NCSI: Dynamics and Transport Properties of Polymer Blends in Dense Nanoparticle Packings

Status	Completed
Mentor Name	Anastasia Neuman
Mentor's XSEDE Affiliation	Research Allocation
Mentor Has Been in XSEDE Community	1-2 years
Project Title	Dynamics and Transport Properties of Polymer Blends in Dense Nanoparticle Packings
Summary	This project aims to support the work of the mentor involving field theoretic study of the thermodynamics of polymer blends in dense nanoparticle packings, which is supported by the mentor's NSF GRFP grant and NSF CBET-1933704, and is the continuation of an XSEDE EMPOWER Fall 2021 project. The effect of nanoconfinement and polymer-nanoparticle interactions on the behavior of polymer blends in highly-filled nanocomposites is poorly understand, in part because the experimental preparation of these systems proved difficult prior to the development of the capillary rise infiltration (CaRI) method in the lab of the mentor's co-advisor. Molecular dynamics simulations in LAMMPS are an alternate method to field theoretic simulations that can shed light on the phase separation dynamics and transport properties of these composites, and will compose the bulk of the work of the student.
Job Description	Polymer nanocomposites have wide-spanning applications as coatings and membranes due to unique functionality of nanoparticles and the ease of processing polymers. Previous studies have revealed loading high concentrations of nanoparticles (NPs) above 50 volume percent in nanocomposite films can lead to fabrication of separation membranes with high selectivity and high permeability, packaging with superb barrier properties, and electrodes for solar cells with high power conversion efficiency. Manufacturing such composites, however, has proved difficult because of the high viscosity and elasticity of mixtures containing a high concentration of nanoparticles and the tendency of nanoparticles to aggregate during processing. A new method to produce highly-loaded nanocomposites has been developed based on capillary rise infiltration (CaRI). This method induces the infiltration of polymer into NP packings by annealing the polymer above the glass transition temperature. The original volume fraction of NP packings is retained after polymer infiltration, leading to formation of nanocomposites with extremely high filler fraction. In CaRI composites, the characteristic length scale of the polymer chain can be much greater than the characteristic pore size in the NP packing, creating an ideal system to study the behavior of polymers under strong nanoconfinement. CaRI also enables preparation of multiphasic nanocomposites of two polymers with complimentary properties, but the microstructure of these composites and the impact of that on composite properties is not yet understood. Confinement can fundamentally alter the thermodynamics of this multiphase system due to increased interfacial effects and reduced configurational entropy, which in turn can drastically influence the transport and mechanical properties of these composites. The student will perform simulations in LAMMPS to understand the behavior of polymer blends in the interstices between nanoparticles (i.e., under extreme nanoconfinement) created using CaRI. Simulations are particularly useful in this system due to the vast parameter space, and the simulations the student runs will be used to guide experimental design of polymer nanocomposites. Student Aim 1: Study the impact of nanoconfinement on the phase separation dynamics of polymer blends. The student will explore the impact of the confinement ratio, the ratio of the polymer radius of gyration (Rg) to the pore radius (Rpore) within the NP packing. The student will alter the nanoparticle radius and distance between nanoparticles in the simulations to alter Rpore. Rg of the polymer will be varied by changing its number of segments (i.e., degree of polymerization). Experimentally, the CaRI system has reached confinement ratios upwards of 50. During the student's XSEDE Empower Fall 2021 tenure so far, we have developed a model for highly-filled polymer nanocomposites with solid nanoparticles in LAMMPS in which the confinement ratio can be more easily altered. Previously, the PNC model relied on changing polymer chain length to change confinement, and the computational resources needed to run highly confined systems scaled unfavorably. During Fall 2021, the student developed a method to analyze the structure factor of these composites in order to understand the critical epsilon value where the system no longer phase separates and the coursening rate, or the speed at which phase separation occurs. We have already analyzed these properties for a bulk polymer blend system and an unconfined composite, and plan to increase the confinement ration and continue this analysis. Student Aim 2: Understand the impact of polymer-NP interactions on the phase separation dynamics of these composites. A substantial fraction of polymers in nanoparticle packings are near nanoparticle surfaces. Thus, we hypothesize that polymer-nanoparticle interactions will have a significant influence on the behavior of the blends under extreme nanoconfinement. The student will explore this behavior by changing epsilon parameters between the polymers and particles in the simulation, which can create attractive or repulsive interactions between either of the polymers and the NPs. NP interactions in a system with high confinement may cause microphase separation of miscible polymer blends, where the polymer with favorable NP interactions coats the NP surfaces and the other polymer remains in the interstitial voids, and the student will seek out this and other interesting equilibrium structures. As with our confinement analysis, we will use the structure factor to analyze how the coursening rate, critical epsilon value, and overall structure of these composites changes with polymer-NP interactions. Student Aim 3: Study the transport properties of highly-confined polymer nanocomposites In addition to completing Aims 1 and 2 continued from the Fall 2021 EMPOWER project, we will begin to explore the transport properties of these composites. Our hypothesis is that the nanoscale morphology will drastically impact the transport of gas molecules. In particular, we believe that composites with very different transport properties can be produced by controlling the nanoscale structure of composites, particularly through the use of polymer-NP interactions. It is of particular interest to investigate composites with 2 polymers with different permeabilities. The student will add small gas molecules to our LAMMPS composite simulations and investigate the diffusivity through the composite and solubility into the composite. This work is very exciting for future applications of composites made with the CaRI method. This work ties in with the work of many graduate students within the mentor's two groups, and the student will be supported and encouraged by many students beyond the mentor.
Computational Resources	The group hosting this project has local computational resources that we regularly use. We have exclusive access to our group cluster, which was purchased in early 2011 and consists of 24 Dell C6100 nodes, each containing dual six-core Intel Xeon X5650 2.66GHz processors. This cluster is maintained by Computing and Engineering Technology Services (CETS) at Penn. Our local resources will be used for code development and debugging steps, while XSEDE will be used for production calculations. The group has an XSEDE allocation for the application and development of multi-scale simulation methods for the study of polymer nanocomposites and soft materials (Charge # TG-DMR150034). The EMPOWER student will use XSEDE's Stampede2 and Ranch supercomputers to run LAMMPS simulations of highly-filled polymer blend nanocomposites to understand both the dynamics and transport properties of these blends. The XSEDE resources are critical due to the long time-scales necessary to reach equilibrium and the complexity of the blend nanocomposite model.
Contribution to Community	This work is conducted with the goal of publication, which will acknowledge XSEDE. Both the mentor and student for this project are women, who have historically been underrepresented in computational work in engineering, and are passionate about outreach for underrepresented groups in STEM. In the spring of 2021, the mentor gave lessons on capillarity to high school students in Philadelphia and explained how XSEDE and computational resources could be applied to make discoveries in the field. We are currently working with schools to repeat this work in Spring 2022. The mentor and student plan to present our work on this project to local students in Philadelphia and introduce them to molecular dynamics and coding in Python, alongside encouraging students to pursue careers involving the work of XSEDE.
Position Type	Intern
Training Plan	The student will have access to the group's local computing resources and XSEDE resources to conduct simulations. The student is continuing this project, and has gained the relevant experience to work independently, but will still meet with the mentor weekly and the mentor and PI bi-weekly. The mentor is also available for additional meetings and assistance when necessary. The student will attend the weekly lab meetings of both of the mentor's lab, as much relevant work is conducted in both labs. The mentor has also created a library of relevant papers and resources for the project which will be provided to the student. The Penn Institute for Computational Science (PICS), an institute which both the mentor and PI are a part of, runs frequent tutorials on the tools and techniques of high-performance computing. The student will be highly encouraged to attend these as well as similar offerings by XSEDE.
Student Prerequisites/Conditions/Qualifications	here are few true prerequisites for this position, as the mentor has TA'd a class on molecular simulations and modeling that required no coding experience and mentored multiple undergrads with little to no prior coding experience and is thus confident in their ability to train a student who has never been introduced to simulation work. The mentor themselves had no coding experience prior to beginning their PhD and has found it a very enriching and fulfilling experience. However, there are a few qualifications that would make a student a stellar fit. A background in polymer science and in particular the thermodynamics of polymer blends would be useful. Any background knowledge of polymer nanocomposites would also be of use. Experience with python, data analysis and visualization, MD simulations, and Linux would make the onboarding process easier, but are again not necessary.
Duration	Semester
Start Date	01/12/2022
End Date	04/06/2022