Session D46: Undergraduate Research IV

Detecting COVID-19 via smartphone: Optofluidic cytometry and viral detection using LIF-LIDAR and Cherenkov Luminescence Imaging.

A variety of analytical techniques are used to test for viruses, including immunoassays, biochemical assays and PCR-based approaches. For distant locations, even if these technologies are more sensitive, the absence of adequate resources typically makes them impractical. I've developed a dynamic multi-channel module that detects SARS-CoV-2. The detection involved the use of QUASR., which also enables multiplexing targets in a single reaction. The device consists of three main components: heating module, assay reaction housing module, and an optical detection module, which houses a Cherenkov Detector and LIF-LIDAR component. Using the CoV-specific primers and without any sample preparation, the virus detection rate at 103 PFU/mL was found to be 100% specific. The smartphone-anchored system can be used for qualitative detection, thereby also making it easier for non-experts to use. The outputs are easily analysed using smartphone application which also enables sharing the results with experts in a different location promoting cross-validation. The module is also capable of flow cytometry and showed a dramatically higher specificity and sensitivity when compared to commercial hemocytometer.   

Computer Simulations of Early Aggregation of Intrinsically Disordered Protein on Nano-Structured Surfaces at Different Spatial and Time Scales

Early aggregation of intrinsically disordered protein (IDP), e.g., beta-amyloid (Aß), on the surfaces of neurons is responsible for the progression of Alzheimer's. However, the molecular details of forming toxic and partially-ordered Aß aggregates on the surface of phase-separated lipid nanodomains, a model of neuronal membrane, are still not available. Using both atomistic and coarse-grained (CG) multiscale MD simulations, we explore early aggregation events of initially disordered Aß on multicomponent, phase-separated lipid membranes at different spatial resolutions and microsecond time scales. Our lipid membranes consist of highly dynamic and heterogeneous liquid-ordered (Lo), liquid-disordered (Ld), and mixed Lo/Ld nano-domains.  Our work involves forward (atomistic-to-CG) and backward (CG-atomistic) mappings of protein/membrane complexes,  as well as, domain preference of protein binding, characterizations of the domain size, area per lipid, and bilayer thickness of the lipid nano-domains at both atomistic and CG resolutions. Our results will provide new insights into understanding the molecular interactions of IDP on nanostructured surfaces, and protein-induced membrane disruption mechanisms of Aß on neurons leading to the early progression of Alzheimer's disease.   

Coarse-Grained Molecular Dynamics Simulations of Molecular Interaction of Human Tau Protein Fragment with Neuronal Membranes

The membrane-binding domain of human tau (K18) belongs to a class of intrinsically disordered proteins (IDPs) that rapidly self-aggregates in solution and on various surfaces. It is believed that the self-aggregation of K18 on neuronal membrane surfaces is responsible for the early progression of Alzheimer's disease. Currently, the molecular details of protein-protein and protein-lipid interactions of K18 aggregates on neuronal membranes are unknown. Using coarse-grained MD simulations, we have investigated the molecular interactions of various sizes of K18 aggregates on phase-separated lipid nanodomains up to tens of microseconds. We have constructed neutral and negatively charged lipid domains containing phospholipid, glycolipid, and cholesterol that mimic both the inner and outer leaflets of neuronal membranes. From the simulation data, we have determined the binding preferences of K18 among the phase-separated liquid-ordered and liquid-disordered lipid domains, as well as the binding energies of protein-protein and protein-lipid interactions. This work improves our understanding of IDP interactions on nanostructured surfaces, as well as the future design of anti-aggregation drugs and treatment for neurodegenerative diseases.   

Role of time dependent single cell stiffness change in collective behavior of tumor invasion

Cell migration plays an essential role in wound healing, tumor invasion, and organ formation during embryo development. While the biochemical factors that support cell migration is comparatively well studied, the effect of physical forces on cell migration needs to be better understood. A physics-based description of cell migration could pave the way towards an improved means of controlling tumor invasion and metastasis. Using computational modeling of cell collectives, we investigate the migration of cancer cells in tumor spheroids during initial stages of tumor growth. Using a three-dimensional model incorporating cell birth, cell death and physical interactions based on cell stiffness and adhesion, we study the impact of time dependent single cell stiffness change on tumor growth and invasiveness. We track and compute how individual cell displacement impacts the temporal evolution of tumor size. We also quantify how spatial patterns in cell stiffness affects tumor spreading. We conclude that time dependent cancer cell stiffness, with enhanced cell stiffness prior to division and softening after division driven by actin build up and decay, leads to heightened cell dynamics and increased tumor size. In comparison, tumors with cells that do not soften in stiffness after cell division leads to more restricted cell dynamics and tumor size growth.    

Sub-10 nm Iron-Oxide Nanospheres and Nanocubes for Hyperthermia Therapy

Magnetic hyperthermia offers promising potential for supplementary cancer therapy. By applying an alternating magnetic field, magnetic nanoparticles generate heat that can raise the temperature of target cells to a point at which they begin to break down. In this research, nanospheres and nanocubes less than 10 nm in size were synthesized and characterized for use in hyperthermia treatment. It was found that these two shapes of superparamagnetic particles exhibit distinct properties below their blocking temperature—namely, the nanocubes display strong exchange bias not present in the nanospheres. This exchange bias, paired with an increase in shape anisotropy, resulted in a higher Specific Absorption Rate (SAR) for nanocubes compared to nanospheres. Finally, measurements on SAR in different mediums were carried out, showing that synthesized nanocubes are superior in hyperthermia treatments compared to nanospheres.   

Extracting the Global Characteristic Time from Partially Sampled Diffusion-Controlled Drug Release Profiles

Recently, Ignacio et al^1,2 have proposed to use a global (or integral) time ??^* to characterize drug release profiles from hydrogels and porous systems. In fact, it was shown that the time ??^* has remarkable mathematical properties for diffusion-controlled release systems. In practice, most release experiments give access to partial data (i.e., a finite number of data points within a finite time range). Here, we present a simple ”geometrical” method to calculate this integral time from partial release data and compare its performance to that of an alternative method based on empirical fitting functions. More precisely, we test these approaches for a uniformly loaded system where the initial total amount of drug is unknown. We show that fitting functions containing a single time scale (such as the Weibull function) are sensitive to the data sampling method and thus perform very poorly.

¹ Maxime Ignacio, Mykyta V. Chubynsky, Gary W. Slater, Physica A, 486, 2017

² M. Ignacio, G.W. Slater, Physica A, 567, 2021   

Withdrawn

Gold nanogratings support surface plasmon waves that can enhance fluorescence. Surface plasmon waves travel along metal surfaces due to the oscillation of conduction electrons in response to light. Förster resonance energy transfer (FRET) is the transfer of energy between a donor and acceptor fluorescent molecules and is strongly dependent on the donor/acceptor spacing. Use of FRET is limited by the Förster radius which is the donor/acceptor spacing where the transfer efficiency is 50%. Gold nanogratings are uniquely suited to enhance the efficiency of FRET because they support a wide range of surface plasmon wavelengths. By enhancing the energy transfer rate and efficiency, surface plasmon enhanced FRET can be utilized more broadly.  

Gratings are fabricated using a soft lithography method then coated with fluorescent molecules. The enhancement of FRET was measured as a function of surface plasmon wavelength by varying the angle of incidence of the excitation laser, and measuring fluorescence emitted as a function of detector angle with respect to the grating surface. We are currently building a setup to measure the lifetime of the fluorescent molecules using a pulsed laser. This measurement will provide a greater understanding of how surface plasmons enhance FRET processes.   

Using DNA as Scaffolding to Better Understand Surface Plasmon Enhancement of FRET

Our lab is utilizing customizable DNA to better understand the surface plasmon enhancement of fluorescence. Currently, we are investigating the enhancement of Förster resonance energy transfer (FRET) with surface plasmons excited on gold nanogratings. FRET is the transfer of energy between a donor and acceptor fluorescent molecule. Enhancements to the FRET efficiency from surface plasmon excitations depend on the spacing of the fluorescent molecules relative to each other and the surface of the grating. We developed a protocol to attach the donor and acceptor molecules to the grating surface using single-stranded, 20 monomer oligonucleotides. One strand contained a six-carbon thiol group on the terminal end, and a complementary strand contained a fluorescent Cyanine group (Cy3). Once annealed, the double-stranded DNA molecule has a thiol group on one end and the Cy3 molecule on the other. This double-stranded DNA is then deposited dropwise onto the gold surface along with a mercaptohexanol (MCH) spacer molecule. The MCH ensures that the DNA orients itself normal to the surface. This protocol allows us to more accurately position fluorescent molecules in such a way that surface plasmon FRET enhancement can be better understood   

Quantum bounds for discriminating mixed states generated by weak measurements and thermal noise

The problem of optimally discriminating between known non-orthogonal quantum states has many important applications in both quantum communications and quantum computation. However, non-orthogonal states cannot be discriminated perfectly, resulting in much work on finding the quantum bounds of discrimination with minimal error. Discrimination of pure states is well understood, but little research has been done on discriminating mixed states.  We compute quantum bounds for discriminating between mixed states that were prepared by a pure state mixed with thermal noise light in a channel versus the same pure state subject to weak measurement in a channel. We calculate the Helstrom bound for this discrimination problem when only one copy of the quantum state is available for measurement, and the quantum Chernoff bound, for the case when copies of the quantum state occupy many modes and can be measured individually or with a joint measurement.  These results have utility in disambiguating between an attacker in a quantum key distribution system with weak measurement capabilities versus thermal noise in the channel.    

Quantum Clocks and Path Integrals: An Intuitive Approach to Quantum Tunneling Time

The quantum tunneling time has been long studied both theoretically and experimentally. However, a consensus on the best way to conceptualize and to calculate the quantity has not been reached. For example, naïve methods using expectation values of position and momentum operators give unphysical negative values of tunneling time for certain wave packets. Other existing methods have various problems, such as affecting the dynamics of the system. Motivated by Page, Wootters, and Deutsch's approach to observable time, we employ the use of quantum clocks and a path integral method to study the problem of quantum transition and tunneling times. We envisage a qubit as it transitions between two quantum states coupled to a quantum clock with finite resolution. By assigning each classical path a probability and characteristic transition time by the quantum clock, we then perform a path sum to determine a characteristic transition time for the qubit. This intuitive method can also be generalized from transition time to tunneling time as the density of states of the observable becomes continuous.   

Probing Two-Level Fluctuators Using a Dynamically Decoupled Qubit

Two-level charge fluctuators (TLFs) plague all solid-state qubit devices, and identifying their switching rates and coupling strengths can shed light on their origin, paving the way to better device design and/or qubit working protocols to mitigate their adverse effects. This work introduces dynamical decoupling protocols that are designed to reconstruct charge fluctuator parameters. Using taylored sets of Mixed Concatenated Dynamically Decoupling sequences we show that the parameters of a  single TLF can be reconstructed with high fidelity over a wide range of coupling strengths. Our protocols are robust against substantial measurement errors and can be extended to consistently identify multiple TLFs, as long as their coupling strengths are within the same order of magnitude. These results suggest that our procedure could be a useful tool for identifying and characterizing TLFs in various solid-state qubit devices.    

Tuning Tunneling Resistance of Josephson Junctions for Precise Qubit Control

Superconducting qubits are leading candidates in the race to build a quantum computer that is capable of realizing computations beyond the reach of modern supercomputers. To build scalable systems of superconducting qubits, it is essential to achieve precise control of qubit frequencies to a level not yet reachable by fabrication. The most commonly employed method to set qubit frequency post-fabrication is by using flux-tunable qubits, but adding flux-bias lines occupies chip space and introduces additional noise channels. In this talk I present progress towards using controlled electromigration to adjust the tunneling resistance of Josephson junctions and tune qubit frequencies.   

Optical Charge Conversion of the SnV Defect in Diamond

Isolated color centers in wide-band-gap materials like diamond have shown promise in quantum information science (QIS) as single-photon emitters and qubits. While the nitrogen-vacancy (NV) defect in diamond has received substantial attention, many potential QIS defects exist, including the group-IV-vacancy defects. Due to the presence of donors in diamond like substitutional nitrogen, these defects typically occur in the negative charge state as spin-½ defects which may limit their applications. The neutral group-IV defects, however, are spin triplets like the NV^- defect. While SiV defects have been prepared in the neutral state, to date, other group-IV defects like the SnV defect have not. Here, we use density-functional-theory calculations to study the optical charge conversion of the SnV defect from the negative to the neutral charge state. We find a deep UV transition with a reasonably strong transition rate that can ionize an electron from the defect into the conduction bands. Furthermore, we study competing processes including deep-valence-band absorption, further conversion to the positive charge state, and band-edge recombination to characterize the robustness of the desired conversion.   

Implementation of the Kitaev-Webb and Klco-Savage algorithms on IBM Q Systems

Recent advancements in quantum algorithms have been significant, yet there is still much to be done in terms of benchmarking noisy quantum computing hardware. Utilizing IBM's Qiskit software development kit and quantum hardware, we have streamlined a novel way of benchmarking and characterizing error on noisy qubits. Specifically, we test the noise levels of IBM's quantum hardware by implementing the Kitaev-Webb state preparation algorithm (Kitaev, Webb 2008). We further examine Kitaev-Webb and Klco-Savage (Klco, Savage 2019) algorithms to prepare a 1D discrete Gaussian and a symmetric exponential distribution as a pseudo-Gaussian respectively. Such simulations provide insight into dominant sources of noise on quantum chips. However, the challenge in manipulating and maintaining states at quantum scales results in poor performance on these problems, even for shallow circuits.   

Improving quantum hardware performance using inverse noise matrices

With the advent of fault-tolerant quantum computing still far on the horizon, highly-entangled state preparation algorithms are not realizable on large numbers of qubits. Due to their versatility and the demand they place on the hardware, these algorithms can be useful for studying the noise on the hardware. We investigate how we can improve the performance of a quantum circuit by exploiting knowledge about the particular noise that affects the qubits.

To this end, we use the noisy output of the state-preparation algorithm to identify the noise on IBM quantum hardware. By identifying the dominant source of error, we reconstruct a noise model of the given hardware. Partial error mitigation is achieved by approximately mapping the noisy channel to a noiseless one. We further discuss our work in the context of the Klco-Salvage state preparation algorithm, as a minimally-entangled approximation to a discretized Gaussian.