Participants
Click on name to see their abstract and elevator pitch video.
Organic photovoltaics (OPVs), also known as organic solar cells, can be an alternative to inorganic solar cells, as they come with a host of benefits. Lower manufacturing and installation cost along with the use of relatively inexpensive organic materials make organic solar cells a more cost effective option. Their flexible displays also open up opportunities for more widespread installation of solar panels in places that are inaccessible to inorganic solar cells. OPVs have both electron donating (donor) and electron accepting (acceptor) materials. Acceptors can be classified as either fullerene or non-fullerene acceptors (NFAs), with NFAs being advantageous due to their ability to better absorb light and be chemically modified. While there is extensive research on benzodithiophene (BDT) based molecules for use as NFAs, this study focuses on benzodifuran (BDF). BDF is a less-known but potential alternative to BDT. Its furan rings make the material more soluble due to the polarized C-O bond. Solubility is important because organic solar cells are solution processed, so the material must be soluble enough to be spin-coated. Using 1,4-dimethoxybenzene as a starting material, a series of reactions were completed to synthesize the BDF TIPS core. Some reactions that were implemented include Sonogashira cross-coupling and Larock iodocyclization. Notable characteristics of the final BDF TIPS core structure are the iodine functional handles and the triisopropylsilane (TIPS) groups. The iodine functional handles allow other groups to be attached to the structure, and the TIPS groups increase solubility and act as protective groups. These properties make the BDF TIPS core a valuable foundation for further chemistry. Future work includes continued synthesis using this core as a base for several synthetic pathways, with the final molecules implemented as non-fullerene acceptors in devices.
When scanning and measuring the effects of laser pulses on a sample, you want to have an efficient set up that will allow you to collect more data points. Experimental factors, including temperature and laser stability, create constraints when it comes to the time frame for data collection. To help reduce the effects of these constraints, adjustments can be made to the original experimental set-up, utilizing redesigned parts. Vacuum pumps are used to create low-pressure environments where air and gas particles are removed, so that they don’t interfere with a sample’s properties and reactions. These vacuum pumps create a pressure differential to evacuate air from an enclosed volume, but this tool does have some efficiency limitations when trying to create this low-pressure environment. Building certain pieces to transfer the placement of the vacuum pump and widen the diameter of space the particles flow through allows for the cooling time frame to last longer as more air and gas particles are being removed. The creation of a piece with the dimensions to fit a cryostat and a back window fit for a larger flange connecting the pump, results in a more efficient air flow. Through the use of CAD software and 3D printing, we were able to prototype a new piece, the rectangular tail, that achieves these goals. Once the rectangular tail is implemented within the setup, the pressure is estimated to decrease by about 0.5-1 arbitrary units. Using the new piece to prolong cooling time, there is more room to scan over the pulses, allowing for more data in a shorter amount of time. This makes the results of the experiments more accurate because a greater sample size minimizes noise. Ultimately, the implementation of this piece allows for the lab's experimental procedure to be more efficient, leading to better results.
2D semiconductors are an emerging field of study, they are flexible, have different band gaps dependent on thickness, and are promising materials for micro-nano electronics. We focused our study on 2D molybdenum disulfide (MoS2), MoS2 is a transition metal dichalcogenide (TMD), it is a van der Waals layered material. MoS2 has been studied in depth, however, most of the works are studying highly crystalline MoS2, while little has been reported on MoS2 with high defect densities. The conversion of metal nitrides to metal sulfide has been largely uninvestigated; we demonstrate an experimental technique to reconstruct the lattice of MoS2 and introduce excess defects into its crystal. We also explored the conditions of this conversion. In order to achieve the lattice reconstruction, MoS2 was first converted into Mo5N6 and then was converted back to MoS2 through the use of a tube furnace. Raman spectroscopy was used to identify the product and obtain structural information of the converted flakes. The structure of the converted material was different from pristine MoS2, the converted MoS2 had many defects based on the optical images. The material we synthesized can be used in future electronic applications.
Neural networks mimic human decision making through a multi-step process in which the input data is passed through several layers of linear and nonlinear transformations to produce their output. The parameters of the linear transformations are learned by choosing them to maximize the accuracy on a training dataset. My project focused on building a convolutional neural network (a particular type of neural network) to label images based on whether or not they contained a Roomba Create robot in them. Toward this goal, the first step was to create a dataset containing both positive and negative examples. A challenge of neural networks is that they need a large training dataset. This was achieved via a procedure called data augmentation. For my positive data, I took a video of the robot on top of a green screen created with several green mats on the floor. I substituted the green pixels in the background of these images with the corresponding pixels from ten other images to generate different training examples for each original image. I repeated that process for my negative data, except random parts of the lab were my subject instead of the robot. To further augment the dataset, I added copies of the original images that were vertically flipped so that my dataset reached 1,520 images. I split my dataset so that 90% of the images were put in a subdirectory called “training” and the remaining 10% in another subdirectory called “testing.” Each of these subdirectories contained two additional folders: “robot” and “not-robot.” To further challenge the model, I corrupted the original robot images in the testing dataset by turning random pixels black. After training a convolutional neural network with seven layers, the model correctly predicted whether the robot was in the image 100% of the time for uncorrupted images while the accuracy decreased from 99.34% to 17.1% for corrupted images. In real-world security issues where robots must monitor one another, this convolutional neural network can be used to sort the data provided by the robots’ video feeds so that there is evidence for whether or not the other robots behaved as expected.
Nanopatterning, or creating extremely small patterns in a surface, is a technology with many potential applications. It can make solar cells more efficient and has uses in optoelectronics and semiconductors. One emerging efficient method of nanopatterning is ion-beam sputtering (IBS), where a surface is bombarded with a stream of ions to form surface dots, ripples, or ultra-smoothness at the nanoscale. However, modern IBS technology doesn’t create reliable, highly-ordered nanopatterns. Our research aims to determine what variables control the size and characteristics of nanopatterns created through IBS. In this experiment, we bombarded thin silicon wafers with a beam of Ar+ ions at 250eV with possible Fe and Cr contaminations inside a vacuum chamber. We varied the incidence angle of the ion beam and the amount of time each sample was bombarded for. After a sample was bombarded, its topography was imaged using the atomic force microscope (AFM). The roughness of a surface as well as the average wavelength of patterns on the surface was measured from the AFM images. Our data demonstrated that more highly ordered ripples may form at 50o than at 60o, while an increase in bombardment time at 50o or 60o may correlate with taller nanostructures. Samples bombarded at 0o retained flat topographies similar to unbombarded samples, which matches previous 0o studies with no metal contaminations. However, more research will need to be done to determine if 0o bombardments with the possible metal contamination still have a flattening effect in Si. Furthermore, samples displayed patterns with slightly different characteristics even when they were bombarded at the same angle for the same amount of time, showing that more research needs to be done to figure out how to make these patterns more consistent. Nevertheless, this experiment provides additional data about ion-beam nanopatterning that may be utilized for future analysis and compared to previous experiments, leading to a greater understanding of IBS.
Though they account for only 1% of the earth’s surface, salt marshes play an important, yet often overlooked, role in carbon sequestration. While we understand salt marshes to be strong carbon sinks, meaning they sufficiently store carbon, little is known about carbon export from the land to the coastal ocean and how carbon stocks in the coastal ocean compare to that of upland and salt marsh ecosystems. In this study, we collected sediment cores from three distinct ecosystems along the land to coastal ocean continuum (the upland, salt marsh, and coastal ocean) and analyzed the organic matter concentrations within samples through a method called loss-on-ignition, which uses combustion to release and measure the organic matter within samples; this produce allows us to draw conclusions regarding the organic matter concentrations and, subsequently, the variance in carbon stocks along the coastal ecosystem gradient. We observed high amounts of organic matter in the salt marshes— approximately three times greater than the levels found in the upland ecosystem and fourteen times greater than the levels found in the coastal ocean. By running an Analysis of Variance (ANOVA) test, a significant difference in organic matter concentrations between the three ecosystems was confirmed. The results from our experiment provide evidence supporting the significance of salt marshes in carbon sequestration and mitigating rising atmospheric carbon dioxide levels.
As the source of one of the most common psychiatric disorders in the world, it is crucial to understand how the human brain processes and responds to the presence of fear. The key to developing effective treatments for anxiety and PTSD could be dependent on the introduction of a safety signal, created to inhibit any feelings of distress. In this study, we trained adolescent mice of both sexes to discriminate between fear and safety cues in order to understand how age and sex interact with their ability to process and respond to unknown stimuli. We hypothesized that an adolescent mouse’s ability to discriminate between cues is not significantly impacted by their sex. Throughout a 4-day training period, mice were conditioned to feel fearful under a tone paired with a shock, but safe under the tone paired with no shock. We measured the amount of time the mice spent “freezing”, as it indicated whenever they felt terrified. To test the efficacy of these cues, we experimented on a 5th day, in a completely new context. We also introduced them to a new, “novel” tone in order to evaluate the mice’s thought processes– are they associating this tone with danger or safety? They were also exposed to two additional compounds– a combination of the fear and safety tones and a combination of the fear and novel tones. Our results revealed that all mice are freezing significantly more under the fear cue than the safety cue. They were also freezing to the novel cue for around as long as they did with the fear cue. We concluded that although adolescent mice are capable of safety signal learning, they are more likely to fear unknown elements than adults. Additionally, our statistics reveal that there is no significant sex differences in the safety signal learning of the male and female adolescent mice. Through more experimentation, we hope to improve our understanding of mice behavior and by default, maximize the potency of safety signal learning.
Vibrations can interfere with many experiments that need to be able to detect very small signals or make precise measurements. In our lab, an experiment that aims to detect axions, a candidate for dark matter, will be placed inside a dilution refrigerator. This uses a pulse tube that creates strong vibrations, which could result in inaccurate measurements or confusion between signals detected from dark matter and those merely generated by vibrations. Our experiment aims to understand how a spring pendulum can be used to isolate vibrations, and how its effectiveness is impacted by different variables in its configuration. We used a loudspeaker and signal generator to create vibrations, then measured the acceleration of a stiff rod (our control case), two spring pendulums with different spring constants, and the bar they were both connected to. We found that at the point of the rod’s peak vibrations, the less-stiff spring isolated vibrations by a factor of 169. Additionally, the driven frequencies that created the most vibrations were similar for all systems and did not depend on the resonance frequencies of the springs or pendulum. This will be used in the future to design a spring pendulum specifically for an experiment that is currently being developed and will eventually go inside the dilution refrigerator with the goal of detecting dark matter.
Neurostimulation is an emerging treatment for neurological diseases as well as a source of insight for basic neuroscience research. Previously, our lab has developed a fiber optoacoustic emitter (FOE) that converts laser pulsed through an optical fiber into ultrasound waves to stimulate neurons with high spatial precision. However, the light absorption of the fiber tip can still be improved to maximize the optoacoustic (light to sound) conversion efficiency. Carbon nanotubes (CNTs) have been used in FOEs due to their high light absorption but are limited by the low maximum concentration of CNTs in polydimethylsiloxane (PDMS), the matrix material that expands and compresses to produce ultrasound waves when heated by absorbed light. An alternative absorption material is candle soot. In this study, we fabricated a candle soot FOE (CSFOE) consisting of a fiber tipped with a layer of candle soot followed by a layer of PDMS. This allowed for a higher absorber concentration and thus greater light absorption. We then compared the neurostimulation capabilities of the CSFOE and the CNT-FOE under the same laser conditions on cultured rat primary cortical neurons expressing the calcium indicator GCaMP6f. The CSFOE induced a fluorescence increase indicating successful activation of the neurons, while the CNT-FOE did not. These results demonstrate that the CSFOE has a superior optoacoustic conversion efficiency compared to other FOE types, thereby boosting FOE technology’s potential for furthering neuroscience research and treating neurological conditions.
Not many alborial treefrogs of the tropics take part in explosive breeding during their matingseasons. In species that reproduce using explosive breeding, members of this species allagglomerate to a specific area in their shared habitat. There has been plenty of research done onaquatic breeding, temperate zone explosive breeders, however, little data has been acquiredregarding those that live in the tropics or those that lay eggs on land. Agalychnis Spurrellitraditionally have their “breeding get-togethers” in the high canopies of the tropical forests theylive in, and this factor makes it a bit difficult to study them. A. Spurelli also have an extremelyshort breeding period–which lasts several hours or up to a day–leaving a short window ofopportunity to study them. During one observational study, it was found that during theseexplosive breeding aggregations, there are two behaviors that some male frogs do: the “eggkicking” behavior, where a male frog kicks away eggs, and “vent positioning”, where a malefrog sneaks into an already mating couple where he aligns his vent with theirs. There were fourdifferent types of samples gathered from an explosive breeding site: frog offspring (eggs andtadpoles), adult frog toe pads (from putative parents), and frog semen swabs. In order to test ourhypothesis, a series of DNA extractions from these samples were performed. During theinternship, only DNA extractions were completed, however the next step in the process wouldhave been to send the already cleaned DNA off for sequencing. The purpose of this study was totry and find out what exactly is the function of these specific behaviors. Competition amongstmale A. spurrelli is very high so not every male is able to find a mare that he can successfullymate with, so we hypothesize that these behaviors may be ways that the un-paired male frogsattempt to fertilize at least some eggs instead of not having fertilized any at all.
Trees in urban areas are dying at twice the rate of rural trees, in part due to increased heat, pH, and pollution and decreased moisture. As urbanization and forest fragmentation increase globally, their impact poses a threat to trees and their fungal symbionts, called ectomycorrhizal fungi. These fungi facilitate many interactions between other microbes, as well as between trees and the soil microbiome. To comprehend how urban conditions are affecting the trees, we studied the fungi in soil across urban and rural forests in Massachusetts to compare their connectivity, or number of interactions. We hypothesized that urban fungi would have less connectivity (fewer correlations) than rural fungi because of the negative impact of urbanization on ectomycorrhizal fungi. To visualize and run statistical tests on the data, we formed networks to represent fungal connections computationally. The output of the networks was the opposite of expected: trees in urban areas had more connections between their fungi than trees in rural areas. The connectivity demonstrates there is more correlation between those urban fungi, but it does not necessarily mean they are interacting more. We further hypothesize that this is because urban microbes have access to more nitrogen, a form of their nutrients. Allowing for microbes to share a common resource without competing, the surplus of nitrogen can increase multiple microbes abundance at a time. When many microbes can grow in quantity, the program understands their relationships as correlations, even if they will never actually interact. Currently we’re extracting new fungal DNA from street trees, which we predict will have even more extreme results in our networks because of the heightened urban conditions.
The impact of osteoarthritis (OA) on the risk of hip fracture is not yet fully understood. The femoral neck (FN) is an important facet of hip fracture because it is a region of complex loading and high failure risk within the proximal femur. Thus, by quantifying OA-related differences in measures of bone microstructure and density in the FN, we can better understand OA pathogenesis. The micro-computed tomography imaging system utilizes x-rays to produce high resolution 3D images of FN samples, and semi-automated image segmentation scripts generate periosteal and endosteal contours to analyze the cortical and trabecular regions. However, these scripts are not as efficient at detecting signs of advanced OA (i.e. osteophytes and 'trabecularization' of the cortical region), so operators must make manual corrections. To ensure that all inherent operator bias is accounted for and eliminated from the method, we conducted an inter-operator variability study. The aim of our investigation was to test the null hypothesis that there is no statistically significant difference between the measurements of bone density and microstructure from two trained operators, independently executing the same segmentation and analysis methods for the same data sets. Both a paired t-test and an intraclass correlation coefficient test were conducted to test this hypothesis. Significance level was set to 0.05 for all tests. We found no evidence to reject the null hypothesis for almost all bone density and microstructure measurements, except for cortical separation. We speculate that cortical separation is the outlier because it is dependent on both the periosteal and endosteal contours as well as the pore network analysis. By proving the repeatability of the method between operators for the majority of the measurements, this investigation furthers the overall research project, which is critical to understanding hip fracture risk, establishing biomarkers of OA pathogenesis, and developing preventative OA treatments.
Regulatory T (Treg) cells regulate or suppress other cells in the immune system, but overexpression has been linked in several types of cancer. Specifically, the acetylation of a transcription factor called Foxp3 by the histone acetyltransferase TIP60 has been the major cause of Treg cell overexpression. TIP60’s interaction with Foxp3 has proved difficult to inhibit, as cancer treatments are facing problematic drug resistance from targeted proteins. A PROTAC, or proteolysis-targeting-chimera, is a molecule made of an enzyme binding domain and a target protein binding domain attached by a linker; the molecule utilizes E3 ligases to ubiquitinate proteins of interest and tag them for degradation via natural cellular processes. Therefore, the goal of this research is to create a PROTAC that will degrade TIP60 in cells rather than simply inhibit its function. The PROTAC was successfully synthesized through a series of reactions by attaching an E3 ligase binding domain and a TIP60 binding domain with an amine linker. Reaction success was confirmed using UPLC and NMR characterization methods. The PROTAC eliminates any formation of drug resistance by the protein of interest, resolving a major obstacle faced by current cancer therapies. The resulting product will be sent to collaborators to perform testing for biological activity. Overall, this PROTAC has the potential to become a new cancer therapy that could save thousands of lives in an effective manner, overcoming some of the major obstacles that current treatments face.
Several million traumatic injuries occur in the United States annually, but mild injuries are often unreported, leaving specific internal effects poorly understood. When injury is induced upon the body, programmed cell death and the removal of apoptotic corpses is necessary in maintaining homeostasis. In these processes, draper is utilized as a cell-surface receptor required for the recognition and engulfment of unwanted or damaged cells, but the body’s response to trauma when draper is knocked down is unclear. Due to the lack of data on female damages after trauma, our investigation seeks to discover the differences between draper mutants and the control group, revealing unknown physiological effects due to mild impact. To begin to investigate this issue, we used the w1118 cross to serve as a control, but four groups of Drosophila Melanogaster were used to study the impact of mild injury: sham w1118, trauma w1118, sham drpr -/-, and trauma drpr -/-. The number of eggs laid for five days straight was counted for fecundity, their ovaries were dissected using a DAPI staining to determine fertility, and erioglaucine disodium salt (blue dye) was used to determine any leakage in the gut. Our results suggest that trauma in the w1118 control group resulted in a negative relationship compared to the sham, as they laid fewer eggs and held on to more stage 14s. Trauma in the draper mutant group resulted in a positive relationship compared to the sham, as they laid more eggs and withheld less stage 14s than the sham group. The results with the draper mutant group are significant as they are unexpected, and have not been seen before. This discovery is crucial, as it suggests that even with the absence of a cell-surface receptor for damaged cells, the body was able to recover to a more biologically healthy state.
One way of exploring how climate change will impact oceans through acidification and warming, is looking at the relationship between coral, their agal symbionts, and their microbiome. The question is, which symbionts and bacteria are specific to certain populations of coral, how that might change the coral's response to climate change through gene expression. Our goal is to extract DNA from samples of coral from Belize in which the success is measured through nanodropping and gel electrophoresis. It was determined that all but three samples' extractions were successful as the rest had high concentrations and clear bands of DNA, allowing for them to be sequenced further down the road (those who failed were successful during a second extraction.) The next step for the samples is to run the samples through ITS2 in which the species of symbionts will be determined. Not only is this research crucial for understanding the fate of coral reef ecosystems and how we might be able to change it for the better, but it also sheds light on the fate of coastal towns' economies and their safety as the likelihood for severe weather increases.
Multiple studies have shown that acoustic stimulation during slow-wave sleep may enhance declarative memory, but the mechanisms behind this effect are not fully understood. Functional magnetic resonance imaging (fMRI) can help us better visualize the effectiveness of sleep stimulation; its high spatial resolution would allow us to gain insights on how the stimulation affects the hippocampus, which is involved in memory consolidation but is too deep inside the brain to be measured by scalp EEG. However, performing simultaneous EEG-fMRI scans is extremely difficult because the MR scanner adds noise to the EEG signals, making the data unusable unless it is cleaned. We built a reference layer cap for subjects to wear during simultaneous EEG-fMRI that effectively allows EEG data to be cleaned both offline and in real time. The layer blocks multiple EEG channels from the scalp so they are only able to detect the pure noise signal, which is later subtracted from the combined noise-brain signals picked up by the unobstructed electrodes. This processing technique isolates the valuable electrical signals from the brain themselves, indicating precisely when to induce stimulation on the subject in the MR machine. Additionally, we developed three lists of 44 semantic word pairs to serve as a word-pairing memory task administered before and after sleep, which will assess the effects of the stimulation on memory consolidation. These tools bring us closer not only to confirming the memory-enhancing effect of acoustic stimulation during sleep, but also to better understanding the way memory functions as a whole. On a broader scale, the reference layer can enable better quality simultaneous EEG-fMRI regardless of experimental paradigm.
Unlike most mammals, neurogenesis and neuron migration continue to occur widespread in adult zebra finches, especially in the HVC region of their brain. In vivo two-photon microscopy has been historically used to study this phenomenon of neuron migration. However, due to the scattering and out-of-plane excitation constraints, imaging is limited to very superficial regions of the HVC, leaving many areas—that may exhibit different movements—unreachable. Therefore, the goal of our experiment was to use a microprism implant, a novel approach that allows deeper imaging into the brain, to investigate the differences in neuron migration patterns in deep and superficial brain regions. To do so, we used transgenic GFP labeled zebra finches and through surgery, implanted a cranial window and prism before imaging through a two-photon microscope and tracking movement using the Fiji application. We found—through analysis via MATLAB software—no statistically significant difference between four characteristics (speed, displacement, distance, and tortuosity) of neuron movement in the two regions. These findings suggest that imaging superficially in the HVC is a good representation of overall neuron dynamics. Furthermore, we found that imaging through an implanted microprism allows neuron tracking in transgenic songbirds, which may be used as a novel approach in future research in hopes of identifying the cellular and molecular mechanisms of neuronal migration and neurogenesis, with the long-term goal of creating a basis for therapeutic technology for neurodegenerative diseases.