquitous igneous mineral series plagioclase. The kinetic experiments are designed to deliberately access a parameter space of disequilibrium behaviors rarely studied experimentally yet likely to be relevant in various natural settings where systems evolve too quickly to follow the predictions of equilibrium theory. Quantitative and qualitative analyses of the recovered experimental products allow us to observe unique textures and chemical gradients that arise from the interplay of thermal and chemical diffusion within the phases, coupled with phase boundary motion and associated surface reactions. Chapter III details the theory and computational methods used to develop a numerical model that describes chemical evolution of melt and crystal phases during two-component melting. Novel application of thermodynamic data is used to describe chemical behavior at the phase boundary, allowing for departure from traditional equilibrium assumptions. Results of the model bring us one step closer to the ultimate goal of understanding disequilibrium in multicomponent rock systems. Chapter IV investigates the kinetics of stable isotopes in biomedicine. Box modeling was used to simulate copper (Cu) stable isotope dynamics in the human body, allowing us to quantify the possible effects of various health conditions (e.g., cancer, liver disease) on isotopic compositions throughout different organs. In turn, we determine whether Cu isotopes can act as diagnostic or prognostic markers for certain diseases using detection by modern mass spectrometry and provide recommendations on their potential uses in the medical field
Software, Tools, and Methods Development for Single-Cell Transcriptomics
No full textAdvances in transcriptomics have transformed the study of gene expression, enabling a shift from low-throughput bulk RNA measurements to high-resolution, large-scale single-cell RNA-sequencing (scRNA-seq). This work refines existing methodologies and introduces new strategies for achieving precise, versatile, and scalable transcriptomic analyses across a broad spectrum of assays and biological contexts.
On the computational front, this dissertation introduces new methods for adaptable preprocessing of sequencing reads, enabling the handling of very complex read structures. It refines existing strategies for efficiently querying large-scale transcriptomic datasets and enhances approaches for quantifying nascent and mature RNA species. A general framework is introduced for discovering and organizing biologically informative sequences directly from raw sequencing data, facilitating the detection of sample-specific or condition-specific variation. On the experimental front, a novel single-cell RNA sequencing method is presented that is cost-effective, open source, and scalable, supporting large-scale studies with substantial cell numbers and high per-cell resolution.
These developments collectively expand the toolkit for transcriptomics, enabling more efficient and comprehensive exploration of RNA biology.</p
Make the Most of What You Have: Resource-Efficient Randomized Algorithms for Matrix Computations
No full textIn recent years, randomized algorithms have established themselves as fundamental tools in computational linear algebra, with applications in scientific computing, machine learning, and quantum information science. Many randomized matrix algorithms proceed by first collecting information about a matrix and then processing that data to perform some computational task. This thesis addresses the following question: How can one design algorithms that use this information as efficiently as possible, reliably achieving the greatest possible speed and accuracy for a limited data budget? This question is timely, as randomized algorithms are increasingly being deployed in production software and in applications where accuracy and reliability is critical.
The first part of this thesis focuses on the problem of low-rank approximation for positive-semidefinite matrices, motivated by applications to accelerating kernel and Gaussian process machine learning methods. Here, the goal is to compute an accurate approximation to a matrix after accessing as few entries of the matrix as possible. This part of the thesis explores the randomly pivoted Cholesky (RPCholesky) algorithm for this task, which achieves a level of speed and reliability greater than other methods for the same problem.
The second part of this thesis considers the task of estimating attributes of an implicit matrix accessible only by matrix–vector products, motivated by applications in quantum physics, network science, and machine learning. This thesis describes the leave-one-out approach to developing matrix attribute estimation algorithms, and develops optimized trace, diagonal, and row-norm estimation algorithms for this computational model.
The third part of this thesis considers randomized algorithms for overdetermined linear least squares problems, which arise in statistics and machine learning. Randomized algorithms for linear-least squares problems are asymptotically faster than any known deterministic algorithm, but recent work of [Meier et al., SIMAX '24] raised questions about the accuracy of these methods when implemented in floating point arithmetic. This thesis shows these issues are resolvable by developing fast randomized least-squares problem achieving backward stability, the gold-standard accuracy and stability guarantee for a numerical algorithm.</p
Functional Stimulated Raman Imaging for Quantitative Cell Biology with Small Bioorthogonal Tags
Get PDFThe development of imaging techniques, particularly optical imaging, has significantly advanced the field of cell biology. Compared to conventional fluorescence imaging, vibrational imaging leverages the intrinsic chemical bond information of molecules, providing multidimensional insights into molecular structures and local environments. So far, stimulated Raman scattering (SRS) microscopy has emerged as a powerful tool for quantitative measurements in biological research. It overcomes several fundamental limitations associated with fluorescence-based techniques and offers high spatial and temporal resolution along with excellent compatibility for live-cell imaging.
In this thesis, I mainly focus on utilizing small bioorthogonal vibrational tags for quantitative investigations in cell biology. These tags, such as alkyne and carbon-deuterium (C-D) bonds, are absent in endogenous biomolecules and smaller than 1 nm in size, enabling minimally perturbative labeling with high molecular specificity. Another key advantage is that the SRS signal scales linearly with bond concentrations, allowing for robust and quantitative analysis. Moreover, because their vibrational modes distinctly reside in the cellular silent region (1800-2700 cm⁻¹), they provide a high signal-to-background ratio, making them particularly well-suited for quantitative applications in complicated cellular environments.
In Chapter 2, we explored the potential of alkyne-tagged probes to serve as environment-sensitive vibrational sensors, extending their utility beyond imaging markers. We developed a generalizable sensing platform based on hydrogen-deuterium exchange (HDX) at terminal alkynes. This subtle isotopic substitution induces a detectable shift in the alkyne vibrational frequency, allowing for real-time monitoring of exchange kinetics. These kinetics, in turn, provide insight into the chemical structures and local environments. We conducted a comprehensive study of the HDX process through both theoretical analysis and experimental validation. This platform was further applied to detect structural changes in DNA and to indicate pH within live cells, demonstrating the broader applicability of alkyne-tagged Raman probes for local environmental sensing in complex biological systems.
In Chapter 3, we utilized deuterated glutamine to label and study polyglutamine (polyQ) aggregates, a pathological hallmark of Huntington’s disease, in neurons. Traditional imaging approaches typically rely on tagging with bulky fluorescent proteins such as EGFP, which can perturb aggregation behavior with their non-negligible sizes. Through deuterium labeling, we achieved EGFP-free imaging of polyQ aggregates, allowing for a more native characterization with live-cell compatibility. This strategy facilitated quantitative analysis of the aggregate composition and growth dynamics of polyQ aggregates in live neurons. Our results revealed significant variations in polyQ aggregates depending on cell types, subcellular localizations, aggregate sizes, and protein constructs. Notably, we identified a previously unknown type of nuclear aggregates, shedding light on the heterogeneity of polyQ pathology.
In Chapter 4, we applied deuterium-labeled small molecules to study neuronal metabolism and its dynamic interactions with neuronal activity. As neuronal firing requires high and tightly regulated metabolic input, it is critical to understand the coupling between neuronal activity and metabolism for elucidating brain function. Using deuterated glucose and fatty acids, we were able to track their downstream metabolites for metabolism studies with high spatial and temporal resolution via SRS microscopy. In parallel, we employed optogenetic stimulation through Channelrhodopsin to achieve precise control of neuronal activity and also used neurotransmitters for longer-term modulation. By correlating different states of neuronal activation with metabolic flux changes, we gained valuable insights into how neuronal activity dynamically regulated glucose and lipid metabolism, advancing our understanding of neuroenergetic mechanisms in live neurons.
Through these studies, I demonstrate that the integration between small bioorthogonal vibrational tags and the advanced vibrational imaging technique, SRS microscopy, can provide powerful, minimally invasive, and highly quantitative tools for tackling fundamental questions in cell biology with high spatial and temporal resolution.</p