AIChE Annual Meeting
Tuesday, November 9, 2021 - 4:10pm to 4:40pm
Many of the most interesting chemical and materials problems---including those related to optical circuitry, batteries, drug delivery, and gas storage--constitute multi-component, often highly complex, systems. Historically, it is difficult and costly to computationally model multi-component and multiscale systems, and many studies are limited to short time scales or reduced degrees of freedom. Therefore, when one of these complex systems has been simulated at the bounds of these limitations, it is even more crucial to extract any information contained within the data, as has often been the focus in the machine learning community.
In this poster, I will preview my vision to develop and implement physics-inspired representations for non-atomic components, expanding successful practices from the atomistic modeling community to include aspects of anisotropy and hierarchy. These models will ultimately lead to a unified multiscale feature space and enable predictive modeling of complex and hierarchical symmetries, ranging from small molecules to complex mixed-scale systems.
PhD Research (2014-2019):
I completed my doctoral work at the University of Michigan (UM) in the lab of Prof. Sharon C. Glotzer. As a part of my doctoral work, I investigated the role of nanoparticle geometry in the spontaneous formation of order, i.e., self-assembly. I proved the role of âpacking argumentsâ in the self-assembly of nanoparticles (published in PNAS), predicted phase transitions that produce multi-state optical materials, and drew connections between phase phenomena at the atomic and nanoscale (published in Phys. Rev. Materials). During this time, I also secured additional independent funding to investigate robust targets for nanoscale photonic materials, resulting in a dataset of 150,000 computations and constituting the first in-depth study of design heuristics for 3D colloidal photonics, as published in Nature Communications. I received several University of Michigan distinctions for this work (including the Charles G. Overberger Award for Excellence in Research and the Biointerfaces Innovator Award) and the Victor K. LaMer Dissertation Award from the Colloids Division of the American Chemical Society.
PostDoc Research (2019-2021):
I am currently a postdoctoral researcher at École Polytechnique Fédérale de Lausanne (EPFL), alongside Prof. Michele Ceriotti in the Laboratory of Computational Science and Modeling (COSMO). My most recent work has focused on developing machine learning methods for atomic systems (published in Machine Learning, Science and Technology),. These efforts include developing open-source software, and the employment of these methods for learning the behavior of hierarchical materials.
Teaching Style and Interests:
In teaching, my goal is to inspire students to develop foundational knowledge, expand and apply it to new problems, and communicate it in educational and professional settings. I am often motivated by the idea that ``A failure of understanding starts with a failure of communication.'' In my acting experience, it is the actor's responsibility to make an audience laugh, not the audience's to get the joke. I believe that this shift in "the burden of communication" is important in the classroom to provide an openness for students to ask questions and helps to combat the feeling of âothernessâ to which students traditionally under-represented in engineering classrooms are particularly susceptible.
My educational background lies in macromolecular and polymer sciences, materials science and engineering, computer science, and machine learning; and I would happily teach courses related to computational modeling and analysis, polymer science, and statistical mechanics and thermodynamics. I have trained in scientific communication with the inter-university organization Engineering Ambassadors, UM Natural History Museum Science Communication Fellows program, and through outreach and communications programs that I have spear-headed such as ACS POLY/PMSE Student Outreach and the REACT Workshop. Based upon this and my theater background, I would also welcome the chance to teach any classes on science communication and presentation skills.
 R. K. Cersonsky, G. van Anders, P. M. Dodd, and S. C. Glotzer, "Relevance of Packing in Colloidal Self-Assembly," (2018). Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.1720139115
 Y. Zhou, R. K. Cersonsky, S. C. Glotzer, "A New Route to the Diamond Colloidal Crystal.â Under Review.
 R. K. Cersonsky, J. Dshemuchadse, J. Antonaglia, G. van Anders, S. C. Glotzer, "PressureâTunable Band Gap in an Entropic Crystal", (2018). Phys. Rev. Mat. https://doi.org/10.1103/PhysRevMaterials.2.125201
 R. K. Cersonsky, J. Antonaglia, B. D. Dice, S. C. Glotzer, "Unexpected Diversity of Three-Dimensional Photonic Crystals.â Nature Communications. https://www.nature.com/articles/s41467-021-22809-6
 R. K. Cersonsky, B. Helfrecht, E. A. Engel, S. Kliavinek, M. Ceriotti, "Improved Data Sub-selection with Principal Covariates Regression." Machine Learning: Science and Technology. https://iopscience.iop.org/article/10.1088/2632-2153/abfe7c
 B. Helfrecht, R. K. Cersonsky, G. Fraux, M. Ceriotti, "Structure-property mapping with Kernel Principal Covariate Regression," (2020) Machine Learning: Science and Technology. https://doi.org/10.1088/2632-2153/aba9ef
G. Fraux, R. K. Cersonsky, M. Ceriotti, "Chemiscope: interactive structure-property explorer for materials and molecules." (2020). Journal of Open Source Software. https://doi.org/10.21105/joss.02117
R. K. Cersonsky, M. Pahknova, M. Ceriotti, "Identifying High-Stability Components of Molecular Crystals." In Preparation.
 A. Travitz, A. Muniz, J. K. Beckwith, and R. K. Cersonsky. âBringing Science Education and Research together to REACT.â Proceedings of the American Society for Engineering Education, 2020. https://peer.asee.org/35030