Microplastics pose multiple health and environmental risks due to their presence in resources such as drinking water. In addition, the increasing amount of plastic waste entering the oceans is resulting in adverse effects such as climate change. Selectively detecting and sorting microplastics still remains a challenge. Here, we aim to carry out studies using Dielectrophoresis (DEP) enrichment and Electrochemical sensing towards the development of a sensor system for detecting microplastics in drinking water. Dielectrophoresis experiments act as a pre-concentrator to manipulate the presence of microplastics in water based on their response to a non-uniform electric field. We will also explore the use of different electrode geometries and dielectric properties for the pre-concentration studies. Electrochemical detection using spectroscopic techniques like Electrochemical Impedance Spectroscopy (EIS) will be carried out based on the pre-concentration studies leading towards the development of label-free detection of microplastics. This novel method of combining two different techniques can lead to the development of a unique sensor platform using various electrode configurations to detect presence of microplastics selectively. In addition, this sensing technique can also be used as a bio- sensor platform for detecting microplastics beyond just drinking water but also presence of microplastics and other harmful bio-particles in human blood.

Current enhanced oil recovery methods lack effectiveness in retrieving oil and care for the environment. Despite this, the development of a cleaner, more efficient alternative has been limited. Currently, enhanced oil recovery methods include thermal recovery methods that involve the flowing of heated gas typically CO2 into an oil reservoir to lower the viscosity and increase its flow or a chemical approach that uses surfactants and polymers to reduce surface tension. Both these methods involve the usage of a brine that can contaminate the environment such as contaminating drinking water supplies with heavy metals. Nanofluids are an effective method for oil removal in porous media that can improve enhanced oil recovery or remediation with significantly less environmental damage. Nanofluids possess a unique ability to concentrate at contact lines and produce pressure called structural disjoining pressure that allows for oil to be removed from complex low permeability geometries that cannot be reached with current EOR methods. By enhancing the structural disjoining pressure created by nanofluids, nanofluid efficiency for oil recovery can be drastically improved offering a cleaner more efficient EOR method. Beyond this, due to its efficiency in complex geometries, its application can be expanded to clean deep intricate three-dimensional designs.

This study will quantify eNOS phosphorylation levels in endothelial cells under differing pulsatile flow regimens. Using an Elve flow microfluidics system, we will adjust the flow pulse rate under healthy blood flow profiles and compare the findings to a previously characterized model of orbital flow; however, it is limited by the need for supra-physiologic pulse rates to achieve healthy flow profiles. Thus, it is unclear how pulse rate factors into healthy endothelial cell phenotype. To characterize the flow velocity of the microfluidics system, we will evaluate the volumetric flow rate of the microfluidics system by collagen coating the microfluidics chips (ibidi) to minimize an observed gradient and ensure the cell surface experiences normal flow profiles. Then, we will run a 24-hour microfluidics experiment at similar flow velocities to the orbital shaker model. To evaluate if the biological response to microfluidic flow is identical to the orbital shaker model, we will use quantitative immunostaining of phosphorylated eNOS. If we observe similar phosphorylation levels, we will conclude that the microfluidics system reliably induces normal flow patterns and perform experiments evaluating the effect of pulse rate alone on endothelial cell eNOS phosphorylation. We expect this method of coating the microfluidic chip will minimize the flow gradient. The findings of this study will allow for the evaluation of how pulsatile flow frequency affects vascular endothelial cell biology to reveal if endothelial cells are sensitive to the frequency of applied mechanical loads, which is yet to be fully clarified.

Retention rates and academic performance in undergraduate STEM (science, technology, engineering, and mathematics) courses are critical to building a robust and competitive workforce. Despite various intervention efforts, traditional university support systems have proven inadequate in addressing the diverse needs of STEM students, often resulting in low retention rates and suboptimal academic performance in STEM degrees. In order to better identify students who might be at-risk, pre-trained Large Language Models (LLMs) were used to predict students’ performance. This study assesses how effectively existing pre-trained LLMs can process and learn from a natural language dataset featuring high-dimensional, experimental time-series data focused on student learning. This data included student grades and assessment data which were run through our computer model to identify students as “at-risk,” “prone,” “average,” or “outstanding”. To improve the adaptability of LLMs for forecasting tasks within this domain, we have developed a data enrichment technique. This technique includes strategies for replacing missing values, augmenting text sequence data, and incorporating specific task instructions and contextual cues. Through this, we aim to understand the applications of LLMs in an educational context, emphasizing both their potential and limitations in forecasting academic outcomes based on experiential data.