Elsevier

Chemical Engineering Science

Volume 64, Issue 24, 16 December 2009, Pages 5182-5191
Chemical Engineering Science

Optimum structured adsorbents for gas separation processes

https://doi.org/10.1016/j.ces.2009.08.029Get rights and content

Abstract

Recent developments in separation technology by adsorption have included the development of new structured adsorbents which offer some attractive characteristics compared to a typical packed bed. These improved features include lower energy consumption, higher throughput and superior recovery and purity of product. However, the exact combination of structural, geometric parameters which yields optimum performance is unknown. This study formulates a methodology for comparison based on a variety of analytical and numerical models and uses it to examine the performance of different adsorbent configurations. In particular, monolithic, laminate and foam structures are evaluated and compared to a packed bed of pellets. The effects of physical adsorbent parameters which govern the performance of a PSA process are considered during model development. Comparisons are carried out based on mass transfer kinetics, adsorbent loading and pressure drop of a PSA system for CO2/N2 separation. The results indicated that structured adsorbents can provide superior throughput to packed beds provided their geometrical parameters exceed certain values. For example, laminate structures can offer superior performance to a packed bed of pellets only if the critical sheet thickness and spacing are less than about 0.2 mm. Each adsorbent structure should be designed to operate at its “optimal” velocity. When operating at velocities higher than the “optimal” value, the increase in pressure drop and length of the mass transfer zone more than offsets gains accrued through reduction in cycle time.

Introduction

Efficient gas separation processes operating at high throughput are compromised when adsorbent materials in the form of beads or granules are used. High pressure drop associated with gas flow through a packed bed of beads/pellets and mass transfer limitations related to gas diffusion into or out of the beads rapidly reduce system performance. To address these drawbacks, different adsorbent structures may be considered.

Recently, various structured adsorbents with enhanced adsorption characteristics such as monolithic, laminate and foam structures have gained considerable attention as substitutes for traditional adsorbent particles (Kodama et al., 1993; Maurer, 1994; Gadkaree, 1997; Li et al., 1998; Golden et al., 2003; Brandani et al., 2004; Keefer et al., 2004; Sawad et al., 2005; Golden et al., 2005; Zabka et al., 2006; Grande et al., 2006; Rode et al., 2007).

Parallel channel monolithic structures with controllable shape, cell density and wall thickness have been reported for their use in adsorptive gas separation systems (Kodama et al., 1993; Gadkaree, 1997; Li et al., 1998; Lee et al., 2000; Yu et al., 2002; Yates et al., 2003). The main advantage of such configurations resides in their low pressure drop and higher mass transfer rates. Although it has been reported that the mass transfer characteristics of monolithic adsorbents is somewhat inferior to a comparable packed bed, this is more a consequence of early monolithic structures with thick walls rather than any intrinsic limitation. It would be possible to improve the mass transfer behaviour of monoliths by utilizing structures with reduced channel width and wall thickness or more appropriate shape. Patton et al. (2004) suggests the employment of hexagonal channels rather than ordinary square or circular channels. These authors applied the linear driving force model (LDF) to optimize practical monolith channels taking mass transfer and pressure drop characteristics into account. In recent study performed by Zabka et al. (2008), the performances of the separation of chiral species by the simulated moving bed (SMB) unit packed with conventional adsorbent beads and monolithic adsorbents were compared. The authors concluded that the selection of particle diameter or different bed morphologies is a trade-off between the productivity and eluent consumption.

The use of parallel laminate structures for adsorption processes is rather new with a few patents describing the use of adsorbent materials in the form of laminate sheets in adsorptive gas separation applications. Rode et al., 2007 demonstrated the application of improved adsorbent sheet based parallel passage structures for PSA, TSA and partial PSA devices. In order for laminates to be effective adsorbents, the thickness of the sheets and the space between adjacent sheets must be as small as possible but not so small so as to result in excessive pressure drop. Ruthven and Thaeron (1996) carried out an experimental and theoretical study on the performance of a parallel passage contactor based on HETP (height equivalent to a theoretical plate) and pressure drop. Their results emphasize the advantage of such an adsorber over the conventional beads for applications in which pressure drop is a key factor.

Ceramic foams have many attractive features as catalyst supports; however, their utilization in adsorptive gas separation applications is scarce. Open cell foams comprising a network of solid struts and pores offer substantial benefits over configurations comprising external supports; namely, a low pressure drop operation with tortuous flow path and high geometric surface area which enhance the rate of external mass transfer (Richardson et al., 2003; Patcas et al., 2007).

Recently, Patcas et al. (2007), performed an experimental comparison of catalysts with different carriers, namely, honeycomb monoliths, ceramic foams and spherical particles based on their pressure drop and mass and heat transfer characteristics in the oxidation of carbon monoxide. Their results showed that the performance of foam catalysts with respect to combined high mass transfer and low pressure drop were superior over particles and inferior to that of honeycomb structures.

In our recent review of the use of structured adsorbents (Rezaei and Webley, 2009), we identified the important structural parameters that impact the overall performance of gas separation processes. These are: (1) amount of adsorbent contained in a given volume (adsorbent loading), (2) pressure drop through the structure per unit length, (3) external surface area per unit volume, (4) total void volume and (5) channelling and dispersion parameter characteristics. Some of these parameters are constants of the zeolite structure (1), (3), (4) while others also depend on gas velocity (2), (5).

As is well established, for cyclic processes such as PSA or TSA, reducing cycle time is the most important factor to improving the process in terms of reducing adsorbent inventory and cost since shorter cycles result in smaller beds. However, product recovery often decreases as cycle time is decreased unless adsorption rate is correspondingly increased (Ackley et al., 2003). In order to capture this feature, the well known mass transfer zone (MTZ) concept (also referred to as length of unused bed or LUB) was used here to assess the performance of a range of adsorbent structures.

It is clear that the parameters of a particular adsorbent structure will influence the separation performance and cost. It is also evident that a trade-off will exist between mass transfer, pressure drop and adsorbent loading as the geometry of the adsorbent structures are altered. It follows that for each structure there is an optimum set of geometric parameters which will provide the “best” performance as defined by a particular objective function. It is the goal of this study to develop the analytical framework required to determine this optimum set of geometric parameters and to use this model to examine laminate, monolith and foam structures.

Section snippets

Mass transfer models

A number of simplified mathematical models were developed for different geometries to describe the adsorption of a single component in an inert carrier. A differential mass balance for the adsorbate in the gas phase flowing through a structure (independent of geometry) givesεbCt-DL2Cz2+UCz=-k¯a(C-C¯p)where C and C¯p are the concentration of the adsorbate in the gas phase and the average concentration in the pores of the adsorbent, respectively, a is specific surface area per unit bed

External surface area comparison

One of the most important parameters which govern the performance of an adsorbent is its external surface area per unit volume, since the mass transfer rate is directly proportional to this parameter. In Fig. 2, the comparison between the external surface areas per unit volume of 1 mm bead diameter, 45 PPI foam and different voidage monolithic structures is shown. It is clear that small beaded packed beds and foam adsorbents have very high external surface area which is difficult to match with

Conclusion

In this study, some important features of different adsorbent structures were derived and their performance compared. Mathematical models were developed based on structural and geometric parameters that impact overall system performance such as voidage, external surface area per unit volume and bulk density. Limiting LDF type expressions were derived for laminates, monoliths and foam structures. Pressure drop and mass transfer characteristics were investigated. Based on the simulation results,

Notation

aspecific surface area per unit bed volume, cm2/cm3
Cbulk concentration, g mol/cm3
C¯paverage solid concentration, g mol/cm3
Cpgas concentration in voids in adsorbent, g mol/cm3
Csgas concentration at adsorbent surface, g mol/cm3
Deeffective diffusivity, cm2/s
DLaxial dispersion coefficient, cm2/s
Dmmolecular diffusion coefficient, cm2/s
kequation constant parameter
k¯total mass transfer coefficient, cm/s
kequation constant parameter
kffilm mass transfer coefficients, cm/s
kppore mass transfer coefficients,

Acknowledgement

The authors would like to acknowledge the Monash Research Graduate School (MRGS) for providing scholarship support for Fateme Rezaei.

References (34)

  • F. Brandani et al.

    Adsorption kinetics and dynamic behavior of a carbon monolith

    Adsorption

    (2004)
  • Cybulski, A., Moulijn, J.A., 1998. Structured Catalysts and Reactors, p....
  • S. Ergun

    Fluid flow through packed columns

    Chemical Engineering Progress

    (1952)
  • Gadkaree, K.P., 1997. System and method for adsorbing contaminants and regenerating the adsorber, United States Patent...
  • L. Giani et al.

    Mass-transfer characterization of metallic foams as supports for structured catalysts

    Industrial and Engineering Chemistry Research

    (2005)
  • L.J. Gibson et al.

    Cellular Solids: Structure and Properties

    (1997)
  • Golden, T.C., Golden, C.M.A., Zwilling, D.P., 2003. Self-supporting absorbent fabric for gas separation, United States...
  • Cited by (150)

    View all citing articles on Scopus
    View full text