Karam Abdullah
Pronouns: He/Him/His
Research Mentor(s): Angela Violi
Co-Presenter:
Research Mentor School/College/Department: Mechanical Engineering / Engineering
Presentation Date: April 20
Presentation Type: Poster
Session: Session 1 – 10am – 10:50am
Room: League Ballroom
Authors: Karam Abdullah, Jacob Saldinger, Angela Violi
Presenter: 17
Abstract
The propensity of aggregation of polycyclic aromatic compounds (PACs) is a very significant property to consider when it comes to studying nanoparticle formation and growth. This paper studies this phenomenon in order to get a deeper understanding of some of the properties that influence the dimerization process between multiple PACs. The formation and growth were stimulated using Molecular Dynamics and the aggregation propensity model was quantitatively studied using Machine Learning. This paper utilizes a dataset of PAC monomers and their free energy of dimerization which would be augmented using the Molecular Dynamics which is enhanced with Metadynamics. Using a least absolute shrinkage and selection operator (Lasso), a machine learning model is capable of quantitatively learning how some molecular features like size can contribute to the physical aggregation which then helps in predicting the free energy of dimerization for new pairs of molecules. The Model used has the ability to compute the stability for homodimerization and heterodimerization pairs. The approach used provides a method to determine the features most important to construct the models. Some of the properties that were determined to be influential to physical dimerization are size, shape, oxygenation, and presence of rotatable bonds. We anticipate that this approach will lead to more effective modeling of the polycyclic aromatic compounds dimerization process as it reflects the efficient prediction of the propensity to aggregate from molecular features that are easy to compute.
Engineering, Interdisciplinary
