Oral Sessions

Redox flow batteries (RFBs) are promising energy storage devices that play a pivotal role in mitigating the gap between grid scale energy demand and utilization. Conventional RFBs rely on acidic electrolytes (H2SO4) and expensive vanadium-based salts. Aqueous organic RFBs that operate at neutral pH are particularly attractive from a safety and cost standpoint. However, very few organic compounds function as efficient redox couples at pH 7. Our group has been studying substituted [3]radialene molecules as catholytes for aqueous organic RFBs. Here, we describe the synthesis and characterization of a series of cyanoacetamide-functionalized [3]radialenes. Specifically, mono-cyanoacetamide-tetracyano[3]radialene supports multielectron transfer in acetonitrile and reversible electrochemistry at neutral pH. We also prepared a highly soluble sulfonoacetamide-functionalized [3]radialene from inexpensive precursors that exhibits higher voltage oxidation in solutions of NaCl. Electrochemical analysis, also presented in this report.

Frost action (heaving and thawing) is a perennial problem encountered in the design, construction, and management of civil engineering structures, particularly road pavements in cold regions and areas that experience seasonal sub-freezing temperatures. Nationally, this leads to recurrent annual maintenance costs estimated at over 2 billion dollars, as well as additional economic impacts because of related vehicle damage, road closures, and weight restrictions. Studies identify three basic requirements for frost action; freezing temperatures, availability of water, and frost-susceptible soils. While advances have been made in the design for freezing temperatures and providing for groundwater separation, very little progress has been made in terms of in-situ soil improvement. A cost and labor-intensive approach is to undercut and replace unsuitable frost-susceptible soils. As an alternative, this presentation will describe Engineered Water Repellency (EWR), a process in which soils are made hydrophobic. This is achieved by combining soils with cost-effective and environmentally compatible polymers and other complex organic molecules. This study proposes an innovative approach for mitigating frost action through engineered water repellency. A frost-susceptible soil was collected from a test plot at the Charlotte Douglas International Airport and treated with a commercially available organosilane. Using the standard proctor test method, changes in density and corresponding water content were obtained. Data on water content, soil temperature, suction, and electrical conductivity were collected every ten minutes using Teros 12 and 21 sensors in both untreated and treated soil samples to monitor water infiltration. Preliminary results indicate an increase in the maximum dry density from 17.54kN/m3 to 17.66kN/m3 and a decrease in the optimum moisture content from 17.36% to 11.75% after treatment. Data obtained from performance tests carried out under sub-freezing weather conditions indicated that the treatment was effective in limiting the infiltration and migration of water into the soil matrix when compared with the untreated soil. As such, engineered water repellency may be a viable solution for Airports and Departments of Transportation seeking methods to mitigate frost action.

Whether it is our smart phone, social media or home security camera, all of them require recognizing your face in the pictures or videos to provide some sort of functionality, be it tagging, searching or providing authentication (eg. FaceID). The technology which allows developers of such a system to provide such functionality requires that end users must first share their data with them. This leaves end users with no choice but to unwillingly share data (facial images) to avail these services. This exposes their privacy and leaves developers under obligation to protect end-user data from unauthorized access. But when data breaches happen, the end user privacy is compromised and the organizations responsible for storing data face lawsuits. Our research devises a method which allows end users to privately share their data with developers such that even in the case of a data breach, the user's privacy will not be compromised. Methodology: Our method leverages recent breakthroughs in few-shot learning to train a facial recognition model just from a few data samples (eg. 5 images per user). Moreover, our model does not even require users to share their raw images. Instead, users obfuscate (add noise) their images by a chosen privacy method (blurring, pixelization, gaussian, DP-Pix) which renders their images indiscernible to the human eye. The protected images are then shared with the developers to train the facial recognition model. Later, when the user wants access to a functionality, such as authentication, their raw images are again obfuscated and sent to the model for authentication. This way user data is always protected, both from developers and any future data breaches. Figure 1 outlines this process (in the submitted PDF). Importance: Our work is a step towards building safe and secure facial recognition systems where necessary. Using the proposed method, the privacy of the user data is always protected both from application developers and potential data breaches, allowing developers to build their applications without needing access to raw user images, hence reducing their burden to safely store user data and as well as allowing users to avail smart services without compromising their privacy.

Light emitting diodes (LEDs) have become ubiquitous and crucial in our lives. However, LED technology often utilizes inorganic phosphors usually built with rare earth elements (REE). The environmental concerns related to mining and refining of REE have brought this technology under public scrutiny. One proposed solution to reduce the use of REE based lights is to develop lighting technology using fluorescent organic materials. But organic materials have complicated synthetic pathways, low thermal stability and prone to oxidation. We propose that thiazolothiazole (TTz) based materials could be ideal candidates to solve this problem of REE based lighting. Multiple groups have studied the high emission of TTz based materials in solution state for various applications in the past. However, the real question is whether this emission can be conserved in solid state for lighting applications? We hypothesize that this can be achieved by isolating the TTz core using bulky molecular spacers. To test this hypothesis, we designed 4 TTz based materials with varying molecular spacers. The crystals were optically characterized and found to be highly emissive, ranging from red to blue. Emission behavior of the crystals was further elucidated by studying the crystal structure under X-Rays. The X-ray diffraction studies revealed that modulating the spacer groups changes the molecular organization in crystals. We report herringbone and J-aggregation type molecular packing in these TTz crystals. The changes in molecular organization thus influence the unique photophysical properties we observe in these crystals. We have established that fluorescence in TTz materials can be conserved in solid state. We have also proved that the emission properties can be modulated with molecular level precision. The fluorescent crystals will be tested as phosphors to prove their efficacy as competitive candidates to industry standards. The success of this study will result in environment friendly, cheap and energy saving lighting solutions.

As populations grow and vehicle density increases, traffic accidents become more common. Bicycle lanes have been proven to reduce these accidents, improving safety for vulnerable road users such as bicyclists and motorcyclists. However, bicycle lanes are expensive to implement and increase road congestion as they’re built. Therefore, city officials need to decide which roads need the most upgrades. To evaluate which roads would benefit cyclists more with the construction of a bicycle lane, we propose a system that collects traffic data to detect which routes are used more and create the most congestion. Our goal is to collect audio, magnetic, and location data from cyclists’ smartphones, which are then analyzed to determine which roads get more congested when cyclists are riding. Audio and magnetic sensors consume low energy and are unaffected by weather and visual disruption. Having this data collected via smartphones means we can collect data from multiple roads from local cyclists without the need to install multiple devices and disrupt traffic. We evaluated various machine learning (ML) models, such as CNN and LSTM, to assess which is most effective at detecting a vehicle passing by the recording device. Models were trained on (1) publicly available traffic datasets and (2) our own recorded roadside data, showing that vehicles can be detected with an estimated accuracy of around ~84%. Published work by other researchers and our own collected data show the potential for magnetic sensor data to determine if a vehicle detected in the audio data is in the same lane as the cyclist, increasing the result's reliability. With the proposed system, local cyclists can help provide essential traffic data to aid government decision-making. This information can be used to help determine which roads are more dangerous and may need a bicycle lane most. Not only is the data collected without road interruptions, but they are collected from users who travel these roads regularly and would be benefited directly from road upgrades.

Nucleic acids (DNA and RNA) are integral biological macromolecules that were once solely revered for their role in processing genetic information. However, as a result of Watson-Crick base pairing (i.e., adenine (A) to thymine (T) or uracil (U), and guanine (G) to cytosine (C)), DNA and RNA are able to be utilized as building blocks to produce a myriad of nanostructures. Elaborate multi-strand assemblies—nucleic acid nanoparticles (NANPs)—take advantage of these intrinsically defined pairing rules and can be designed in silico to spontaneously assemble into predicted three-dimensional structures. NANPs are composed of several strands of synthetic DNA or RNA (either purchased or synthesized in-house) and are made via a one-pot assembly protocol. To do so, individual strands are combined at equimolar concentrations in water and are subjected to incubation at two different temperatures. The first incubation period is required to melt individual strands and reduce any secondary and tertiary structures, whereas the second promotes the designed interactions between individual strands to finally construct the desired NANP. With programmed shape, size, and functionality, NANPs are highly customizable and have broad applications within the field of nanotechnology. One use of this promising technology lies in its potential as a therapeutic. Owing to their dynamicity, NANPs can be designed to operate as a therapeutic by incorporating therapeutic nucleic acids (TNAs) into their structure. By decorating NANPs with TNAs, the production of abnormal proteins related to a disease can be inhibited by downregulating the production of that specific protein’s RNA precursor (i.e., gene) upon intracellular delivery of the NANP. This gene knockdown (or downregulation) is accomplished by a naturally occurring regulation process known as RNA interference (RNAi), which utilizes specific TNAs (Dicer Substrate RNAs) to guide and trigger gene silencing. Recently, this pathway has received tremendous attention for its potential to treat various maladies including cancer, inflammatory diseases, and viral infections. Herein, a conditionally-activated NANP switch that capitalizes on this cellular phenomenon is investigated. The NANP switch is designed such that if presented with a prognostic biomarker (disease-related gene), then the therapeutic functionality of the NANP will be “turned on.” This switch to the “on” state promotes the release of the previously mentioned Dicer Substrate RNAs that are capable of binding to that biomarker, ultimately marking it for destruction. Utilizing electrophoretic mobility shift assays (EMSA) and atomic force microscopy (AFM), preliminary results affirm successful fabrication of a uniformly structured switch. The switch’s ability to release Dicer Substrate RNAs targeting survivin (a gene responsible for inhibiting programmed cell death (apoptosis) and is highly expressed in cancer cells) is assessed and confirmed by using EMSA. Subsequent in vitro studies will be conducted in a human breast cancer cell line (MDA-MB-231) as well as in a prostate cancer cell line (PC3) to assess the efficacy of gene regulation of the NANP switch and quantify cell viability by using the MTS assay. It is expected that upon successful intracellular delivery of the NANP switch, its activated state will promote a decrease in production of survivin via RNAi, leading to an increase in quantifiable cell death.

Gold nanoparticles (GNP) in combination with photothermal therapy (PTT) shows great promise for cancer treatment. As compared to chemotherapy and radiotherapy, PTT is minimally invasive, and is extremely targeted and localized. GNPs have excellent thermal and optical properties that include concentrating and converting light to heat, thus enabling hyperthermia and PTT, but their temperature profiles and PT efficiency have only been assessed in solutions and do not correlate with that in vitro or in vivo. One of the main reasons for the lack of PTT’s clinical translation is the lack of understanding of “true” temperature profiles of GNPs. Without that knowledge, clinicians cannot determine the GNPs dosage or laser irradiation time required for PTT. Currently, infrared thermal imaging is used for remote monitoring of tumor temperature during PTT, and thermocouples are used for measuring GNP temperature profiles in solutions, but these methods cannot determine the “true” local temperature profile of GNP to evaluate the PT efficiency. To solve this problem, a “nanothermometer” is being developed that can identify the local temperature profile of GNP in real time, in solutions as well as biological environments. The objective here is to synthesize, characterize and optimize a fluorescent dye that would successfully bind to a GNP to form a probe which would work in vivo as a temperature sensor. A class of novel fluorescent stable molecules called thiazolothiazoles (TTz) have been successfully synthesized, and they demonstrate characteristics that are suitable for work in vitro. The TTz dyes developed so far are very pure with almost no side products and they exhibit very high yield of >95% in certain solvents. So far, we have measured their photoluminescence decay, quantum yields, light absorbance in different product concentrations and different solvents, and confirmed their structures using mass spectrometry and nuclear magnetic resonance spectroscopy. Their solvent polarity, fluorescence emission and Raman scattering change with temperature, and these properties can be exploited to make them work as temperature sensors in conjugation with GNPs. The innovation of this approach relies on the unique integration of thermofluorochromism and Raman scattering of fluorophores for local temperature sensing, none of which have been used together or even separately for this purpose. This work would set a foundation for an optical thermometer for in vivo real time temperature sensing, which is virtually non-existent.