(371g) Identification of Hybrid Population Balance Models for Mechanochemical Depolymerization

Mechanochemical depolymerization of waste plastics is a method of chemical recycling that successively breaks polymer chains via mechanical impacting. By leveraging local high-energy collisions, mechanochemical milling can reduce the need for process heating in plastics recycling [1, 2]. Since the milled plastics leaving these processes require downstream separations and polymer reconstructions, detailed models predicting molecular weight distributions (MWDs) of mechanochemical mills are necessary to incorporate them into flowsheet analysis and broader process optimization [3, 4]. This presentation focuses on modeling the temporal evolution of the MWD of polystyrene (PS) pellets in a lab-scale vibratory ball mill with a hybrid Population Balance Model (PBM) that considers diverse experimental and simulation data.

The hybrid model uses the traditional PBM structure to track the evolution of discrete MWD classes by modeling their rates of change as a set of coupled ordinary differential equations (ODEs). The hybrid PBM is parameterized with physically interpretable values by using a hybrid-mechanistic kernel with parameters determined by high-fidelity Discrete Element Method (DEM) simulations of the milling process and experimental reaction data. DEM simulations capture the kinematics of the moving entities within the mill. Although hybrid DEM-PBM models have been developed for ball-mills, models that capture coupled reaction with milling have added complexity that has not been explored before [5, 6]. This work explores how to identify the type and form of the hybrid-mechanistic kernel and the overall form of the PBM for reactive milling. To do so, we investigate and compare different methods for parameter estimation of these large dynamic models, by solving the inverse problem using equation-based, surrogate-based and stochastic optimization methods.

This work lays a foundation for the implementation of hybrid PBMs to describe mechanochemical depolymerization recycling processes with a physically interpretable model, which can be used for process design and industrial scale-up in future research.

Citations

[1] A. W. Tricker, G. Samaras, K. L. Hebisch, M. J. Realff, and C. Sievers, “Hot spot generation, reactivity, and decay in mechanochemical reactors,” Chemical Engineering Journal, vol. 382, p. 122954, Feb. 2020, doi: 10.1016/j.cej.2019.122954.

[2] A. W. Tricker et al., “Stages and Kinetics of Mechanochemical Depolymerization of Poly(ethylene terephthalate) with Sodium Hydroxide,” ACS Sustainable Chem. Eng., vol. 10, no. 34, pp. 11338–11347, Aug. 2022, doi: 10.1021/acssuschemeng.2c03376.

[3] E. Anglou et al., “Process development and techno-economic analysis for mechanochemical recycling of poly(ethylene terephthalate),” Chemical Engineering Journal, vol. 481, p. 148278, Feb. 2024, doi: 10.1016/j.cej.2023.148278.

[4] E. Anglou, Y. Chang, A. Ganesan, S. Nair, C. Sievers, and F. Boukouvala, “Discrete Element Simulation and Economics of Mechanochemical Grinding of Plastic Waste at an Industrial Scale,” in Computer Aided Chemical Engineering, vol. 52, A. C. Kokossis, M. C. Georgiadis, and E. Pistikopoulos, Eds., in 33 European Symposium on Computer Aided Process Engineering, vol. 52. , Elsevier, 2023, pp. 2405–2410. doi: 10.1016/B978-0-443-15274-0.50382-6.

[5] M. Capece, R. N. Davé, and E. Bilgili, “A pseudo-coupled DEM–non-linear PBM approach for simulating the evolution of particle size during dry milling,” Powder Technology, vol. 323, pp. 374–384, Jan. 2018, doi: 10.1016/j.powtec.2017.10.008.

[6] N. Metta, M. Ierapetritou, and R. Ramachandran, “A multiscale DEM-PBM approach for a continuous comilling process using a mechanistically developed breakage kernel,” Chemical Engineering Science, vol. 178, pp. 211–221, Mar. 2018, doi: 10.1016/j.ces.2017.12.016.