Embedded Files
Research portfolio spanning molecular simulation, statistical mechanics, and data-driven polymer discovery across work at the American University of Sharjah, Brigham Young University, the University of South Florida, and the University of Wisconsin–Madison.
Polymer Simulation Benchmarks and Data-Driven Materials Discovery
Polymer Simulation Benchmarks and Data-Driven Materials Discovery
My current work at the University of Wisconsin–Madison focuses on building trustworthy molecular simulation workflows for polymer property prediction and materials discovery. Polymer informatics and AI-guided design promise to accelerate the search for new materials, but those approaches are only as reliable as the data used to train, validate, and benchmark them. For polymers, that data problem is especially difficult: properties depend on chemistry, chain architecture, molecular weight, processing history, morphology, and the simulation model used to represent them.
I use molecular dynamics, multiple atomistic force-fields, statistical mechanics, and automated analysis workflows to ask a deceptively simple question: when can molecular simulations be trusted to guide polymer design? This means benchmarking thermal, mechanical, and transport properties across diverse polymer chemistries; quantifying the uncertainty introduced by force-field choice and simulation protocol; and identifying which properties can be predicted robustly versus which require more careful molecular representation.
This work sits at the intersection of polymer physics, materials science, and artificial intelligence. By building high-quality simulation datasets and clarifying the limits of molecular modeling, I aim to support AI-enabled polymer discovery while keeping the physics honest: the goal is not simply to generate more data, but to generate data that is interpretable, reproducible, and useful for designing better materials.
Software Development for Molecular Dynamics Analysis
Software Development for Molecular Dynamics Analysis
A major part of my computational work involves developing and extending software tools for analyzing molecular dynamics simulations. I contributed to AMDAT — the Amorphous Molecular Dynamics Analysis Toolkit — an open-source C++ package for post-processing molecular dynamics trajectories of amorphous, glassy, polymeric, and complex-fluid systems.
AMDAT is designed for high-performance analysis of structure and dynamics, including observables such as radial distribution functions, structure factors, intermediate scattering functions, and neighbor correlations. The toolkit is especially useful for studying supercooled liquids and glass-forming materials, where long-timescale dynamics and heterogeneous motion require efficient trajectory handling and carefully designed sampling strategies.
This software-development work complements my broader research philosophy: molecular simulations are only as useful as the analysis tools used to interpret them. By helping build reusable, open-source infrastructure, I aim to make mechanistic simulation studies more reproducible, extensible, and accessible to the broader soft-matter and polymer physics communities.
Configurations of polymer nanocomposite system rendered using OVITO. a) Filler particles illustrated using different colors of gray, polymer beads using green, and crosslinker beads in sky blue. b) Only filler beads are shown with contacting beads (determined using AMDAT create_bin_list distance method) shown in red. c) Interfiller polymer, determined using find_between, are highlighted in apple green. d) Shells around filler particles (find_between) are shown in different colors.
Average counts and potential energies of different subsets/regions within the polymer nanocomposite computed using AMDAT’s custom_manual read-in method and create_bin_list / find_between methods for identifying contacting filler,interfiller polymer, and polymer shell regions. Each bar has a standard error from the average measurements at different frames.
Selected Outputs
[1] P. Kawak, W. F. Drayer, and D. S. Simmons, “AMDAT: An open-source molecular dynamics analysis toolkit for supercooled liquids, glass-forming materials, and complex fluids,” arXiv, arXiv:2602.05865, 2026, doi: 10.48550/arXiv.2602.05865.
[2] D. S. Simmons, P. Kawak, W. F. Drayer, and M. Mackura, “Amorphous Molecular Dynamics Analysis Toolkit (AMDAT),” Zenodo, 2025, doi: 10.5281/zenodo.17417166.
Origin of Toughness in Elastomeric Nanocomposites
Origin of Toughness in Elastomeric Nanocomposites
During my postdoctoral work in the David S. Simmons Lab at the University of South Florida, I used molecular simulations to investigate the molecular origins of toughness in elastomeric nanocomposites. These materials — rubbers reinforced with nanoscale fillers such as carbon black or silica — combine large stretchability with enhanced strength, but the physical mechanisms behind that reinforcement have remained difficult to isolate.
My work showed that nonlinear reinforcement arises from a coupled filler–polymer response during deformation. As the elastomer stretches, filler aggregates experience large local compressive stresses while the surrounding polymer network constrains their motion and maintains mechanical contact. This feedback between polymer incompressibility, filler structure, and local stress transmission helps explain how nanofillers reinforce soft elastomers at large strain.
a) Snapshots from stretching MD simulations at ε=0.0 (top, ai), ε=1.0 (middle, aii), and ε=2.0 (bottom, aiii) created with OVITO visualization software. Filler particle clusters are displayed in different shades of blue. To illustrate our stress spatial decomposition algorithm, filler beads near particle contacts ("Filler Near-Contact") are shown in black and polymer beads between two filler particles ("Interfiller Polymer") are shown in red. b) Spatially-resolved extensional stress response from "Interfiller Polymer" and "Filler Near-Contact" regions. c) Same as b, but with normal pressure. All scatter points include standard error bars calculated from 15 distinct deformation simulations.
This work received broader coverage for its implications in understanding why reinforced elastomers are strong, tough, and durable across applications ranging from tires to soft materials technologies.
Selected Press Coverage
[1] I. Dumé, “Why is rubber so resilient?” Physics World, May 19, 2026.
[2] “Study unlocks century-old riddle behind rubber’s strength,” Rubber Journal Asia, May 4, 2026.
[3] E. Roth, “After 100 Years, Engineers Finally Discover Why Rubber Is So Tough,” Gizmodo, 2026.
[4] TOI Science Desk, “How aeroplane tyres handle the massive impact of jet landings,” The Times of India, May 20, 2026.
[5] “Scientists solve 100-year-old mystery behind rubber that powers modern life,” Phys.org, Apr. 15, 2026.
[6] “Team cracks 100 year-old rubber mystery,” Futurity, Apr. 26, 2026.
Selected Outputs
[1] P. Kawak, H. Bhapkar, and D. S. Simmons, “Glassy interphases reinforce elastomeric nanocomposites by enhancing percolation-driven volume expansion under strain,” Proceedings of the National Academy of Sciences, 2026, doi: 10.1073/pnas.2528108123.
[2] P. Kawak, H. Bhapkar, and D. Simmons, “Dataset for Glassy interphases reinforce elastomeric nanocomposites by enhancing volume expansion under strain,” Dataset, 2026, doi: 10.17632/dg9jxp8sj6.
[3] P. Kawak, H. Bhapkar, and D. Simmons, “Young’s modulus and Poisson’s ratio versus temperature data from uniaxial extension molecular dynamics simulations of model elastomers,” Dataset, 2026, doi: 10.17632/c4bnns89g2.
[4] P. Kawak, H. Bhapkar, and D. Simmons, “Bulk modulus versus temperature data from equilibrium molecular dynamics simulations of a model neat elastomer,” Dataset, 2026, doi: 10.17632/rfrb249t77.
[5] P. Kawak, H. Bhapkar, and D. S. Simmons, “Origin of heating-induced softening and enthalpic recovery in elastomeric nanocomposites,” ACS Macro Letters, 2025, doi: 10.1021/acsmacrolett.5c00442.
[6] P. Kawak, H. Bhapkar, and D. S. Simmons, “Central role of filler–polymer interplay in nonlinear reinforcement of elastomeric nanocomposites,” Macromolecules, vol. 57, no. 19, pp. 9466–9478, 2024, doi: 10.1021/acs.macromol.4c00489.
Glass Transition of Sequence Specific Copolymers
Glass Transition of Sequence Specific Copolymers
As part of my postdoctoral work at the University of South Florida, I investigated how monomer sequence affects glass formation in copolymers. Copolymers can combine chemically distinct repeat units within a single chain, but their properties depend not only on composition, but also on sequence, stereochemistry, regioregularity, and local molecular packing. That design space is enormous, and its connection to glass transition behavior is difficult to predict from intuition alone.
Molecular simulations provide a way to isolate how sequence-specific interactions and local structure influence segmental dynamics and the glass transition temperature, Tg. This project used simulation-based screening and analysis to connect molecular sequence to glassy dynamics, helping clarify which sequence features are most important for tuning polymer thermal behavior.
Tg for system for a copolymer of Styrene and Methyl-Methacrylate simulated using the atomistic OPLS force-field.
More broadly, this work connects to my interest in choosing the right molecular representation for the question at hand: enough chemical detail to resolve sequence effects, but enough computational efficiency to explore broad design spaces.
Schematic of the enabling technology, the predictive stepwise quench algorithm, that allows efficient and accurate simulation of Tg.
Crystal Nucleation in Polymer Melts
Crystal Nucleation in Polymer Melts
Flowchart illustrating the connections between the methods used and the presented data. Yellow rectangles represent the methods, and green ovals correspond to one or more figures in this paper.
During my Ph.D. work at Brigham Young University with Douglas R. Tree, I used advanced Monte Carlo simulations to study crystal nucleation and phase behavior in polymer melts. This work focused on how chain stiffness, orientational ordering, and crystallinity combine to produce ordered phases from disordered melts.
By calculating free-energy landscapes and analyzing molecular configurations, I showed that semiflexible oligomers can crystallize through a cooperative transition involving coupled nematic and crystalline ordering. This work helped clarify how molecular stiffness reshapes the pathway from melt disorder to crystalline order.
(i) Representative MCMC snapshots of configurations and [(ii)–(iv)] averaged 2D structure factors for (a) a disordered melt phase at ϕ = 0.471 and Tr = 2.0, (b) a nematic phase at ϕ = 0.407 and Tr = 0.001, and (c) a crystal phase at ϕ = 0.471 and Tr = 0.001. All qi are in units of σ −1 and qx , qy , qz ∈ [−4.2π, 4.2π].
Selected Outputs
[1] P. Kawak, C. Akiki, and D. R. Tree, “Effect of local chain stiffness on oligomer crystallization from a melt,” Physical Review Materials, vol. 8, 2024, doi: 10.1103/PhysRevMaterials.8.075606.
[2] P. Kawak, “Simulation of crystal nucleation in polymer melts,” Ph.D. dissertation, Brigham Young University, Provo, UT, USA, 2022.
[3] P. Kawak, D. S. Banks, and D. R. Tree, “Semiflexible oligomers crystallize via a cooperative phase transition,” The Journal of Chemical Physics, vol. 155, 2021, doi: 10.1063/5.0067788.
Ultrasonically controlled estrone-modified liposomes for estrogen-positive breast cancer therapy
Ultrasonically controlled estrone-modified liposomes for estrogen-positive breast cancer therapy
During my M.S. work in the Ghaleb Husseini Lab at the American University of Sharjah, I studied ultrasound-responsive liposomal drug carriers for targeted cancer therapy. The goal was to design liposomes that could encapsulate chemotherapeutic cargo, reduce nonspecific toxicity, preferentially interact with estrogen-receptor-positive breast cancer cells, and release their contents in response to ultrasound.
This project combined soft materials, drug delivery, and biomedical engineering. While it sits outside my current focus in polymer physics, it shaped my broader interest in using molecular-scale structure and physical triggers to control material function.
Key design goals included selective delivery, reduced toxicity to healthy cells, estrone-mediated targeting, and ultrasound-triggered release.
An illustration of a liposomal nanocarrier.
An illustration of liposomal nanocarriers entering a tumor microenvironment and binding with tumor cells' receptors.
Selected Outputs
[1] P. Kawak, “Ultrasound triggered release of estrone-targeted liposomes,” M.S. thesis, American University of Sharjah, Sharjah, United Arab Emirates, 2017.
[2] N. M. Salkho, V. Paul, P. Kawak, R. F. Vitor, A. Martins, M. H. Al Sayah, and G. Husseini, “Ultrasonically controlled estrone-modified liposomes for estrogen-positive breast cancer therapy,” Artificial Cells, Nanomedicine, and Biotechnology, 2018, doi: 10.1080/21691401.2018.1459634.
Page updated
Google Sites
Report abuse