(94e) Monitoring Dispersion and Heat Effects in Dynamic Column Breakthrough Experiments with Different Particle Shapes Using the Digital Adsorption Method

Industrial gas adsorption separations are commonly carried out as cyclic processes, in which the adsorbent bed undergoes alternately saturation and regeneration. By enabling independent study of each individual step in the cycle, dynamic column breakthrough (DCB) experiments remain a fundamental component in the design of these systems [1]. A major challenge to achieve a predictive description of these experiments is the estimation of unknown parameters in either the mass or energy balance partial differential equations.

One of such parameters is the axial dispersion coefficient [2]. The commonly adopted estimations of the dispersion coefficient account, depending on the Reynolds number, for contributions from molecular diffusivity and eddy diffusion, i.e., turbulent mixing [3]. Although, a dependence of the axial dispersion on the shape and size of adsorbent particle therefore exists [4], the effect of these properties, particularly the shape, on breakthrough results has not been fully understood. Additionally, depending on the solid/gas heat capacity ratio and the equilibrium sorption/gas concentration step ratio, different velocities of the thermal and mass transfer front are possible [5], which in turn must result in varying dispersion of the mass-transfer zone along the adsorbent bed.

The common estimation of the dispersion coefficient can in principle reflect such local variations. However, this requires information about the local internal conditions, which are difficult to obtain experimentally. In practice, numerical models are implemented, in which this coefficient is commonly considered an adjustable parameter to fit classical experimental breakthrough curves [1, 2]. However, recently even this practice has been demonstrated to potentially produce inaccurate estimations of the axial dispersion coefficient due to secondary effects such as flow channeling, and that the more reliable approach is to measure breakthrough curves inside the bed for the fitting [6]. This highlights the need to further assess the accuracy of estimations of the axial dispersion coefficient through accessing transient internal processes in the packed bed.

With the digital adsorption (DA) framework, an experimental method has been introduced that allows the measurement of internal, transient concentration profiles through the application of X-ray computed tomography [7]. Here, we utilize this technique to monitor the internal adsorption profiles during CO2-He DCB experiments with zeolite 13X particles of different shapes, spherical beads, and cylindrical rods, respectively. The packed bed’s length is 273mm and its diameter 30mm. In total six, i.e. three per particle shape, DCB experiments are conducted at the same CO2 inlet flow rate of 100ccm, and CO2 inlet concentration of 1.0, 0.5, and 0.25, which are achieved through varied He inlet flow rates. All DCB experiments are conducted under ambient conditions T ≈ 293.25 K and p ≈ 99 kPa. The DCB product is measured using mass a flow meter and CO2 analyzer, while the internal temperature is simultaneously monitored at four different axial locations along the center of the bed, at 1/8, 3/8, 5/8, and 7/8 bed lengths, respectively.

Using DA and temperature measurements, the propagation velocities of the mass transfer and thermal fronts are determined independently for each experiment. By analyzing the temporal change of adsorbed amount and of the temperature at the exact same locations, the relation between heat effects and locally varying dispersion is established. Breakthrough response, internal temperature, and concentration profiles are compared for the two different particle shapes. Based on these results, the role of heat convection through the bulk gas phase and conduction through the solid adsorbent phase and their impact on dispersion and the mass transfer zone is discussed. A comparison is made between experimental breakthrough curves measured within, through the DA measurements, and at the outlet of the adsorption column, and their suitability to extract dispersion coefficients is assessed.

1. Wilkins, N. S., Rajendran, A., Farooq, S. (2021). Dynamic column breakthrough experiments for measurement of adsorption equilibrium and kinetics. Adsorption, 27(3), 397–422.
2. Knox, J. C., Ebner, A. D., Levan, M. D., Coker, R. F., Ritter, J. A. (2016). Limitations of Breakthrough Curve Analysis in Fixed-Bed Adsorption. Industrial and Engineering Chemistry Research, 55(16), 4734–4748.
3. Edwards, M. F., Richardson, J. F. (1968). Gas dispersion in packed beds. Chemical Engineering Science (Vol 23).
4. Rasmuson, A. (1985). The effect of particles of variable size, shape and properties on the dynamics of fixed beds. Chemical Engineering Science (Vol. 40, Issue 4).
5. Farooq, S., Ruthven, D. M. (1990). Heat Effects in Adsorption Column Dynamics. 1. Comparison of One-and Two-Dimensional Models. In Ind. Eng. Chem. Res (Vol 29).
6. Son, K. N., Weibel, J. A., Knox, J. C., & Garimella, S. v. (2019). Limitations of the Axially Dispersed Plug-Flow Model in Predicting Breakthrough in Confined Geometries. Industrial and Engineering Chemistry Research, 58(9), 3853–3866. https://doi.org/10.1021/acs.iecr.8b05925
7. Pini, R., Joss, L., Hosseinzadeh Hejazi, S. A. (2021). Quantitative imaging of gas adsorption equilibrium and dynamics by X-ray computed tomography. Adsorption, 27(5), 801–818.