11/07/2024
By Suzanne Young
Degree: Master's
Date: Thursday, November 21st 2024
Time: 2:30 p.m.
Location: Olney Hall, Room 518
Committee Chair: James F. Reuther, Department of Chemistry, University of Massachusetts Lowell
Committee Members:
Marina Ruths, Department of Chemistry, University of Massachusetts Lowell
Jerome Delhommelle, Department of Chemistry, University of Massachusetts Lowell
Abstract:
Globally, especially in third-world countries, lack of access to clean water has been the direct cause of millions of deaths every year. Current methods of water treatment (e.g., adsorption with activated carbon) although effective, lack the ability to be easily regenerable limiting their sustainability. Activated carbon requires high energy/temperature (~800 °C) in controlled atmospheres to burn off organic pollutants that are adsorbed to the pores. These conditions are not able to be reached in a modern home, necessitating the development of low-energy regenerable adsorbent materials. Herein, the described research explores the synthesis of nanostructured gels using photo-controlled atom transfer radical polymerization-induced self-assembly (PhotoATR-PISA) for water treatment. These adsorbent materials offer two distinct advantages over activated carbon: 1) chemical tailorability allowing for targeted adsorption characteristics and 2) dynamic network characteristics to facilitate pollutant desorption and adsorbent regeneration using stimuli-triggered crosslink bond-exchange.
The advent of nanoparticle imaging has revolutionized the analysis of synthesized compounds and their assemblies, enabling detailed examinations at the nanoscale. However, the laborious task of classifying and annotating these nanoparticles has posed significant challenges, often consuming hours, if not days, of researchers' time. Leveraging the power of machine learning (ML), this study proposes a solution aimed at expediting nanoparticle classification and annotation processes. By employing Support Vector Machine (SVM) algorithms, a program was developed capable of swiftly classifying hundreds of transmission electron microscopy (TEM) images of nanoparticles, reducing the time required for classification and annotation. Results demonstrate that the proposed ML framework achieves classification accuracy exceeding 90%, alongside impressive recall and F1 scores. This approach holds promise for streamlining nanoparticle analysis workflows, paving the way for more efficient and effective research in nanoscience and materials chemistry.
Polyurethane (PU) vitrimers represent a significant advancement in polymer science through the integration of dynamic covalent bonds (DCBs) into conventional PU networks. This integration enables the material to retain the mechanical strength and durability characteristics of thermosets while incorporating recyclability and reprocessability typical of thermoplastics. This unique combination of stability and recyclability positions PU vitrimers as promising candidates for sustainable material applications. Herein, we will also discuss recent research findings into incorporation of dynamic conjugate acceptor (DCA) groups into PU materials to impart vitrimer properties and controlled degradability via addition of synthetic DCA-diol monomers into commercial resins for footwear. Three DCA-diol monomers were synthesized with different core DCA groups to investigate their influence on reprocessability and degradability. For this, we investigated the mechanical properties of distinct DCA-PU-vitrimer formulations with varied amounts of DCA-diol incorporated (ca. 1 – 10 w%) and assessed their potential for seamless incorporation into existing polyurethane systems. Further, research was conducted on DCA-PU degradation using ethylene diamine as a chemical trigger clipping specifically at DCA bonds creating PU byproducts well suited for upcycling.
All interested students and faculty members are invited to attend.