Dear Members of the CRC1261 Board,
Thomas Barret received his B.Sc. degree in Mechanical Engineering from the University of Rochester. He then completed an M.Sc. in Mechanical Engineering, specializing in Mechanics and Design, at Northeastern University, where he also pursued a PhD in Materials Science and Engineering.
Thomas has a background in high performance nanocomposite materials, with his PhD focused on the processing and understanding of semi-crystalline polymer morphologies and their effects on mechanical behavior. During this work, he became familiar with polymer nanocomposite processing and development of nonwoven films and gel-spun fibers to supplement his primary computational focus. Much of his work utilizes a multiscale approach, and therefore requires deep understanding of various computational techniques, both in application and background theory. In developing his dissertation work, he primarily used molecular dynamics and finite elements, with these techniques supplemented with large amounts of programing in C++ and Python. These skills are good addition to CRC1261 and will supplement ongoing work through: (1) Molecular Dynamics and Machine Learning models for biosensor development and (2) multiscale, multiphysics, and finite element modeling, and general polymer processing. Given the overarching goal of developing bioinspired nanocomposites for surgical complication detection, advanced screening, measurement, and material handling techniques must be employed.
Thomas has extensive experience with molecular dynamics, both using LAMMPS and GROMACS. In 2023, he joined my research group as a visiting PhD researcher as a PACK International Research Fellow. During this time, he learned about our approach to Molecular Imprinted Polymer (MIP) development and performed computational simulations with the goal of coarse-graining the interactions between functional monomers and the biomarker IL-6. As a result of this stay, we have developed a dataset of more than 50 functional monomers in atomistic and coarse-grained form, and we are currently compiling and calibrating the results for publication. This work is fundamental in nature and will allow us to develop faster screening and modeling techniques for determining target-functional monomer interactions in the future.
During the PACK research stay, work also began on creating a polymerization model to better understand the interactions of functional monomers and cross-linkers during MIP production. The crosslinking step is crucial in creating a usable and repeatable product; however, it has not seen much attention in MIPs simulation due to its difficulty. Towards the end of his stay, Thomas developed a basic model for the pseudo-polymerization of a coarse-grained model, allowing us further insight into the chemical processes of MIP development. With additional time, this model can be applied to the complete functional monomer dataset, providing us a flexible and easily implemented model to study the MIP polymerization stage in-silico. This will significantly enhance batch-to-batch reproducibility of MIPs and facilitate the development of high-performance MIPs for real-world applications. This is especially crucial for the successful implementation of MIPs in gastroanastomotic leak cases.
Combined, the MIPs dataset and polymerization approach also provides a basis for machine learned approaches, with specific attention placed on diffusion models. Diffusion models are generative models that slowly add Gaussian noise to a system, before reverse engineering the random noise application. It is well suited to molecular simulation, as our standard binding simulations can be used to train the model, providing a faster generative approach to determining the necessary recipes for MIP development. Our group has successfully developed a machine learned screening model to determine the most suitable functional monomers for a specified target molecule. Combined with this diffusion model, we will be able to both determine the optimal functional monomer combination for each biomarker, and rapidly generate a range of theoretical binding sites to better understand and estimate the necessary processing steps to achieve a successful MIP.
Beyond the B12 project, Thomas’s background is well suited for collaboration with other projects, such as A2, “Hybrid magnetoelectric sensors based on mechanically soft composite materials”. Given Thomas’s background with polymers and polymer composites, Thomas can provide both experimental and simulation expertise to the modeling of micro- and nanoparticle filled soft composites. One such application might be the study of mechanical flexion resistance, where the hard, magnetic nanoparticles imbedded in the soft material behave differently under load than the surrounding matrix, potentially leading to debonding and void formation with repeated flexion loads. Using his background in multiscale modeling, Thomas could assist in developing a multiscale model to better understand the effect these particles have on the mechanical performance of the material, allowing us to make more accurate models.
Overall, Thomas would be an excellent addition to the B12 project and to the entire CRC. Further development of these models is of particular interest, as they directly contribute to the development and optimization of our MIPs and biosensors. In addition to this, Thomas’s background with polymer nanocomposite systems makes him well suited to assist with further development of our sensor technology, whether than be through films, fibers, or gels; all of which he has experience with.
Yours sincerely
Zeynep Altintas
Cost
The cost will be 30.000 Euros in total. This money should come from the lumpsum money of the CRC for 2025. The cosequence will be that we will have only 70.000 Euro to distribute at the end of year 2025.
Do you agree with the propsal of Zeynep Altintas
Your options are yes/no/abstention:
