Abstract
The sorption kinetics of some aromatics (toluene, p-xylene, o-xylene and p-diethylbenzene) was studied experimentally with a high sensitivity microbalance (± 1µg) at temperatures between 150 and 210ºC and concentrations in the adsorbed phase of a maximum of 1 mmol/g. In general, the sorption rates decreased in the following order: toluene > p-xylene > o-xylene > p-diethylbenzene. Two diffusion models were tested against the experimental results. One model considers the mass transfer inside the zeolitic micropores as the controlling diffusion step. The other model supposes that resistance to mass transfer in the macropores (formed during pelletization of the crystals with the amorphous ligand) is the limiting step. The results observed for both models are presented and analyzed
Sorption kinetics; gravimetric method; diffusion models
SORPTION KINETICS OF AROMATICS IN Y ZEOLITE PELLETS USING THE GRAVIMETRIC METHOD
C.L. Cavalcante Jr. 1, * * To whom correspondence should be addressed. , V.E. Lima 2 , I.G. Souza 1 and O.L.S. Alsina 2
1Universidade Federal do Ceará - Departamento de Engenharia Química - GPSA - Grupo de Pesquisa em
Separação por Adsorção - Campus do Pici, Bl. 710 - Fortaleza,CE - 60.455-760, Brazil
Fax: (085) 288-9601, Phone: (085) 288-9611
2Universidade Federal da Paraíba - Departamento de Engenharia Química - Campus II, Aprígio Veloso, 882
58109-970 - Campina Grande, PB, Brazil - Fax: (083) 333-1650
(Received: March 5, 1997; Accepted: August 5, 1997)
Abstract: The sorption kinetics of some aromatics (toluene, p-xylene, o-xylene and p-diethylbenzene) was studied experimentally with a high sensitivity microbalance (± 1µg) at temperatures between 150 and 210oC and concentrations in the adsorbed phase of a maximum of 1 mmol/g. In general, the sorption rates decreased in the following order: toluene > p-xylene > o-xylene > p-diethylbenzene. Two diffusion models were tested against the experimental results. One model considers the mass transfer inside the zeolitic micropores as the controlling diffusion step. The other model supposes that resistance to mass transfer in the macropores (formed during pelletization of the crystals with the amorphous ligand) is the limiting step. The results observed for both models are presented and analyzed.
Keywords: Sorption kinetics, gravimetric method, diffusion models.
INTRODUCTION
Separation and purification of xylene isomers is of great interest to the petrochemical industries, primarily for the production of p-xylene, due to its importance for textile and industrial fibers manufacturing (Santacesaria et al., 1982).
The separation of aromatics using Y zeolites has been widely studied during the last 10 years (Broughton et al., 1970; Storti et al., 1985; Morbidelli et al., 1986; Ruthven and Goddard, 1986; Goddard and Ruthven, 1986a,b; Cavalcante and Gubulin, 1990; Marra Jr., 1991; Neves, 1995). Most of these studies report adsorption in zeolite crystals. However, industrial applications generally use pellets to avoid high pressure drops in the commercial adsorbers.
This paper will present experimental kinetic data obtained for the adsorption of toluene, p-xylene, o-xylene and p-diethylbenzene in pellets of zeolite Y, using the gravimetric method.
SORPTION KINETICS
For pelletized adsorbents, several diffusional resistances may be identified (Ruthven, 1984):
- micropore resistance in the zeolitic crystals;
- macropore resistance in the extracrystalline channels formed by the pelletization process;
- fluid film resistance, around the pellet.
Kärger and Ruthven (1992) present several models for sorption kinetics in zeolites, considering different alternatives for the controlling step of the diffusion process. The models that will be tested in this study will now be presented.
Micropore Diffusion
If intracrystalline diffusion controls the process, and there is no significant change in the adsorbed phase concentration (Dc can be assumed to be constant), the transient diffusion equation may be written for a spherical particle (Ruthven, 1984) as follows:
(1)
The initial and boundary conditions for a gravimetric uptake experiment are:
,The solution, in terms of the uptake of sorbate by the solid, is given by Crank (1975) as:
(3)
Macropore Diffusion with Irreversible Equilibrium
In a macroporous pellet, even if most of the adsorption capacity is in the micropores, it is possible that the extracrystalline diffusion may be the dominant mass transfer resistance. For systems with irreversible equilibrium (rectangular isotherm), the "shrinking core" model may be applied. Under these conditions, adsorption occurs through an adsorption front, from the surface to the center of the particle, that is, a sorbate-free core in the adsorbent diminishes with time. In the region Rp > R > Rf the flux through the pores is constant. Ruthven (1984) presents the solution, for isothermal systems:
(4)
where
(5)
If the model applies, the experimental plot of t versus time shall yield a straight line through the origin. The slope will thus give an estimate of the macropore diffusion coefficient.
EXPERIMENTAL
Commercial zeolite Y, in the form of spherical pellets, was used. Some of its properties are shown in Table 1. All chemicals (o-xylene, p-xylene, toluene and p-diethylbenzene) had a purity of over 99%wt.
The experimental setup is shown in Figure 1. A Cahn model #2000 electrobalance was used. The zeolite sample was previously treated under a flow of nitrogen inside the system oven at 400oC for at least 12 hours, after having slowly raised the temperature during 5 hours. Concentration in the fluid phase was established by bubbling nitrogen (previously dried with molecular sieve 4A) through liquid sorbate under temperature-controlled conditions. The vapor pressure of the sorbate at the temperature of the thermostatized bath was used to calculate the fluid phase concentration. Uptake curves were measured at 150, 180 and 210 oC in the oven.
Figure 1: Gravimetric experimental setup.
RESULTS AND DISCUSSION
Figures 2 and 3 show all the experimental uptake curves. It may be observed that the process is very fast in the early stages of adsorption, specially for toluene, which reaches equilibrium after about 20 minutes (Figure 3a). After this fast initial step, adsorption continues slowly, taking between 60 and 100 minutes to reach complete equilibrium.
It is expected that the rate of adsorption increases as the temperature increases. In our experiments, only p-diethylbenzene showed that behavior. However, for o-xylene, p-xilene and toluene, the curves at 210° C showed anomalous behavior, as compared to those at 150° C and 180° C. This may indicate that an external factor is affecting the sorption rate, thereby inverting the usual trend of an increase in diffusion with temperature.
Kinetic Selectivity
It is well established that, under hindered conditions through the zeolitic channels, o-xylene normally diffuses slower than benzene, toluene and p-xylene (Shah and Liou, 1994). These molecules present smaller critical diameters (about 5.8 Å) than o-xylene (» 6.8 Å). At temperatures of 150oC and 180oC, the uptake curves of o-xylene and p-xylene confirm that behavior (Figure 4). Again, results at 210oC were anomalous.
Figure 2: Gravimetric uptake curves on commercial Y zeolite adsorbent (a) o-xylene; (b) p-xylene.
Figure 3: Gravimetric uptake curves on commercial Y zeolite adsorbent (a) toluene; (b) p-diethylbenzene.
Figure 4: Uptake curves for o-xylene and p-xylene on commercial Y zeolite, showing adsorbent p-selectivity. (a) 150oC; (b) 180oC.
Model for Micropore Diffusion Control
The experimental results, correlated using equation 3, yielded the estimates for the intracrystalline diffusion coefficients shown in Table 2. Diffusivity values increase from 150oC to 180oC, but, again, are completely anomalous at 210oC. Figure 5 shows a comparison of our results with those presented by Goddard and Ruthven (1986a) for crystals of natural faujasite (with a structure similar to that of zeolite Y). The large difference among the experimental results (about four orders of magnitude) highlights the inadequacy of the micropore diffusion model to describe the process. This is also shown by the large errors shown in Table 2. So, it is highly conceivable that the values obtained for these diffusion coefficient estimates are in gross error, due to the inadequacy of the model for this specific process.
Shrinking Core Model for Macropore Diffusion
Equations 4 and 5 were used to estimate macropore diffusion coefficients from our experimental data. Figures 6 and 7 show the plots obtained for o-xylene and p-xylene, respectively, at 150oC. It is quite evident that the model did not fully represent the process at all times. However, the diffusion coefficients (epDp), estimated from the experimental data (Table 3), are within the range of values expected for xylenes in the gaseous phase in other macroporous Y zeolite adsorbents (Morbidelli et al., 1985).
If we compare the results obtained for the two models that were tested, the macropore shrinking core model may be more coherent. This is not surprising, since in most commercial adsorbents the size of the crystals is very small ( 1mm radius). Therefore the resistance to mass transfer is normally negligible as compared to diffusion through the large macropores, which generally dominates the kinetic process.
*Crystals radius ~ 1m m.
Figure 5: Intracrystalline diffusivities of o-xylene. Comparison with data from Goddard and Ruthven (1986a).
Figure 6: Shrinking core model for o-xylene on commercial Y zeolite adsorbent at 150oC.
Figure 7: Shrinking core model for p-xylene on commercial Y zeolite adsorbent at 150oC.
CONCLUSIONS
A study of the sorption rates on a commercial Y zeolite adsorbent was performed. The rate of adsorption increased with increasing temperature in the range of 150 to 180oC, but behaved anomalously at 210oC. A possible explanation is the eventual intrusion of external heat effects on our experiments.
Under the same conditions, there is kinetic selectivity for p-xylene over o-xylene, which was specially observed at 150oC.
The two models that were applied did not fully represent the data, even though the macropore diffusion model seemed more reasonable regarding the calculated diffusivity values. Due to the multiple resistances to mass transfer normally involved in diffusion in pellets, this may suggest that a more complex diffusional model, with contributions from the various kinetic processes that may be identified in the sorption process, will have to be applied for these data.
ACKNOWLEDGEMENTS
The authors wish to acknowledge support from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), COPENE Petroquímica do Nordeste S/A and the Department of Organic and Inorganic Chemistry of Universidade Federal do Ceará.
NOMENCLATURE
c Fluid phase concentration
co Fluid phase concentration at time zero, at equilibrium
c¥ Fluid phase concentration at equilibrium
Dc Intracrystalline diffusivity
Dp Diffusivity in macropores
mt Mass adsorbed at time t
m¥ Mass adsorbed at time t Æ¥
P Sorbate partial pressure
Ps Saturation pressure at a given temperature
q Adsorbed phase concentration
qo Adsorbed phase concentration at time zero
qs Saturation concentration in adsorbed phase
R Particle radial distance
Rf Sorbate-free core radius
Rp Radius of particle
r Crystal radial distance
rc Crystal radius
T Temperature
t Time
Greek Letters
eb Porosity of bed
ep Porosity of adsorbent particle
g Fractional approach to equilibrium (equation 3)
t Dimensionless time (equation 5)
REFERENCES
Broughton, R.W.; Neuzil, R.W.; Pharis, J.M. and Brearley, C.S., The Parex Process for Recovering Paraxylene. Chemical Engineering Progress, 66, 9, 70 (1970).
Cavalcante Jr., C.L. and Gubulin, J.C., Liquid Phase Selective Adsorption of Xylenes. I. Chem. E. Symp. Series, 118, 83 (1990).
Crank, J., The Mathematics of Diffusion. Oxford Press, London (1975).
Goddard, M. and Ruthven, D.M., Sorption and Diffusion of C8 Aromatic Hydrocarbons in Faujasite Type Zeolites. II Sorption Kinetics and Intracrystalline Diffusivities. Zeolites, 6, 283 (1986a).
Goddard, M. and Ruthven, D.M., Sorption and Diffusion of C8 Aromatic Hydrocarbons in Faujasite Type Zeolites. III Self-Diffusivities by Tracer Exchange. Zeolites, 6, 445 (1986b).
Kärger, J. and Ruthven, D.M., Diffusion in Zeolites and Other Microporous Solids. John Wiley and Sons, New York (1992).
Marra Jr., W.D., Adsorção de Xilenos Sobre Zeólitas em Leito Fixo: Interpretação das Curvas Breakthrough. Masters thesis, Universidade Federal de São Carlos (1991).
Morbidelli, M.; Santacesaria, E.; Storti, G. and Carrà, S., separation of xylenes on y zeolites in the vapor phase. 2. breakthrough and pulse Curves and Their Interpretation. Ind. Eng. Chem. Process Des. Dev., 24, 83 (1985).
Morbidelli, M.; Storti, G. and Carrà, S., Comparison of Adsorption Separation Processes in the Liquid and Vapor Phase. Application to the Xylene Isomer Mixture, Ind. Eng. Chem. Fundam., 25, 1, 89 (1986).
Neves, S.B., Modelagem de um Leito Fixo de Adsorção para Sistemas Líquido-Sólido. Masters thesis, Universidade Federal da Bahia (1995).
Ruthven, D.M., Principles of Adsorption and Adsorption Processes. John Wiley and Sons, New York (1984).
Ruthven, D.M. and Goddard, M., Sorption and Diffusion of C8 Aromatic Hydrocarbons in Faujasite Type Zeolites. I. Equilibrium Isotherms and Separation Factors, Zeolites, 6, 275 (1986).
Santacesaria, E.; Morbidelli, M.; Servida, A.; Storti, G. and Carrà, S.; Separation of Xylenes on Y Zeolites. 2. Breakthrough Curves and their Interpretation. Ind. Eng. Chem. Process Des. Dev., 21, 3 (1982).
Shah, D.B. and Liou, H.Y., diffusion of aromatics through a silicalite membrane. IN: Zeolites and Related Microporous Materials: State of the Art 1994, J. J. Weitkamp, H.G. Karge, H. Pfeifer and W. Hölderich (eds.), Studies in Surface Science and Catalysis, 84, 1347 (1986).
Storti; G.; Santacesaria, E.; Morbidelli, M. and Carrà, S., Separation of Xylenes on Zeolites in the Vapor Phase. 3. Choice of the Suitable Desorbent. Ind. Eng. Chem. Process Des. Dev., 24, 89 (1985).
Yan, T.Y., Effects of Moisture in Separation of C8 Aromatics Using Medium-Pore Zeolites, Ind. Eng. Chem. Res., 27, 1665 (1988).
Publication Dates
-
Publication in this collection
09 Oct 1998 -
Date of issue
Sept 1997
History
-
Received
05 Mar 1997 -
Accepted
05 Aug 1997
