Journal of Colloid and Interface Science 313 (2007) 12–17

www.elsevier.com/locate/jcis

Equilibrium and kinetic analysis of CO2–N2 adsorption separation

by concentration pulse chromatography
Peiyuan Li, F. Handan Tezel ∗
Department of Chemical Engineering, Faculty of Engineering, University of Ottawa, 161 Louis Pasteur, Ottawa, Ontario K1N 6N5, Canada
Received 28 November 2006; accepted 11 April 2007
Available online 13 April 2007

Abstract
CO2 and N2 adsorption kinetics and equilibrium behaviours have been studied with silicalite, NaY and 13X by using concentration pulse
chromatography for the separation of these gases in the present study. Adsorption Henry’s Law constants, the heat of adsorption values, micropore
diffusion coefficients and corresponding activation energies are determined experimentally and the three different mass transfer mechanisms are
discussed. From the equilibrium data, the corresponding separation factors are obtained for the adsorption separation processes. The heat of
adsorption values as well as the Henry’s Law adsorption equilibrium constants of CO2 are much higher than those of N2 for all the adsorbents
studied. 13X, NaY and silicalite all have good separation factors for CO2 /N2 system based on equilibrium processes. The order of the equilibrium
separation factors is 13X (Ceca) > 13X (Zeochem) > NaY (UOP)  silicalite (UOP). Equilibrium selectivity favours CO2 over N2 . Micropore
diffusion resistance is the definite dominant mass transfer mechanism for CO2 with silicalite and NaY.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Adsorption of carbon dioxide; Adsorption of nitrogen; Flue gas separation; Silicalite; NaY; 13X; Henry’s Law constant

1. Introduction the present study. Henry’s Law constants are calculated from
the corrected first moment of the response peaks. The heat of
The global environment is a major issue today, and global adsorption and the pre-exponential factors are obtained by us-
warming in particular is the focus of much attention. Accumu- ing the Van’t Hoff equation. From the equilibrium data, the
lation of greenhouse gases (GHG) in the atmosphere is respon- equilibrium selectivities are obtained for the adsorption sep-
sible for increased global warming of our planet. The increasing aration process for the four different adsorbents studied. The
concentration of carbon dioxide mainly from flue gas, auto- micropore diffusion (Dc /rc2 ) is determined from the corrected
mobile and landfill emissions in the atmosphere is the major first and second moments of the response peaks, and the three
contributor to this problem with more than 80% of total GHG different mass transfer mechanisms are discussed.
emissions [1]. For this reason, together with a tighter control
in emissions to meet Kyoto Protocol targets, separation of car- 2. Concentration pulse chromatographic technique
bon dioxide from nitrogen (representing the major component
The use of concentration pulse chromatography for adsor-
of air) has been the focus of the present study. A variety of tech-
bent screening is very attractive since it is relatively inexpensive
niques have been used for CO2 –N2 adsorption separation in the
to setup and is very efficient. It is a dynamic method of conduct-
literature [2–11].
ing adsorption measurements, which measures the response of a
Adsorption separation of carbon dioxide from nitrogen is in-
chromatographic column to a pulse in adsorbate concentration.
vestigated by using concentration pulse chromatographic tech-
The first and second moments of the response peak produced
nique, with silicalite, NaY and 13X zeolites as adsorbents, in
by a pulse in carrier gas concentration would yield the ini-
tial slope of the adsorption isotherm and the diffusivity. This
* Corresponding author. Fax: +1 613 562 5172. method, which indirectly measures the adsorption parameters,
E-mail address: tezel@eng.uottawa.ca (F.H. Tezel). has been shown to agree well with gravimetric and volumet-
0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2007.04.015
 P. Li, F.H. Tezel / Journal of Colloid and Interface Science 313 (2007) 12–17 13

ric data. Concentration pulse chromatographic technique was is the macropore diffusivity (based on free pore area in the pel-
found to be simpler and quicker for the determination of binary let), Dc is the diffusivity within the crystal within the pellet (or
adsorption isotherms, as well [4,12–16]. micropore diffusivity), and Dm is the molecular diffusivity. The
In this technique, a pulse of the sample gas is injected into terms on the right hand side of this equation represent different
the carrier stream and passes through the column that is packed mass transfer resistances that exist for adsorption to take place
with the adsorbent. The response of the column to this injection in a column, packed with adsorbent pellets that are composed of
is measured as concentration vs time at the exit of the column. zeolite crystals. The first term represents axial dispersion. The
This response peak gives the mean retention time of the sam- second term represents external film mass transfer. The third
ple, μ, defined as the first moment of the chromatogram [17]. and fourth terms represent macropore and micropore diffusion
Dimensionless Henry’s Law constants can be calculated from resistances, respectively. According to Eq. (5), a plot of the dis-
the corrected first moment (with respect to the dead time) of the persion (σ 2 L/2μ2 v, which is the left-hand side) vs v −2 should
response peaks by using following equation [14]: be approximately linear with slope DL and intercept given by
∞   the summation of the last three terms on the right-hand side of
c(t − μD ) dt L (1 − ε)K
μ= 0 ∞ = 1+ , (1) Eq. (5). By using these plots, the dominant mass transfer mech-
0 c dt
v ε anisms, as well as the dispersion and mass transfer parameters
where t is the time, c is the adsorbate concentration measured can be determined.
at the outlet of the column, L is the column length, ε is the bed The molecular diffusivity, Dm , in cm2 /s, can be estimated
porosity, v is the interstitial fluid velocity, K is the dimension- by [20–22]:
less Henry’s Law adsorption equilibrium constant, and μD is T 3/2 (1/MA + 1/MB )1/2
the dead time. Dm = 1.8583 × 10−3 × 2 Ω
, (6)
P σAB AB
The dimensionless Henry’s Law constants, K, can be con-
verted to a dimensional form [14,18] as follows: where T is the temperature in K, M is the molecular weight,
P is the total pressure in atm, σ is a constant in the Lennard-
K
Kp = , (2) Jones potential-energy function in Å, and Ω is the collision
RT ρP integral (dimensionless). Subscripts A and B indicate compo-
where T is temperature, ρP is the density of the pellets of the nents A and B.
adsorbent and Kp is the dimensional Henry’s Law adsorption The macropore diffusivity, Dp , is going to be given by a
equilibrium constant given in mmol/(g atm). combination of molecular diffusivity and Knudsen diffusivity,
The Van’t Hoff equation is given as [19]: and can be estimated by [13]:
Kp = K0 e−(H /RT ) , 1
(3) Dp = , (7)
1/Dm + 1/Dk
where T is the column temperature, R is the gas constant, H
is the heat of adsorption, and K0 is the pre-exponential factor. where Dk is the Knudsen diffusivity in cm2 /s, which is given
H and K0 may be obtained from the experimentally deter- by [13,23]:
 1/2
mined Kp values at several different temperatures. These values T
can then be used to calculate equilibrium separation factor, the Dk = 9700rp , (8)
M
ratio of the Henry’s Law constants, which is useful in predicting
where rp is the pore radius in cm, M is the molecular weight,
promising separation conditions.
and T is temperature in K.
The rate of diffusion in micropores can be determined by this
After axial dispersion is determined from the slopes of dis-
method from the second moment of the response peak defined
persion (σ 2 L/2μ2 v) vs v −2 plots, using the estimated values
as [17,19]:
∞ for molecular diffusivity (from Eq. (6)) and the macropore dif-
c(t − μ)2 dt fusivity (from Eq. (7)), the micropore resistance and the cor-
σ = 0 ∞
2
, (4) responding value of Dc /rc2 can be determined experimentally
0 c dt
from Eq. (5).
where μ is the corrected retention time, t is time, and c is the ad- If external film mass transfer and macropore diffusion re-
sorbate concentration at the outlet of the column. At low flow sistances are much smaller than micropore diffusion resistance,
rates, a useful general form for the second moment of the re- the micropore diffusion would be the dominant mass transfer
sponse peak is given as [19]: mechanism. These will be the ideal conditions, under which the
 2  latter can be determined. Therefore, it is necessary to check the
σ 2 L DL ε Rp Rp2 rc2
· = 2 + + + , (5) relative values of different mass transfer resistances under the
2μ2 v v (1 − ε) 3Dm 15εp Dp 15KDc
experimental conditions studied.
where μ is the corrected retention time (the first moment of
the response peak), σ 2 is the second moment of the response 3. Experimental
peak, L is the column length, ε is the bed porosity, v is the
interstitial fluid velocity, K is the dimensionless Henry’s Law The schematic diagram of the experimental apparatus
constant, DL is the axial dispersion coefficient, Rp is the pel- (Fig. S1), as well as the experimental details are given in sup-
let radius, rc is the crystal radius, εp is the pellet porosity, Dp porting information.
14 P. Li, F.H. Tezel / Journal of Colloid and Interface Science 313 (2007) 12–17

Table 1
Details of adsorbents packed in the columns
Type Commercial Si/Al Commercial Shape Size as supplied Supplier
name ratio number (diameter)
Silicalite MOLSIV adsorbents >1000 HISIV 3000 Cylinder 1/16 inch Universal Oil Products (UOP), Des Plaines, IL, USA
NaY MOLSIV adsorbents 1.8 NAY Cylinder 1/16 inch Universal Oil Products (UOP), Des Plaines, IL, USA
13X Molecular sieve adsorbents 1.0–1.5 Z10-01 Bead 8 × 12 mesh Zeochem, Louisville, KY, USA
13X G5 13X 1.25 GS 13X Bead 8 × 12 mesh Ceca, France

Fig. 1. Linear regression of the relation between Henry’s Law adsorption equilibrium constant (Kp ) and column temperature (T ): Kp vs T −1 for different adsorbents
and adsorbates.

The details about the adsorbents packed in the column are to the fact that, although the molecular weight of CO2 (44) is
listed in Table 1. larger than that of N2 (28), its kinetic diameter is smaller (3.3 Å
The sample and carrier gases used in the experiments are as opposed to 3.64 Å for N2 ). Therefore, it can easily get into
available (in Table S1, supporting information). The experimen- the pores of the adsorbents. Ceca 13X has significantly larger
tal and column specifications and conditions are available (in Henry’s Law constants for both of the adsorbates, compared to
Table S2, supporting information). the other adsorbents studied. N2 Henry’s Law constants were
almost identical for silicalite, NaY and Zeochem 13X. Table 2
4. Results and discussion lists the values of H and K0 , which give the temperature in
dependence of the Henry’s Law constants according to Eq. (3).
By using Eqs. (1) and (2) Henry’s Law constants are cal- From the values of H , it can be seen that CO2 heats of ad-
culated from the corrected first moment of the response peaks sorption values are also higher than those for N2 with all the
(Determination of System Dead Time, supporting information). adsorbents studied. This is due to the fact that CO2 has a much
Representative plots for CO2 and N2 are shown in Fig. 1 for all higher quadropole moment and a much higher polarizability,
the adsorbents studied. Henry’s Law constants are much higher compared to N2 , therefore its interaction with the adsorbents
for CO2 than those for N2 for all the adsorbents. This is due are going to be much stronger. The difference in heat of ad-
 P. Li, F.H. Tezel / Journal of Colloid and Interface Science 313 (2007) 12–17 15

Table 2
Heat of adsorption and pre-exponential factor values and their comparisons
with literature
Adsorb- Adsorb- H H (kJ/mol) K0
ents ates (kJ/mol) (literature) (mmol/(g atm))
Silicalite N2 −18.5 −15.1 [24] 1.14 × 10−4
−17.6 [25]
CO2 −27.4 −21.7 [26] 8.18 × 10−5
−24.1 [24]
−23.4 [27]
−27.2 [25]
NaY N2 −15.4 – 3.58 × 10−4
CO2 −33.6 −30 to 6.33 × 10−5
−33 [28]
13X N2 −15.4 (Zeochem) −25.0 [26] 3.36 × 10−4 (Zeochem)
−14.9 (Ceca) 9.01 × 10−4 (Ceca)
CO2 −34.9 (Zeochem) −36.0 [26] 6.43 × 10−5 (Zeochem)
−26.2 (Ceca) 4.58 × 10−3 (Ceca)

sorption values is more pronounced with more heterogeneous

Fig. 2. The relation between adsorption equilibrium separation factor
adsorbents (NaY and 13X), as opposed to silicalite, as expected. (αe,CO2 /N2 ) and column temperature (T ): αe,CO2 /N2 vs T −1 for different ad-
In Table 2, literature values of heat of adsorption were com- sorbents studied.
pared to the ones obtained in the present study. As can be seen
from this table, all values agree with the literature, except the where KpA and KpB are Henry’s Law adsorption equilibrium
values for 13X, for which, the heat of adsorption value seems constants for components A and B, respectively. When using
to be sensitive to the origin of the adsorbent. There was even concentration pulse chromatography, we can make sure that the
a discrepancy between the 2 different 13X samples we tried in system has the concentrations low enough to follow Henry’s
the present study. Law, by injecting a very small amount of the sample gas into
By using the parameters in Table 2, the Henry’s Law con- the carrier gas stream.
stants at other temperatures were estimated from the experi- The separation factors, determined from the experimental
mental data for CO2 and N2 and are presented in Table S3, data obtained, are shown as a function of inverse tempera-
in which the data from Harlick and Tezel [29] are also listed ture in Fig. 2. Henry’s Law adsorption equilibrium constants
for CO2 for comparison. They used nitrogen as the carrier gas and adsorption equilibrium separation factors are given in Ta-
while helium was used as the carrier gas in the present study. ble S3 in supporting information. The equilibrium separation
As can be seen from Table S3, Henry’s Law constants for CO2 factors decrease as temperature increases, depending on the rel-
are very similar for NaY zeolite. For 13X zeolite, on the other ative magnitudes of the heats of adsorption values. Following
hand, the samples obtained from different manufacturers be- Eqs. (3) and (10), the slope of separation factor vs inverse tem-
haved quite differently for CO2 . The Henry’s Law constants perature plot is given by −(HCO2 − HN2 )/R. In this case,
were the highest for Ceca 13X for all the temperatures studied,
since the heat of adsorption of CO2 is more negative than that
followed by UOP 13X, and followed by the Zeochem sample.
of N2 , this plot will have a positive slope, which will cause
The primary requirement for an economic separation process
the separation factor to decrease when T increases. Accord-
is an adsorbent with sufficiently high selectivity, capacity, and
ing to the data obtained, all the adsorbents have good separa-
life. The selectivity would depend on difference in either ad-
tion factors for CO2 /N2 system for equilibrium processes. The
sorption kinetics or adsorption equilibrium. Most of the adsorp-
order of the equilibrium separation factors follows the Si/Al ra-
tion processes in current use depend on equilibrium selectivity.
tio of the adsorbents: 13X (Ceca) > 13X (Zeochem) > NaY
In considering such processes it is convenient to define an ad-
(UOP)  silicalite (UOP). The reason for that is the homoge-
sorption equilibrium separation factor as follows:
neous surface characteristic of the silicalite adsorbent which
xA /xB has a Si/Al ratio in the thousands. The more Al a zeolite will
αe,A/B = , (9)
yA /yB contain, the more cations will be needed to balance the charge
where xA , xB , yA , and yB are the mole fractions of components difference between Si4+ and Al3+ , the more heterogeneous its
A and B in adsorbed and gas phases at equilibrium, respectively. surface will be. Since zeolites X and Y have a much lower
For Langmuir type favourable adsorption isotherms, the equi- Si/Al ratio, compared to silicalite, CO2 will be more attracted
librium separation factor is constant, and is simply given by the to these adsorbents, because of its higher quadrupole moment,
ratio of the Henry’s Law constants: compared to silicalite, resulting in higher heats of adsorption,
compared to N2 .
KpA
αe,A/B = , (10) It needs to be emphasized that the separation factors consid-
KpB ered here are based on pure component data in Henry’s Law re-
16 P. Li, F.H. Tezel / Journal of Colloid and Interface Science 313 (2007) 12–17

Table 3
Mass transfer resistances for CO2
Adsorbent T Intercept×(1−ε)/ε Rp2 /(3Dm ) Rp2 /(15εp Dp ) rc2 /(15KDc )
(◦ C) second Second (%) Second (%) Second (%)
Silicalite (UOP) 40 0.06587 0.00038 0.58 0.00098 1.48 0.06451 97.94
60 0.06712 0.00034 0.50 0.00093 1.39 0.06586 98.11
80 0.06800 0.00031 0.45 0.00089 1.31 0.06680 98.23
100 0.07700 0.00028 0.37 0.00086 1.12 0.07586 98.51
NaY (UOP) 40 0.15563 0.00038 0.24 0.00080 0.51 0.15445 99.24
60 0.12501 0.00034 0.27 0.00077 0.61 0.12391 99.12
80 0.09752 0.00031 0.32 0.00074 0.76 0.09648 98.93
100 0.09865 0.00028 0.29 0.00071 0.72 0.09765 98.99

Fig. 3. Linear relation of σ 2 L/2μ2 v vs v −2 in the lower flow rate region for Fig. 4. Linear regression of the relation between micropore diffusivity and tem-
CO2 with 13X (Zeochem) at 100 ◦ C. perature for CO2 .

gion adsorption equilibrium properties. Although, realistically from the difference between column 3 in Table 3 (intercept ×
speaking, the separation factor should be based on actual ca- (1 − ε)/ε) and the addition of columns 4 and 6 according to
pacities under mixture conditions; the one calculated from the Eq. (5). From the table, it can be seen that the contributions
pure component data would be an initial indicator of how easy from external film mass transfer and macropore diffusion are
the separation would be. negligible with micropore diffusion (with more than 98% of to-
Equation (5) suggests that a plot of the dispersion (σ 2 L/ tal resistance) being the definite dominant mass transfer mech-
2μ v) vs 1/v 2 should be approximately linear with slope DL
2
anism for CO2 with silicalite and NaY the systems studied.
and intercept given by the last term in Eq. (5). Such plots pro- Identical results were reported by Triebe and Tezel [30] for dif-
vide a convenient means of extracting the mass transfer pa- ferent systems. In their work, maximum macropore diffusion
rameters. From a comparison of the mass transfer resistances resistance consisted of 2–4%. The contribution from external
measured under different experimental conditions, it is possible mass transfer was less than 1%, and micropore diffusion resis-
to establish which mass transfer resistance is dominant. A typ- tance was dominant for CO and N2 in clinoptilolite. For other
ical plot of dispersion (σ 2 L/2μ2 v) vs 1/v 2 is given in Fig. 3 systems looked at in this study, the intercepts are so small that
for CO2 with Zeochem 13X adsorbent. These plots are qualita- it is difficult to obtain accurate values.
tively similar for all the systems studied. Knowing the values of K calculated from the first moments,
For CO2 with silicalite and NaY as the adsorbents, the con- the values of the micropore diffusion (Dc /rc2 ) for CO2 are de-
tributions of different mass transfer resistances to total disper- termined from the micropore diffusion resistances calculated at
sion are given in Table 3, in which, external film mass transfer different temperatures and the results are plotted in Fig. 4. The
(Rp2 /3Dm in column 4), and macropore diffusion resistances corresponding parameters of the Arrhenius type equation
(Rp2 /15εp Dp in column 6) are estimated and the micropore  
Ea
diffusion resistances (rc2 /15KDc in column 8) are calculated Dc = D0 exp − (11)
RT
 P. Li, F.H. Tezel / Journal of Colloid and Interface Science 313 (2007) 12–17 17

Table 4 uate Scholarship, the Ontario Graduate Scholarship (OGS),

Parameters D0 /rc2 and Ea of CO2 according to Arrhenius equation (12) and the Canadian Society for Chemical Engineering (CSChE)
Adsorbents D0 /rc2 (s−1 ) Ea (kJ/mol) Scholarship are gratefully acknowledged.
Silicalite 64.7 22.4
NaY 6606 41.9 Supporting information

or The online version of this article contains additional support-

  ing information: Fig. S1, Tables S1–S4, Experimental Details,
Dc D 0 Ea
= exp − (12) Determination of System Dead Time, Nomenclature.
rc2 rc2 RT
Please visit DOI: 10.1016/j.jcis.2007.04.015.
are tabulated in Table 4. The micropore diffusion values
(Dc /rc2 ) are given in Table S4 in supporting information for References
different temperatures studied for CO2 . In Fig. 4 and Table S4,
the micropore diffusion (Dc /rc2 ) of CO2 with silicalite is higher [1] J. Hansen, I. Fung, A. Lacis, D. Riud, J.S. Levedeff, R. Ruedy, G. Russell,
than that with NaY. The reason for that is the homogeneous sur- J. Geophys. Res. 93 (1988) 9341–9364.
face characteristic of the silicalite adsorbent which has a Si/Al [2] F. Dong, H. Lou, A. Kodama, M. Goto, T. Hirose, Sep. Purif. Technol. 16
ratio in the thousands, so the force between silicalite and CO2 (1999) 159–166.
[3] F. Dreisbach, R. Staudt, J.U. Keller, Adsorption 5 (1999) 215–227.
is smaller than that between NaY and CO2 . Therefore, CO2 is
[4] P.J.E. Harlick, F.H. Tezel, Can. J. Chem. Eng. 79 (2001) 236–245.
easier to move around silicalite than NaY so that the micropore [5] P.J.E. Harlick, F.H. Tezel, Sep. Sci. Technol. 37 (2002) 33–60.
diffusion (Dc /rc2 ) of CO2 with silicalite is higher than that with [6] P.J.E. Harlick, F.H. Tezel, Sep. Purif. Technol. 33 (2003) 199–210.
NaY. [7] R. Hernandez-Huesca, L. Diaz, G. Aguilar-Armenta, Sep. Purif. Technol.
15 (1999) 163–173.
5. Conclusions [8] E.S. Kikkinides, R.T. Yang, G. Aguilar-Armenta, Ind. Eng. Chem. Res. 32
(1993) 2714–2720.
Following conclusions can be drawn from the present study: [9] N. Nishikawa, J. Jpn. Inst. Energy 71 (1992) 1090–1098.
[10] S. Pakseresht, M. Kazemeini, M.M. Akbarnejad, Sep. Purif. Technol. 28
(2002) 53–60.
1. CO2 heats of adsorption values are higher than those for
[11] Y. Takamura, S. Narita, J. Aoki, S. Hironaka, S. Uchida, Sep. Purif. Tech-
N2 with different adsorbents studied, which agrees with the nol. 24 (2001) 519–528.
results from the literature. [12] H.W. Haynes, P.N. Sarma, S. Prasertmanukitch, AIChE J. 19 (1973) 1043–
2. The Henry’s Law adsorption equilibrium constants of CO2 1046.
are much higher than those of N2 for the different ad- [13] R.T. Yang, Gas Separation by Adsorption Process, Butterworths, Stone-
sorbents studied and equilibrium selectivity favors CO2 ham, MA, 1987, pp. 9–338.
[14] P.J.E. Harlick, F.H. Tezel, Adsorption 6 (2000) 293–309.
over N2 .
[15] P.J.E. Harlick, F.H. Tezel, Adsorption 9 (2003) 275–286.
3. 13X, NaY and silicalite all have good separation factors for [16] J. Nokerman, X. Canet, P. Mougin, S. Limborg-Noetinger, M. Frere, Meas.
CO2 /N2 system for equilibrium processes. The order of the Sci. Technol. 16 (2005) 1802–1812.
equilibrium separation factors is 13X Zeolite (Ceca) > 13X [17] D.B. Shah, D.M. Ruthven, AIChE J. 23 (1977) 804–810.
Zeolite (Zeochem) > NaY (UOP)  Silicalite (UOP) for [18] P. Li, F.H. Tezel, Microporous Mesoporous Mater. 98 (2007) 94–101.
the temperature range studied. [19] D.M. Ruthven, Principles of Adsorption and Adsorption Processes, Wiley,
4. The equilibrium separation factors decrease as column tem- New York, 1984, pp. 1–433.
[20] S. Chapman, T.G. Gowling, Mathematical Theory of Non-Uniform Gases,
peratures increase for CO2 /N2 separation for all the adsor- second ed., Cambridge Univ. Press, Cambridge, 1951, chaps. 10, 14.
bents studied. [21] J.O. Hirschfelder, C.F. Curtiss, R.B. Bird, Molecular Theory of Gases and
5. Micropore diffusion resistance is the definite dominant Liquids, Wiley, New York, 1954, p. 539.
mass transfer mechanism for CO2 with silicalite and NaY. [22] R.B. Bird, W.E. Stewart, E.N. Lightfoot, Diffusivity and the Mechanisms
The order of the contributions of three mass transfer re- of Mass Transport in Transport Phenomena, Wiley, New York, 1960,
pp. 495–518.
sistances is micropore diffusion  macropore diffusion >
[23] W. Kauzmann, Kinetic Theory of Gases, Benjamin, New York, 1966,
external film mass transfer, which is identical with the re- pp. 1–248.
sults for other similar systems from the literature. [24] T.C. Golden, S. Sircar, J. Colloid Interface Sci. 162 (1994) 182–188.
6. The micropore diffusion (Dc /rc2 ) of CO2 with silicalite is [25] J.A. Dunne, R. Mariwala, M. Rao, S. Sircar, R.J. Gorte, A.L. Myers, Lang-
higher than that with NaY. muir 12 (1996) 5888–5895.
7. The adsorbents studied have potential applications for en- [26] V.R. Choudhary, S. Mayadevi, Sep. Sci. Technol. 28 (1993) 2197–2209.
[27] T. Yamazaki, M. Katoh, S. Ozawa, Y. Ogino, Mol. Phys. 80 (1993) 313–
vironmental industry, specifically for flue gas separation.
324.
[28] J. Pires, M. Brotas de Carvalho, F. Ramoa Ribeiro, E.G. Derouane, J. Mol.
Acknowledgments Catal. 85 (1993) 295–303.
[29] P.J.E. Harlick, F.H. Tezel, Microporous Mesoporous Mater. 76 (2004) 71–
Financial supports received from the Natural Sciences and 79.
Engineering Research Council of Canada (NSERC) Postgrad- [30] R.W. Triebe, F.H. Tezel, Gas Sep. Purif. 9 (1995) 223–230.