(249c) Bridging Molecular Insights to Industrial Scale: Developing a Multiscale Modeling Framework for Predicting Catalyst Deactivation for Dry Reforming of Methane Process

The process of dry reforming of methane (DRM) presents a promising strategy for converting greenhouse gases CH4 and CO2 into syngas, a valuable precursor for producing long-chain alkanes through Fischer-Tropsch synthesis[1], [2]. A reactor scale model is essential for optimizing DRM processes by simulating and improving packed bed reactor performance based on operating conditions. However, the formation of carbon deposits under reforming conditions can quickly deactivate these catalysts, posing a significant challenge[3]. It is crucial to delve into the reaction mechanisms of DRM under real conditions through detailed kinetic studies at a microscale for identifying and mitigating catalyst deactivation mechanisms, enhancing process efficiency of nickel-based catalysts [4]. Previously, the integration of multiscale models was hindered by undefined kinetics for carbon diffusion on Ni catalysts [5]. A pellet scale model is needed to bridge microscale reactions and macroscale reactor behavior, thereby predicting the coke whisker formation [6].

Addressing this need, our study presents a comprehensive multiscale model that simulates the DRM process from atomic-scale interactions to reactor-scale phenomena. The model incorporates a kinetic Monte Carlo (kMC) framework, detailing the adsorption, reaction, and desorption steps of 33 reversible reactions on nickel catalyst surfaces [7]. At the pellet scale, the model employs a Whisker Model to monitor the evolution of coke whiskers, predicting their growth length based on reaction kinetics based on the kMC simulations. This model considers factors such as the surface area of catalyst pellets and diffused carbon as unreactive coke, delivering granular insights into coking dynamics. The reactor scale component leverages a modified Ergun Equation[8], integrating the predicted pressure drop and resulting changes in catalyst bed porosity due to coke accumulation which in turn affects the kMC models reaction kinetics. The sheer number of possible reaction pathways and the stochastic nature of kMC simulations, require considerable computational resources to accurately capture the dynamics of coking at the atomic scale.

This computational challenge is addressed through the methodological advancement of applying a coarse time stepper facilitating the dynamic interaction between the microscale and macroscale phenomena, which drastically improves computational efficiency [9]. The model demonstrates that increasing the coarse time stepper from 1 second to 100 seconds reduces the computation time from 17 days to 6 hours, with minimal loss in accuracy, as validated by the close congruence of simulated coke growth for a simulated time of an hour.

Through this modeling approach, we achieve a predictive tool that accurately simulate the DRM process from the atomic scale of catalyst interaction up to the operational scale of industrial reactors, taking into account the complex dynamics of coke whisker formation and its impact on reactor performance. This work deepens the understanding of the complex interplay between reaction kinetics and catalyst deactivation and offers a pragmatic solution to one of the major challenges in industrial scale DRM operations which were previously unattainable.

References

[1] K. Wittich, M. Krämer, N. Bottke, and S. A. Schunk, “Catalytic Dry Reforming of Methane: Insights from Model Systems,” ChemCatChem, vol. 12, no. 8, pp. 2130–2147, Apr. 2020, doi: https://doi.org/10.1002/cctc.201902142.

[2] J. Wei and E. Iglesia, “Isotopic and kinetic assessment of the mechanism of reactions of CH4 with CO2 or H2O to form synthesis gas and carbon on nickel catalysts,” J Catal, vol. 224, no. 2, pp. 370–383, Jun. 2004, doi: 10.1016/J.JCAT.2004.02.032.

[3] B. Yuan, T. Zhu, Y. Han, X. Zhang, M. Wang, and C. Li, “Deactivation Mechanism and Anti-Deactivation Measures of Metal Catalyst in the Dry Reforming of Methane: A Review,” Atmosphere (Basel), vol. 14, no. 5, 2023, doi: 10.3390/atmos14050770.

[4] M. Andersen, C. Panosetti, and K. Reuter, “A Practical Guide to Surface Kinetic Monte Carlo Simulations,” Front Chem, vol. 7, 2019, doi: 10.3389/fchem.2019.00202.

[5] S. Helveg et al., “Atomic-scale imaging of carbon nanofibre growth,” Nature, vol. 427, no. 6973, pp. 426–429, 2004, doi: 10.1038/nature02278.

[6] S. Helveg, J. Sehested, and J. R. Rostrup-Nielsen, “Whisker carbon in perspective,” Catal Today, vol. 178, no. 1, pp. 42–46, Dec. 2011, doi: 10.1016/J.CATTOD.2011.06.023.

[7] C. Fan, Y.-A. Zhu, M.-L. Yang, Z.-J. Sui, X.-G. Zhou, and D. Chen, “Density Functional Theory-Assisted Microkinetic Analysis of Methane Dry Reforming on Ni Catalyst,” Ind Eng Chem Res, vol. 54, no. 22, pp. 5901–5913, Jun. 2015, doi: 10.1021/acs.iecr.5b00563.

[8] A. G. Dixon, “General correlation for pressure drop through randomly-packed beds of spheres with negligible wall effects,” AIChE Journal, vol. 69, no. 6, p. e18035, 2023, doi: https://doi.org/10.1002/aic.18035.

[9] C. Theodoropoulos, Y. H. Qian, and I. G. Kevrekidis, “‘Coarse’ stability and bifurcation analysis using time-steppers: A reaction-diffusion example,” Proceedings of the National Academy of Sciences, vol. 97, no. 18, pp. 9840–9843, Aug. 2000, doi: 10.1073/PNAS.97.18.9840.