International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

Diffusion of Toluene and Ethylene in ZSM-12 zeolite:

Experimental and numerical study
Insaf Daldoul1, Hicham Chaouki2, Serge Kaliaguine3
1,3

Department of Chemical Engineering, Universit Laval, Qubec, Canada, G1V 0A6

2
Department of Civil Engineering, Universit Laval, Qubec, Canada, G1V 0A6

Abstract The Na+ and Ag+ forms of ZSM-12 zeolite microporous were used as effective adsorbents for hydrocarbons trap.
Diffusion of a single adsorbed component including ethylene as a light molecule or toluene a heavier one inside a narrow
one-dimensional channel was studied. Experimental desorption profiles for these two gases desorbing from both Ag-ZSM-12
and Na-ZSM-12 were obtained at very high heating rates comparable with the one observed during the cold start of a
combustion engine. Numerical simulation based on Ficks law and a transient diffusion boundary value yielded a good
fitting of these experimental results. For both ZSM-12 forms considered in this study, diffusion activation energies E and E
for toluene are much higher than those for ethylene. The values of these parameters are higher for the Ag + exchanged zeolite
than for Na-ZSM-12 and related to the presence of strong Lewis acid sites in the former material. In addition each activation
energy for Na-ZSM-12 and Ag-ZSM-12 decreases at increasing heating rate. This was explained by thermal vibrations of the
zeolite lattice which allow higher diffusivity at high temperature.
Keywords single component, single file diffusion, temperature-programmed desorption, ZSM-12

I.

INTRODUCTION

Faced with increasingly strict emission standards, automobile and catalyst manufacturers have investigated a variety of
approaches to reduce cold-start emissions. Attention has been focused on the start-up phase, when 70-80% of the
hydrocarbons generated (HCs) by the combustion engine are emitted [1]. Cold-start is referring to the short time period of 12 min, after engine ignition and before the three-way catalyst reaches its light-off temperature which is around 200C.
Zeolites have been found to be the preferred molecular sieves and the best candidates for this application due to their stability
under a variety of conditions and their affinity to adsorb HCs. In literature, different types of zeolites have been proposed and
tested as adsorbents for HC emissions [2][11]. It has been reported that ZSM-12, which possesses one dimensional channels
with 12-member ring pores, performs better in the envisaged application of hydrocarbon trap compared to more typical threedimensional zeolites [9][14]. With the growing use of one dimensional zeolites in separation and reaction processes, the
need to better understand the diffusion and transport properties of molecules within the channels of these materials is
important not only from a fundamental, but also from a practical point of view. The mechanism used to describe the
simultaneous passage of ethylene (light molecules) and toluene (heavy molecules) inside narrow one dimensional channels
was designated as the single-file diffusion[15]. This phenomenon has become a topic of current research interest. Heavy
aromatic molecules are strongly adsorbed in the zeolite, therefore they need a higher temperature in order to desorb from
channels. On the other hand, light hydrocarbon molecules are more weakly adsorbed and easily desorb from channels at
lower temperature. Thus in simultaneous diffusion, the desorption of lighter molecules such as ethylene from one
dimensional narrow channels may be retarded by co-adsorbed heavier ones such as toluene.
Despite the large number of theoretical studies published on the single-file diffusion over the years, experimental and
numerical single component loading (small molecules or aromatic molecules) are still rare. It is important to understand the
transport mechanism of these molecules within the pores of zeolites in order to design more efficient applications for these
materials. The aim of this paper is to provide a comparison between experimental and simulation results for each single kind
of gas after developing a model for diffusion in zeolite pores.

II.

EXPERIMENTAL DETAILS

The detailed synthesis procedure and physicochemical characterization results for Na-ZSM-12 and Ag-ZSM-12 samples with
molar ratio Si/Al = 52 used in this research were described in our previous study[15]. Fig. 1 shows the scanning electron
micrographs of Na-ZSM-12 and Ag-ZSM-12 zeolites as synthetized samples. They have uniform shape and particle size;
crystals have spherical form with average crystal diameter of 1m related to the use of methyltriethylammonium chloride as
organic template. This is agreement with the X-ray diffraction patterns of these hydrocarbon adsorbents that were recorded
using a Siemens D5000 diffractometer using the CuK ( = 1.54184) as shown in Fig. 2.
Page | 50

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

FIG.1. SCANNING ELECTRON MICROGRAPH OF SYNTHESIZED SAMPLE (A) Na-ZSM-12 and (B) Ag-ZSM-12 (52).

FIG.2. X-RAY POWDER DIFFRACTION PATTERNS OF AS-SYNTHESIZED Ag-ZSM-12 AND Na-ZSM-12 SAMPLES.
The high intensity of the XRD lines and the low background intensity point out that the samples are highly crystalline. The
BET surface area and micropore volume were determined using an Omnisorp-100 automatic analyzer after degassing the
sample at 150 C for at least 4h under vacuum (10 -5 to 10-4 Torr). Using Fourier transform infrared spectroscopy of
chemisorbed pyridine, Na-ZSM-12 has a higher measured CBrnsted/CLewis ratio compared to Ag-ZSM-12 indicating larger
content of Lewis acid sites. Brnsted and Lewis acid centers are detected respectively by peaks around 1550 cm-1 and
1448cm-1 (Fig. 3, Table 1). The acid site density of zeolites was quantified by ammonia thermodesorption. All experimental
results are reproduced here from our previous work (Table 2)[15].
Page | 51

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

TABLE 1
FTIR RESULTS OF CHEMISORBED PYRIDINE
Peak area (a. u.) 1448 cm-1

Peak area (a. u.) 1550 cm-1

Sample

Si/Al

CBrnsted / CLewis

Na-ZSM-12

52

0.012

0.322

35.5

Ag-ZSM-12

52

0.482

0.152

0.42

CBrnsted/CLewis = 1.88*A1550 / 1.42*A1448 23

TABLE 2
NH3 TPD RESULTS.
Sample

Tpeak1 (C)

NH3 (mol/g cat

5

Tpeak2

*10 )

(C)

NH3 (mol/g cat

Tpeak3 (C)

NH3 (mol/g cat

*105)

*10 )

Na-ZSM-12 (52)

92

32

150

13.4

345

204.6

Ag-ZSM-12 (52)

145

38.5

213

131.4

360

64.7

FIG.3 FTIR SPECTRA OF CHEMISORBED PYRIDINE (DESORPTION TEMPERATURE 150C) OF SAMPLES WITH
SI/AL=52; (A) NA-ZSM-12; (B) AG-ZSM-12.
Here we report the application of laser diode thermal desorption (S-960 LDTD) sample introduction source, coupled with an
atmospheric pressure chemical ionization (APCI) source mounted on a quantum ultra AM triple quadripole mass
spectrometer for analyte detection. This device allows reaching ramping rates similar to those reached in a muffler which is
around 3-5C/s. Samples preliminarily loaded with adsorbed single-component ethylene and toluene separately at room
temperature were deposited in the Lazewell plates of the LDTD source (Phytronix).

III.
3.1

SINGLE-COMPONENT GAS DESORPTION MODELING

Model description

The desorption process of a single-component gas can be simulated using a transient diffusion boundary value problem
coupled with Ficks law [16], [17]. In the present study, due to the structure of Na-ZSM-12and Ag-ZSM-12samples, a onedimensional diffusion problem is considered[9], [15], [18]. For this purpose, let L and

tT f denote the zeolite sample length

Page | 52

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

and the total time of the desorption process, respectively. At a given time t 0, tT
at the position x

0, L .

[Vol-2, Issue-4 April- 2016]

, let t , x denotes the gas density

The transient diffusion problem is given by the following equations [17], [19]:

t , x
t

div J t , x 0 , t , x 0, tT f 0, L

J t , x t, x t , x

(1)
(2)

with the Neumann and Dirichlet boundary conditions:

J x 0, t 0 , t

x L, t 0, t

(3)

and the initial condition:

x, t 0 0 x , x 0, L
In expressions (1)-(2),

(4)

div . and are divergence and gradient operators, respectively. J t , x and t , x represent

the gas flux and the diffusion coefficient. Substituting the flux expression (2) into the formula (1) leads to the following
partial differential equation:

t , x
t

div t, x t , x , t , x 0, tT f 0, L

(5)

In the sequel, the material is assumed to be homogeneous. Therefore, the diffusion coefficient is only time-dependent and it
is defined according to the Arrhenius law [16], [17], [20]:

kT

t C exp

where E kJ .mol
temperature.

, k J . mol

. K 1

and

(6)

T K are the activation energy, the Boltzmann constant and the

and C are two constants. is defined as follows [17], [18]:

E
exp
kT f

(7)

Moreover, during the experiments the temperature is increased according to a constant heating rate :

T t T0 t

(8)

where T0 is the initial temperature.

In some research works [16], [17], [19], [21], it has been shown, through numerical simulations, that the above mentioned
diffusion model fails to predict the general shape of the gas flux, especially at the beginning of the desorption process. In
order to overcome such a drawback, it has been suggested in the work of Chatterjee el al. [17][19]to use two diffusion
coefficients: i) the first one defined inside the physical domain and ii) an additional diffusion coefficient defined on the
domain boundary with a higher activation energy. Thus, two diffusion coefficients will be considered:

Page | 53

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

E
t C exp kT , x 0, L

E
t C exp
, if x L
kT

exp

kT f

exp E

kT
f

E E

(9)

For more details about this extra-diffusion barrier the reader is referred to the following references [16], [17], [20].
3.2

Time and Space Discretization

The boundary value problem (1)-(5) is solved using the finite difference method [22]. To this end, the spatial domain
is discretized into a grid of nodes
a set of time points

t
j

xi i 1
NS

such that x1 0 and xNS L . The time domain

0, L

0, tT is also discretized into

f

NT
j 0

. Without loss of generality, nodes and time points are supposed equidistant with fixed space step

x and fixed time step t . Therefore, one can show that:

t j j t , j 0,.., NT

xi x1 (i 1) x, i 1,.., NS

Let denotes by

ij

(10)

the approximation of the solution t j , xi at given time t j and point xi . Using a central difference

scheme, the right hand term of equation (5) is approximated by the following scheme [22]:

div t j t j , xi t j

i j1 2 i j i j1

(11)

For the time discretization of equation (5) an explicit scheme is used [22] Thus, the time partial derivative is approximated as
follows:

t j , xi
t

i j 1 i j

(12)

Combining expressions (11)-(12) leads to the following finite difference approximation scheme:

i j 1 i j t j

j
i 1

2 i j i j1

(13)

Therefore, knowing the solution at time step t j , the solution is updated in a straightforward way for the time step t j 1 . In
expression (13) the time step t is considered small enough compared to the square of the space step
the stability condition for the explicit scheme [22].

in order to fulfill

Page | 54

International Journal of Engineering Research & Science (IJOER)

IV.
4.1

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

EXPERIMENTAL AND NUMERICAL RESULTS

Experimental results

In our previous study using Fourier transform infrared spectroscopy of chemisorbed pyridine characterization, it has been
shown that Ag-ZSM-12 has low measured CBrnsted/CLewis indicating largest content of Lewis acid sites. On the other hand,
Na-ZSM-12 has a lower 1448 cm-1 peak area, a characteristic band of Lewis acid centers, than Ag-ZSM-12. Consequently
the ration CBrnsted/CLewis for Na-ZSM-12 increased. The Ag exchanged zeolite has the highest content in Lewis acid sites and
the lowest content in Brnsted acid sites (lowest 1550 cm-1 peak area)[15].
Three high heating rates have been used for the desorption process: 3C/s, 5C/s and 9C/s. Fig.4 reports TPD profiles of the
single-component, either ethylene and toluene, from Na-ZSM-12 and Ag-ZSM-12. All desorption profiles of both ethylene
and toluene display a single peak, except for the desorption curves of ethylene from Ag-ZSM-12 (Fig.5), which show two
peaks. In the latter case the low temperature peak is assigned to the ethylene desorption from Brnsted acid sites whereas the
high temperature desorption peak is from Lewis acid centers.

FIG.4. SINGLE COMPONENT (ETHYLENE OR TOLUENE) TPD PROFILES OF (A) Na-ZSM-12 AND (B) Ag-ZSM12 SAMPLES AT DIFFERENT HEATING RATES.

Page | 55

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

FIG.5. COMPARISON OF EXPERIMENTAL AND NUMERICAL DATA OF SINGLE-COMPONENT TPD PROFILES IN

AG-ZSM-12: (A) ETHYLENE AND (B) TOLUENE AND NA-ZSM-12: (A) ETHYLENE AND (B) TOLUENE WITH
3C/S AS HEATING RATE. VALUES OF

Ew

AND

REPORTED IN TABLE 4 WERE USED WITH

.
According to Table 3 the amounts of both ethylene and toluene detected are essentially not affected by the heating rate which
is an indication of results reproducibility. However a clear dependence of desorption temperature on heating rate for toluene
and ethylene was observed for the two zeolites. At increasing ramp rate from 3 to 9C, the maximum desorption peak
temperature is raised by 10-25C. Regarding the desorption temperature of both ethylene and toluene molecules, it seems that
the desorption temperature of Na-ZSM-12 is always lower than the desorption temperature of Ag-ZSM-12. This result
emphasizes the importance of the strong Lewis acid sites in Ag-ZSM-12 and confirms that the zeolite exchanged with Ag+ is
a better candidate as hydrocarbon adsorbent as indicated by its higher desorption temperatures [9], [15].

TABLE 3
SINGLE COMPONENT TPD RESULTS
Gas

Ethylene

Toluene

Desorbed amounts (mmol/g)

Rate (C/s)

Desorption temperature (C)

Na-ZSM-12

Ag-ZSM-12

Na-ZSM-12

Ag-ZSM-12

0.021

0.042

63

(70*;105**)

0.023

0.044

73

(81*;114**)

0.020

0.042

80

(89*;126**)

1.73

1.94

112

126

1.71

1.92

126

137

1.72

1.95

133

144

9
*

Desorption temperature obtained from the low T peak

Desorption temperature obtained from the high T peak

**

Page | 56

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

Independently of the zeolite and heating rate, the amount of toluene adsorbed and the desorption temperature are much
higher than those of ethylene.This indicates that part of ethylene is either not adsorbed at all or so weakly adsorbed that it is
desorbed at room temperature. It is found that toluene is more strongly adsorbed than ethylene on both Brnsted and Lewis
acid sites[15]. As shown in Table3, the amount of ethylene adsorbed on Ag-ZSM-12 is higher than the amount adsorbed on
Na-ZSM-12 which also indicates that the silver form is a better adsorbent for the smaller gas molecules.
4.2

Numerical results

In order to show the ability of the transient diffusion boundary value problem (1)-(2) to predict the obtained experimental
results, numerical simulations were carried out using the finite difference method (13). As mentioned earlier, two diffusion
coefficients will be considered for each gas (9). The ZSM-12 zeolite has a typical channel length nearly equal to 5 m and a

pore size of 5.7 6.1A . Molecular diameters of ethylene and toluene are

1.4 A and 5.7 A , respectively.

In Fig. 5 (A and B), numerical simulation results for ethylene and toluene desorption process using Na-ZSM-12 zeolite are
compared to experimental trends. It is shown that the diffusion model predicts desorption curves in a good agreement with
experimental results. Values of the activation energy used for these simulations are reported in Table 4. For the three heating
rates, activation energies associated to the ethylene are smaller than those of toluene. This trend could be explained through
the dimension of ethylene molecule which is smaller compared to the dimension of zeolite channels, in contrast to the
toluene molecule. As a consequence, ethylene can desorb with a lower energy of activation. For both gases and for all
heating rates the activation energy inside the domain
boundary

is lower than the activation energy at the lattice

E . Such a result corroborates the hypothesis made in Chatterjee et al.[17] Suggesting the introduction of an

additional activation energy at the boundary. Furthermore, we notice that increasing the heating rate leads to decreasing
activation energy (see Table 4). Indeed, by increasing temperature, the cages of zeolite channels undergo higher vibration and
thermal agitation which allows the adsorbed hydrocarbon molecules to be diffuse more easily.

TABLE 4
ACTIVATION ENERGY OBTAINED FROM NUMERICAL PROGRAM (kJ.mol-1)
Sample

Na-ZSM-12

Toluene

Ew

Tf (K)

Ew

Tf (K)

77

129

369

84

184

416

70

120

378

80

180

431

61
*

110
**

(50 ; 105 )

5
Ag-ZSM-12

Ethylene

Rate (C/s)

**

**

(43 ; 73 )

(34 ; 55 )

385

75

175

438

**

391

88

187

431

**

410

85

185

443

**

413

75

180

449

(114 ; 410 )
(111 ; 388 )
(107 ; 350 )

Activation energy value obtained from the low T peak

Activation energy value obtained from the high T peak

**

For the Ag-ZSM-12 zeolite, numerical and experimental results for the desorption process are compared in Fig. 6. Since
ethylene desorption curves (Fig. 6 (A)) exhibit two peaks corresponding, respectively, to Brnsted acid sites (low
temperature peak) and Lewis acid centers (high temperature peak), a deconvolution operation was carried out using FitPeak
Matlab code in order to uncouple peaks (Fig. 6). Fig. 5 (A and A) show a benchmark between numerical and experimental
results for the ethylene at 3C/s as an example of a heating rate while Fig. 5 (B and B) illustrate results for toluene. It is
shown that numerical results are in a good agreement with experimental trends. The activation energy values used in
Equations (9) and (11) for the Ag-ZSM-12 zeolite are reported in Table 4. The aforementioned remarks about the Na-ZSM12 zeolite, namely: i) ethylene having activation energy lower than toluene, ii) activation energy inside the zeolite lower than
the activation energy at the lattice boundary and iii) decreasing activation energy at increasing heating rate, still hold for the
Ag-ZSM-12 zeolite.
Page | 57

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

Finally, it may be concluded that, for both ethylene and toluene, activation energies for Na-ZSM-12 are lower than those of
Ag-ZSM-12 (see Table 4). This can be explained by the presence of strong Lewis acid sites in Ag + exchanged zeolite while
the Na-ZSM-12 zeolite is characterized by the presence of mainly Brnsted acid sites.

FIG.6. ETHYLENE THERMODESORPTION PROFILE OF AG-ZSM-12 WITH (A) 3C/s, (B) 5C/s AND (C) 9/s
HEATING RATE

V.

CONCLUSION

Diffusion of a single adsorbed component ethylene or toluene, inside the microporous ZSM-12 zeolite was studied
experimentally and numerically. The Na + and Ag+ forms of zeolite ZSM-12 were synthetized and characterized. The
thermodesorption process was conducted at three different heating rates (3, 5 and 9C/s) selected to be similar to the muffler
heating rate during the cold start process.
Experimentally, it was verified that the amount of both ethylene and toluene detected are essentially not affected by the
heating rate. Contrary to the ethylene adsorbed on Ag-ZSM-12, where this molecule desorbed as two peaks, only one peak
was observed for toluene. Moreover the amounts of ethylene adsorbed on ZSM-12 are low compared to those of toluene, due
to the small size of this molecule in comparison to the pore diameter of this zeolite adsorbent.
Numerically, using a transient diffusion boundary value problem coupled with Ficks law, it was possible to explain all
qualitative features of the temperature dependent desorption profiles observed experimentally. In this model, a coefficient C
was included in the diffusion coefficient

(t ) (6) that has units of m2/s and consequently the term t , x which describes

the probability that a molecule occupies a site at distance x in the zeolite channels at time t is dimensionless.
For the two solids considered in this study, activation energies Ew and E for toluene are much higher than those for
ethylene owing to its bigger size. We have also shown that the values of these parameters are higher for the Ag+ exchanged
zeolite than for Na-ZSM-12 because of the presence of strong Lewis acid sites in the former material. In addition each

Page | 58

International Journal of Engineering Research & Science (IJOER)

ISSN: [2395-6992]

[Vol-2, Issue-4 April- 2016]

activation energy for Na-ZSM-12 and Ag-ZSM-12 decreases at increasing heating rate. This was explained by thermal
vibrations of lattice which allow higher diffusivity at high temperature.

ACKNOWLEDGEMENTS
The authors gratefully acknowledge Natural Science and Engineering Research Council of Canada for financial support.

REFERENCES
[1]
[2]
[3]
[4]
[5]

[6]

[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]

F. F. Zhao, Technologies for Near-Zero-Emission Gasoline-Powered Vehicles. 2006.

S. P. Elangovan, M. Ogura, S. Ernst, M. Hartmann, S. Tontisirin, M. E. Davis, and T. Okubo, A comparative study of zeolites SSZ33 and MCM-68 for hydrocarbon trap applications, Microporous Mesoporous Mater., vol. 96, pp. 210215, 2006.
S. P. Elangovan, M. Ogura, M. E. Davis, and T. Okubo, SSZ-33: A promising material for use as a hydrocarbon trap, J. Phys.
Chem. B, vol. 108, no. 35, pp. 1305913061, 2004.
P. Brauer, J. Karger, and N. Neugebauer, Reaction and Particle Distribution in Networks of Single-File Systems, Europhys. Lett.,
vol. 53, no. 1, pp. 814, 2001.
J. M. Lpez, M. V. Navarro, T. Garca, R. Murillo, a. M. Mastral, F. J. Varela-Ganda, D. Lozano-Castell, a. Bueno-Lpez, and D.
Cazorla-Amors, Screening of different zeolites and silicoaluminophosphates for the retention of propene under cold start
conditions, Microporous Mesoporous Mater., vol. 130, no. 13, pp. 239247, 2010.
B. Purtolas, M. Navlani-Garca, J. M. Lpez, T. Garca, R. Murillo, a. M. Mastral, M. V. Navarro, D. Lozano-Castell, a. BuenoLpez, and D. Cazorla-Amors, Molecular simulation design of a multisite solid for the abatement of cold start emissions, Chem.
Commun., vol. 48, p. 6571, 2012.
K. F. Czaplewski, T. L. Reitz, Y. J. Kim, and R. Q. Snurr, One-dimensional zeolites as hydrocarbon traps, Microporous
Mesoporous Mater., vol. 56, no. 1, pp. 5564, 2002.
S. Gopal, K. Yoo, and P. G. Smirniotis, Synthesis of A1-rich ZSM-12 using TEAOH as template, Microporous Mesoporous
Mater., vol. 49, no. 13, pp. 149156, 2001.
Z. Sarshar, M. H. Zahedi-Niaki, Q. Huang, M. Ei, and S. Kaliaguine, MTW zeolites for reducing cold-start emissions of
automotive exhaust, Appl. Catal. B Environ., vol. 87, no. 12, pp. 3745, 2009.
A. Iliyas, Z. Sarshar, H. M. Zahedi-Niaki, M. Ei, and S. Kaliaguine, One-dimensional molecular sieves for hydrocarbon cold start
emission control, Top. Catal., vol. 52, pp. 18601867, 2009.
Z. Sarshar, M. H. Zahedi-Niaki, Q. Huang, and S. Kaliaguine, Synthesis, structural and acidity characterizations of the large-pore
zeolite SSZ-42 for controlling cold-start emissions, Microporous Mesoporous Mater., vol. 118, no. 13, pp. 373381, 2009.
A. Iliyas, M. Eic, M. H. Zahedi-Niaki, and S. Vasenkov, Toward observation of single-file diffusion using the tracer zero-length
column method, J Phys Chem B, vol. 112, no. 12, pp. 38213825, 2008.
A. Iliyas, M. H. Zahedi-Niaki, M. Ei, and S. Kaliaguine, Control of hydrocarbon cold-start emissions: A search for potential
adsorbents, Microporous Mesoporous Mater., vol. 102, no. 13, pp. 171177, 2007.
A. Iliyas, H. M. Zahedi-Niaki, and M. Ei, One-dimensional molecular sieves for hydrocarbon cold-start emission control: Influence
of water and CO2, Appl. Catal. A Gen., vol. 382, no. 2, pp. 213219, 2010.
I. Daldoul, S. Auger, P. Picard, B. Nohair, and S. Kaliaguine, Effect of temperature ramp on hydrocarbons desorption profiles from
zeolite ZSM-12., Can. J. Chem. Eng., vol. 94, article in press, 2015.
J. Krger, R. Valiullin, and S. Vasenkov, Molecular dynamics under confinement to one dimension: Options of measurement and
accessible information, New J. Phys., vol. 7, pp. 115, 2005.
S. Chatterjee and G. M. Schtz, Diffusion of a hydrocarbon mixture in a one-dimensional zeolite channel: An exclusion model
approach, Microporous Mesoporous Mater., vol. 125, no. 12, pp. 143148, 2009.
S. Chatterjee and G. M. Schtz, Diffusion of a hydrocarbon mixture in a one-dimensional zeolite channel: An exclusion model
approach, Microporous Mesoporous Mater., vol. 125, no. 12, pp. 143148, Oct. 2009.
S. Chatterjee and G. M. Schtz, Diffusion in a One-Dimensional Zeolite Channel: An Analytical and Numerical Study, DiffusionFundamentals.Org, vol. 11, no. 2009, pp. 113, 2009.
J. Karger, M. Petzold, H. Pfeifer, S. Ernst, and J. Weitkamp, Single-file diffusion and reaction in zeolites, J. Catal., vol. 136, no. 2,
pp. 283299, 1992.
P. Bruer, A. Brzank, and J. Krger, Adsorption and Reaction in Single-File Networks, J. Phys. Chem. B, vol. 107, no. 2, pp.
18211831, 2003.
N. Ozisik, Finite Difference Methods in Heat Transfer, CRC Press,. 1994.

Page | 59