Kinetics of Aqueous-Phase Catalytic Dehydration of 2-Propanol

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Kinetics of Aqueous-Phase Catalytic

Dehydration of 2-Propanol
Jean-Marc Rivard, Qinglin Zhang and Karl T. Chuang*

Department of Chemical and Materials Engineering, University of Alberta, Edmonton, AB TGG 2GG, Canada

T
he vapour-phase dehydration of 2-propanol is widely used as a
reaction to qualitatively characterize the acidity of solid catalysts The aqueous-phasecatalytic dehydration of 2-propanol
and to study the mechanism of the catalytic dehydration of was investigated in a batch slurry reactor. Alumina,
alcohols over solid acids (lain and Pillai, 1967, as per references; zeolite 13X, SAPO-5 and silicalite are all active in the
Jacobset al., 1977; Cervasini and Auroux, 1991; Cervasini et al., 1997). liquid phase dehydration of 2-propanol at 463 K with
The vapour-phase dehydration of 2-propanol over a gamma alumina silicalite being the most active catalyst. Propylene was
catalyst has also been proposed for propylene production (Fukuhara et found to be the major reaction product, with
diisopropyl ether and acetone formed in trace amounts.
al., 1993), although it is not common practice to do so.
The reaction kinetics over silicalite was determined at a
Several solid acid catalysts have been reported to be active in the
temperature range of 434 to 463 K and at a concentration
vapour phase dehydration of 2-propanol. These catalysts include range of 4 to 10 mol% 2-propanol in water. A single
alumina (Cervasini and Auroux, 1991; Cervasini et al., 1997; Mostafa et site Langmuir-Hinshelwood-Hougen-Watson (LHHW)
al., 1991), zeolite 13X (Bink et al., 1985; Yue and Olaofe, 1984a; 1984b) type mechanism was found to describe the kinetic
and Amberlite I. R. 200 ion exchange resin (Cottifredi et al., 1968). The data well. Water was found to inhibit the dehydration
dehydration reaction is often carried out in a plug-flow reactor system reaction. The activation energy over silicalite was
with no water in the feed. determined to be 226.8 kJ/mol while the heat of
The liquid phase dehydration of 2-propanol has potential application adsorption for 2-propanol and water was 4 5 . 5 and
for wastewater purification where 2-propanol is present as an impurity. -9.6 kJ/mol, respectively.
To design a process for a 2-propanol/water separation, kinetic data for
On a etudie la deshydratation catalytique en phase
2-propanol dehydration in an aqueous media are essential. Although the
aqueuse du 2-propanol dans un reacteur b suspen-
vapour-phase dehydration of 2-propanol has been well studied, there sions discontinu. Les catalyseurs d‘alumine, zeolite
exists no information in the literature on the liquid phase dehydration of 13X, SAPO-5 et silicalite sont tous actifs dans la
2-propanol. Furthermore, the influence of a high concentration of water deshydratation en phase liquide du 2-propanol b 463 K,
on the rate of 2-propanol dehydration was not well understood. la silicalite &ant le plus actif. Le propylhe est apparu
In this study, performances of several well-known acid catalysts iitre le principal produit de reaction, avec formation
including alumina, zeolite 13X, SAPO-5 and silicalite were evaluated with tie diisopropylkther e t d’acetone b I’etat de traces. La
an initial 2-propanol concentration of 10 mol% a t 463 K in a batch slurry cinhtique de reaction sur la silicalite est determinee a
reactor. The kinetic study was carried out a t temperatures ranging from Line gamme de temperatures de 434 b 463 K et b une
gamrne de concentrations de 4 b 10% molaires de 2-
434 to 463 K and initial 2-propanol concentrations of 4 to 10 molYo. The
propanol dans I’eau. On a trouve qu’un mecanisme de
kinetic equation was derived based on a Langmuir-Hinshelwood- type Langumuir-Hinshelwood-Hougen-Watson (LHHW)
Hougen-Watson (LHHW) type reaction mechanism. The influence of b site unique decrivait bien les donnees cinetiques. On
water on the rate of 2-propanol dehydration was also examined. a trouvk que I‘eau inhibait la reaction de deshydratation.
LCnergie d‘activation sur la silicalite a 6te determinee
Experimental b 226,8 kJ/mol alors que la chaleur d’adsorption du
Experimental Apparatus 2-propanol et de I’eau sont de 45,s et -9,6 kJ/mol,
All experimental runs were conducted in a batch slurry reactor system respectivement.
(Figure 1). The reactor was a high pressure Parr reactor (Model 4841,
Parr Instruments Inc. USA) made of SS-316 stainless steel. The vessel has Keywords: 2-propanol dehydration, kinetics, silicalite,
sllJrry reactor.
a volume of 320 mL and is equipped with an impeller. A thermocouple
(]-type) was used to measure and control the reactor temperature. The
heating device/controller was used to maintain a constant reactor
temperature within *1 K. The liquid sampling line consists of a 0.125 m m 0.d. stainless steel tube connected to a stainless steel
sampling valve. A pressure transducer (Foxboro
electronic transmitter, Model 841 CM-D) was used to
meawre the reactor pressure.
‘Author to whom correspondence may be addressed. E-mail address: karlt.chuang@
ua1berta.m

The Canadian Journal of Chemical Englneerlng,Volume79, August 2001 517


The fugacity coefficient for a component in a gas mixture I$
Helium Purge Outlet
I#'~
4
Helium Purge Inlet
and for a pure liquid can be calculated using a modified

f i A
Peng-Robinson equation of state (Stryjek and Vera, 1986). The
Peng-Robinson equation of state was chosen because of its
ability to accurately reproduce vapour-pressure data for a wide
variety of substances, such as polar, nonpolar and associating
compounds. For the vapour phase, conventional mixing rules
were used with the binary interaction term kii set equal to zero
Liquid Sampling Line He
because the non-idealities associated with polar compounds
and their mixtures are not as pronounced in the vapour-phase as
in the liquid-phase. The liquid phase activity coefficient yi must be
determined experimentally by fitting experimental pressure-
composition data with an excess Cibbs free energy model. Because
%
of the small range in liquid concentration (4 to 10 mol%) used in
Stirrer this study, the van Laar activity coefficient model was used:

Cij
Figure 1. Schematic diagram of experimental setup. lnyi =
(3)
[1+$]2

Analysis Techniques and Experimental Procedure


The binary interaction parameters C, and ,C, (1 = 2-propano1, 2
The liquid samples were analyzed using a Hewlett Packard 5890
= water) can be determined a t different temperatures by fitting
Series II Gas Choromatograph with a TCD detector. A 1.0 m
the pressure-liquid composition diagram.
long column with Poropak R packing (mesh 50-80)was used to
separate 2-propano1, water, diisopropyl ether and acetone. The
column temperature was kept constant at 423 K.
During the kinetic runs, a known amount of 2-propanol (analytical (4)
grade, BDH), deionized water and dried catalyst was fed to the
reactor. The 2-propanol/water mixture corresponds to a 2-propanol
The binary interaction parameters for Equation (3) above
concentration ranging from 4 to 10 mol% with a total volume of
correspond to the limiting activity coefficient at infinite dilution.
150 mL. This concentration range was chosen to represent
The governing thermodynamic relationship for the limiting
conditions one may encounter in wastewater purification, where 2-
activity coefficient at infinite dilution is:
propanol concentration is normally less than 10 molYo. The reactor
was then purged with helium to remove any air that might be
present initially in the reactor. The inlet and outlet of the reactor
were then closed. The amount of helium initially present in the
reactor prior to heating can be determined based on the initial
[-]p,x =
hi"
R
reactor temperature, pressure and vapour volume. The reactor was
Over a small temperature range, a plot of Cii versus 1/ J should
then heated to the desired reaction temperature, between 434 and
yield a straight line.
463 K. The mixture of 2-propano1, water and catalyst was stirred at
If we neglect trace amounts of diisopropyl ether and acetone
1080 rpm for 1.5 to 2.5 h. During the catalyst screening tests,
formed during the dehydration, for the liquid phase we have:
liquid samples (-1 mL each) were taken every half-hour to
determine product distributions using a gas chromatograph. In all xw =1-x, (6)
cases, gas phase data (pressure versus time) were used to establish
2-propanol conversion as described in the model development
The cumulative moles of propylene (a)formed at reaction time
section, thus avoiding the need for liquid sampling.
t can thus be obtained from the reaction stoichiometry and the
overall material balance:
Kinetic Modeling v = MTo + a - L (7)
Rate Determination
The rate of propylene formation per gram of catalyst in a batch
Under reaction conditions, the solubility of propylene in the
slurry reactor can be written in the following manner:
liquid phase is small enough to be negligible. From the propylene
and helium material balance we obtained:
a
Propylene mole fraction:
"=MTO+a-L
Since the dehydration occurs only in the liquid phase, 2-propano1,
water and product propylene can be assumed to be at their
vapour-liquid equilibrium a t any given reaction time, t. For any
of these components, i, vapour-liquid equilibrium is given as: Helium mole fraction: MH
Y"=MTO+a-L (9)
@:Y, = P@P'Y,X, (2)

518 The Canadian Journal of Chemical Engineering, Volume 79, August 2001
The other vapour phase components, 2-propanol mole catalytic dehydration of 2-propanol in the liquid phase. The
fraction, yA and water vapour mole fraction, yw, are determined adsorption/clesorption of 2-propanol and/or water is assumed
using the modified Peng-Robinson equation of state and the to either occur via a single-site mode, where acid sites are
van Laar model. The liquid phase fugacity coefficient is involved or ei dual-site mode, where both acid and basic sites are
determined for a pure component and the vapour phase fugacity involved. The rate of propylene formation is considered to be
is determined for a component in a gas mixture (Stryjek and Vera, either inhibited or unaffected by the presence of water in our
1986). It is sufficient to say that $'!, is a function of all vapour phase model screening process. The chemical equilibrium constant for
composition and the compressibility factor for a gas mixture Z. the liquid-phase dehydration of 2-propanol at a temperature
We are now left with 6 unknowns: xA, yA, yw, L, a and Z. The range of 433 to 463 K ranges from 51.8 to 114.2. Because of
following outlines the set of nonlinear equations that must be the high values of the equilibrium constant, the reaction was
solved for each data point (pressure and temperature). assumed to be irreversible. The derived kinetic equations are
expressed a:i the following:
Compressibility Factor Z
Single-site mode adsorption, water term excluded (SSM - I ) :
Z3 + (5- l)Zz + ( A -3B2 -25)Z + ( B 3 + 52 - A61 = 0 (1 0)

A and 5 are parameters which depend on the vapour phase


composition, temperature and pressure. The largest root of this Dual-site mode adsorption, water term excluded (DSM - I ) :
cubic equation corresponds to the vapour phase compressibility
factor.

Water component balance: Singlesite mode adsorption, water term included (SSM - 2):
L(I - x A ) + ( M + a~ - L~ ) ~ W= MWO + a kb:AC,q
r=-
1 + KACA + KwCw

Vopour-liquidequilibrium for 2 - propanol:


Dual-site mode adsorption, water term included (DSM - 2):
Y A ~ V A = xAYAQOA~

Vapour-liquidequilibrium for water:


For models SSM-2 and DSM-2, the rate of propylene
YW& = (l- xA)YW@E formation is inhibited by the adsorption/desorption of water on
the active sites. The temperature dependence of the kinetic and
adsorption parameters can be expressed mathematically as:
Summation of vapour-phase mole fractions = I:
a+M,
=1 k = k"exp
\ RT
(3)
Y A YW
+ + M-..+ n - I

Equation of state for a closed system:

The pararrieters were determined by fitting the above equations


ZRT(MT, +u -L) with experimental rate data. The rate equation was integrated
numerically and the model parameters were determined from
Saturated densities pA and pw can be calculated using the least-squares nonlinear regression.
COSTALD correlation (Hankinson and Thomson, 1979).
From the above analysis, the rate of reaction can thus be
Results and Discussion
described using only pressure and temperature data. As many Determination of the Liquid-Phase van Laar
data points can be taken as required without influencing the Parametlers for 2-Propanol/Water Mixture
reactor condition. The only requirement for using such a Total pressure-liquid composition data were collected over a
method is to have adequate vapour-liquid equilibrium (VLE) temperature range of 433 to 463 K and a 2-propanol concentration
data at our disposal. Since VLE data for 2-propanoVwater at range of 4 to 10 mot% using a similar procedure described
elevated temperatures are not readily available, experiments earlier but without added catalyst. The liquid-phase van Laar
must be performed to acquire the required information. binary pairameters were determined by nonlinear least-squares
regression, of Equation (15) and with assumption of the binary
Kinetic Equation vapour-interaction parameters are negligible (Figure 2). The
With the surface reaction considered as the rate-determining temperature dependence of the binary parameters, as
step, several LHHW type rate equations were developed for the expressed mathematically in Equation (5), is shown in Figure 3.

The Canadian Journal of Chemical Engineering,Volume 79, August 2001 519


Catalyst Screening
Samples of active alumina, zeolite 13X (Aldrich), SAPO-5
1378 L (MHZN2-34, Lava1University, PQ, Canada), silicalite with a silica
binder (S-115 SO i , UOP, LOT 15228-32) and silicalite with an
m
a alumina binder (S-115 AO ,I , UOP, LOT 09296-29C) were
1034~
9; tested for activity in aqueous phase 2-propanol dehydration.
1 Alumina and zeolite 13X were selected for screening due to
$ 0689~ their activity in vapour-phase dehydration of 2-propanol.
b -------* Silicalite is hydrophobic (Flanigen et at., 1978) and is considered

P
0345
structurally similar to HZSM-5 catalysts, which are found to be
active in ethanol dehydration (Olson et al., 1980; Rees, 1982;

0
0 00
> 0 02 004 008 0 08 0 10 0 12
Oudejans et al., 1982; Le Van Mao et al., 1990; Phillips and Datta,
1997). Silicoaluminophosphates (SAPO-5) are known to be
active in reactions requiring strong acids (Hedge et al., 1988).
The results of the catalyst screening are shown in Figure 4.
2PropanoJ Mole Fradion, XA’
The 2-propanol conversion is defined as the mole percent of
2-propanol that is converted to propylene. For all catalysts
Figure 2. The equilibrium pressure versus liquid composition: PRSV tested, propylene is the major reaction product. Only trace
equation of state with the van Laar excess Gibbs free energy model. +: amounts of diisopropyl ether and acetone were found. These
433 K; 0:443 K; A : 453 K; W : 463 K; - : van Laar model fit. results justify our assumption of a negligible amount of other
components in the liquid phase except water and 2-propanol.
Analysis of the liquid samples taken prior to starting of agitation
indicated negligible 2-propanol conversion during heating
period. The amount of 2-propanol converted to propylene over
It should be noted that these parameters were determined 5-115 AI,O, a t 463 K after 1.5 h is 5.1 46, which is far from the
under a relatively small range of 2-propanol concentrations. calculated equilibrium conversion of 35.3% at 463 K for 1.5 h
Caution should be exercised in extrapolating these parameters and an initial 2-propanol concentration of 10 mo146. The
to other concentration ranges. An accurate description of the assumption of an irreversible reaction appears to be justified.
vapour-liquid equilibria for 0 < xA c 1 is outside the scope of this S-115 AI,O, was found to be the most active catalysts of the
paper. The temperature dependence of the van Laar binary screened samples followed by S-115 SO,. In aqueous media,
constants was determined to be: the rate of propylene formation not only depends on the
number and strength of acid sites, but on the hydrophobicity of
the catalyst as well. It has been found previously that silicalite is
both more hydrophobic and more acidic than SAPO-5 (Hedge
c,2 =--2299’1 3.624, r 2 = 0.995 et al., 1988). This explains the higher activity of the silicalite
T
catalysts compared to the SAPO-5 catalyst. It should be noted
that the acidity and hydrophobicity of zeolites, such as silicalite,
depend on the silica/alumina ratio (Olson et al., 1980).
However, more tests need to be performed on the catalyst with
-+ 5.953,
c2, = -2228.6 r 2 = 0.893
respect to its acidity and hydrophobicity in order to make
T

10-

10.
0
.
14.
0
12’
0

10-

..
J 0
08-

..
“‘L
00
2 14 218 210 220 222 224 228 220 230 232
* 0 10 20 30 40 50 60 70 80 90 1W
100011. K
Time. min

Figure 3. Tempemture dependence of the b i ~ parameters


y for the van bar Figure 4. Activities of different catalysts at 463 K with a 1.6 wt% catalyst
excess Cibbs free energy model. 1 : 2-propanol; 2: water; V: C12;W : CZl. loading and initial 2-propanol concentration of 10 mol% in water.

520 The Canadian Journal of Chemical Englneerlng, Volume 79, August ZOO1
Table 1. Parameters of the developed LHHW kinetic models.

Model E* r2 AH/lA z AHa, W z model fit


(klimol) (kJ/mol) (kJ/mol) (%I
SSM-1 209.1 0.994 -26.9 0.991 18.2
DSM-1 199.7 0.996 -36.4 0.842 26.7
SSM-2 226.8 0.996 45.5 0.987 -9.6 0.994 8.8
DSM-2 166.6 0.981 -30.3 0.870 -30.5 0.835 11.8

*%Model Fit = -

definite conclusions with regards to the influence of these The kinetic parameters established using the run with 10 mol%
parameters on the catalytic activity in the aqueous phase. All initial 2-propanol concentration were used to predict the rate of
kinetic runs were carried out with silicalite S-115 AI,O, catalyst. propylene formation at other initial 2-propanol concentrations.
It can be seen in Table 1 that the SSM-2 model, which has
Kinetic Study single-site adsorption of both 2-propanol and water in the rate
True surface reaction rates can only be determined if the experiments model predict well the rate of propylene formation for 2-propanol
are performed under conditions where the observed rate is not concentrations ranging from 4 to 10 mol%. Both models that
limited by diffusional processes. Our preliminary tests with include the water adsorption/desorption parameter predict well
silicalite 5-1 1 5 AI,O, catalyst at an experimental condition of the rate of propylene formation, which indicates a possible
463 K, 1.57 wt% catalyst loading and 10 mol% 2-propanol water inhibiting effect on propylene formation. This partly
indicate that that the reaction rate was not limited by interphase explains the poor dehydration activity of some of the
mass transfer and pore diffusion at stirrer speeds above 1000 hydrophilic catalysts such as zeolite 13X and SAPO-5 found
rpm and over the range of particle sizes tested (20-170 mesh). during the catalyst screening.
No apparent catalyst deactivation was observed in duplicate Figure 5 illustrates the amount of propylene produced with
experiments with the used catalysts. All kinetic experiments varying timi? a t several reaction temperatures with the fitted
were conducted with fresh silicalite S-115 AI,O, catalyst a t stirrer SSM-2 model. The model gives a good fit of the experimental
speeds of 1080 rpm with 30-40 mesh catalyst particle sizes. data for temperatures between 434 and 463 K. The temperature
Typical catalyst concentration was 1.57 wt% in 2-propanol/ dependence of the rate constant, as seen in Equation (20), is
water/catalyst mixture. The experimental data were fitted with shown in Figure 6. As can be seen from Figures 5 and 6, the rate
several LHHW rate equations (Equations 16, 17, 18, and 19). of propylent! formation is strongly dependent on temperature.
Activation energies and heats of adsorption calculated from the The high activation energy confirms that the kinetic data were
above fitted models are given in Table 1. obtained in a region where the diffusion was not rate limiting.
A good kinetic model should not only fit the experimental The temperature dependence of the adsorption constants for
data but also to predict the reaction rates under different the SSM-2 rate model is shown in Figure 7. The adsorption plots
conditions. To this end, additional kinetic runs were carried out have positive slopes, which indicates that the heat of adsorption
at 463 K for 2-propanolfeed concentrations of 4, 6 and 8 molOh. is exothermic. Figure 8 shows a reasonable agreement between
SSM-2 model prediction and experimental rates of propylene

"""I

0 20 40 60 80 100 120 140 160 214 216 218 220 222 224 228 220 230 232
100M. K "
Time, nin

Figure 5 . Comparison of experimental rate data with fitted SSM-2 Figure 6. Teinperature dependence of the kinetic parameters k,
model. *:434 K; A:444 K; 0 :453 K; M: 463 K; -: SSM-2 model fit. determined from the fitted SSM-2 LHHW model.

The Qnadlan Journal of Chemlcal Engineerlng,Volume 79, August ZOO1 52 1


Conclusions
Alumina, zeolite 1 3X, silicoaluminophosphate and silicalite were
active for the catalytic dehydration of 2-propanol at 463 K. Of
these catalysts, silicalite (5-1 1 5 AI,O,) was the most active
catalyst for 2-propanol dehydration in aqueous media. The
main product of 2-propanol dehydration is propylene with
trace amounts of diisopropyl ether and acetone.
A single site Langmuir-Hinshelwood-Hougen-Watson type
mechanism was found to describe well the kinetic data obtained
over a 5-1 15 AI,O, catalyst for 2-propanol concentrations
ranging from 4 t o 10 mol% and temperatures ranging from
434 t o 463 K. The model indicates a possible water inhibition
00 I I
214 218 218 220 222 224 220 228 230 232 on the rate of propylene formation. The activation energy was
1OOOTT. K ” determined t o be 226.8 k]/mol and heat of adsorption for
2-propanol and water was -45.5 and -9.6 k]/mol, respectively.
Figure 7. Temperature dependence of the adsorption/desorption
equilibrium constants determined from the fitted SSM-2 LHHW model.
Nomenclature
+ : 2-propanol; W : water. a, b equation of state parameters for PRSV equation of state
dimensionless terms, [ A = PA/(RQ2; 6 = Pb/RTj
A, 6
C binary constant for van Laar equation
liquid phase concentration of component, (mol/L)
EA activation energy, (kJ/mol)
AHA heat of adsorption, (kJ/rnol)
h; partial molar excess enthalpy of component i at infinite
dilution, (kJ/mol)
k binary interaction parameter for PRSV equation of state
rate constant for 2-propanol dehydration, (mol.g-lmin-l)
K adsorption/desorption equilibrium constant, (Umol)
ko pre-exponential factor for rate factor, (mo1.g-lmin-1)
KO adsorption/desorption pre-exponential factor, (L/mol)
L moles of liquid, (mol)
m mass, (9)
M moles of component prior to reaction, (mol)
Mw molecular weight of component, (g/mol)
0 20 40 60 80 100 120 140 160
NP number of data points
Time, min
P pressure, (kPa)
R gas constant, (8.31451 Jmo1-l.K-l)
Figure 8. Comparison of SSM-2 model prediction with the rate of t time, (min)
propylene formation under different initial 2-propanol concentrations: T temperature, (K)
A: 4 mol%; A: 6 mol%; W: 8 mol%; 0 : 10 mo196;-: SSM-2 model. V moles of vapour, (mol)
X liquid mol fraction
Y vapour phase mol fraction
Z compressibility factor
formation a t lower 2-propanol concentrations. All other models
are not sufficient to describe the data, especially a t high water Greek Symbols
concentrations. The rate equation established is: a cumulative amount of propylene produced at time t, (mol)
Y activity coefficient
P saturated liquid density, (g/mL)
z summation
+ fugacity coefficient

Subscripts
k = 1 .406.1023exp-
r2Y48) 0
A
C
initial, t = 0
2-propanol
catalyst
EX P experimental
K, = 8.32.10-’exp
Y21
(- H
..
‘,I
P
helium
component
propylene
Kw =1.66.10-2exp(T)
1154 PRED value predicted from model
T total
W water

522 The Canadlan Journal of Chemical Englneerlng, Volume 79, August ZOO1
Superscripts lain, J.R. and C.N. Pillai, “Catalytic Dehydration of Alcohols over
0 purr phase Alumina - Mechanism of Ether Formation”, J. Catal. 9, 322-330
m infinite dilution (1 967).
I liquid phase Le Van Mao, R., T.M. Nguyen and J. Yao, “Conversion of Ethanol in
V vapour phase Aqueous Solution over ZSM-5 Zeolites. Influence of Reaction
Parameters and Catalyst Acidic Properties as Studied by Ammonia
References TPD Technique”, Applied Catalysis 61, 161-1 73 (1990).
Bink, R.H., W.] Thomas and P.L. Yue, ”Effect of Mass Transport and Mostafa, M.F., A.M. Youssef and S.M. Hassan, “Conversion of Ethanol
Catalyst Deactivation on the Kinetics of Ethanol Dehydration on and IsoproiDanol an Alumina, Titania and Alumina-T‘itania Catalysts”,
Zeolite 13X”, Chem. Eng. Res. Des. 63, 338-344 (1985). Materials Letters 12, 207-21 3 (1991).
Flanigen, E.M., J.M.Bennett, R.W. Crose, J.P. Cohen, R.L. Patton, R.M. Olson, D.H., W.O. Haag and R.M. Lago, “Chemical and Physical
Kirchner and J.V. Smith, ”Silicalite, a New Hydrophobic Crystalline Properties of the ZSM-5 Substitutional Series”, J.Catal. 61, 390-396
Silica Molecular Sieve”, Nature 271, 51 2-51 6 (1978). (1 980).
Fukuhara, H., F. Matsunaga, M. Yasuhara, 5. Araki and T. Isaka, Oudejans, J.C., P.F. van den Oosterkamp and H. van Bekkum,
“Preparation of Propylene by Dehydration of lsopropanol in the “Conversion of Ethanol over Zeolite H-ZSM-5 in the Presence of
Presence of a Pseudo-Boehmite Derived Camma Alumina Catalyst”, Water”, Applied Catalysis 3, 109-1 15 (1982).
U.S. Pat. 5, 227-563 (1993). Phillips, C.B. and R. Datta, “Production of Ethylene from Hydrous
Cervasini, A and A. Auroux, “Acidity and Basicity of Metal Oxide Ethanol or1 H-ZSM-5 Under Mild Conditions”, Ind. Eng. Chem. Res.
Surfaces II Determination by Catalytic Decomposition of 36, 4466-4475 (1 997).
Isopropanol”. I . Catal. 1 3 1 , 190-198 (1991). Rees, L.V.C., “When is a Zeolite not a Zeolite?”, Nature 296, 491 -492
Cervasini, A,, Fenyvesi and A. Auroux, “Study of the Acidic Character (1 982).
of Modified Metal Oxide Surfaces Using the Test of lsopropanol Stryjek, R. arid J.H. Vera, “PRSV An Improved Peng-Robinson Equation
Decomposition”, Catalyst Letters 43, 21 9-228 (1 997). of State for Pure Compounds and Mixtures”, Can. J. Chem. Eng. 64,
Cottifredi, J.C., A.A. Yeramian and R.E. Cunningham, “Vapour-Phase 323-333 ((1986).
Reactions Catalyzed by Ion Exchange Resins 1 . lsopropanol Yue, P.L. and 0. Oloafe, “Kinetic Analysis of the Catalytic Dehydration
Dehydration”, J, Catal. 12, 245-256 ( 1 968). of Alcohois over Zeolites”, Chem. Eng. Res. Des. 62, 81-91 (1984a).
Hankinson, R W. and C.H. Thomson, “A New Correlation for Saturated Yue, P.L. and 0. Oloafe, “Molecular Sieving Effects of Zeolites in the
Densities of Liquids and Their Mixtures”, AlChE J. 25, 653-662 (1979). Dehydration of Alcohols”, Chem. Eng. Res. Des. 62, 167-1 72
Hedge, S.C., P Ratnasamy, L.M. Kustov and V.B. Kazansky, “Acidity and (1 984b).
Catalytic Activity of SAPO-5 and AIPO-5 Molecular Sieves”, Zeolites
8, 137-141 (1988).
lacobs, P.A, M. Tielen and 1.6. Uytterhoeven, ”Active Sites in Zeolites.
Part 6 Alcohol Dehydration over Alkali Cation-Exchanged X and Y Manuscript received June 12, 2000; revised manuscript received May
Zeolites”, ] Catal 50, 98-108 (1977). 7, 2001; accepted for publication May 29, 2001.

The Canadian Journal of Chemical Engineering,Volume 79, August 2001 523

You might also like