Upload
sandip-talwalkar
View
235
Download
1
Embed Size (px)
Citation preview
www.elsevier.com/locate/apcata
Applied Catalysis A: General 302 (2006) 140–148
Synthesis of methyl isobutyl ketone from acetone over
metal-doped ion exchange resin catalyst
Sandip Talwalkar, Sanjay Mahajani *
Department of Chemical Engineering, Indian Institute of Technology Bombay, Powai, 400076 Mumbai, India
Received 21 September 2005; received in revised form 29 December 2005; accepted 4 January 2006
Available online 14 February 2006
Abstract
The kinetics of one-step synthesis of methyl isobutyl ketone from acetone was studied in the presence of the bifunctional commercial ion
exchange resin, Amberlyst CH28 over a wide range of temperature, total pressure and catalyst loading in a batch reactor. An activity-based kinetic
model is proposed to predict the observed results, with the non-idealities of the liquid phase being described using the UNIQUAC method.
Formation of mesityl oxide was found to govern the overall rate of reaction. Low reaction rates were observed at higher conversion, possibly due to
a pseudo-equilibrium caused by reversible deactivation of the catalyst as a result of formation of water in the reaction system. Simultaneous
removal of water during the course of the reaction may result in an enhanced conversion.
# 2006 Elsevier B.V. All rights reserved.
Keywords: Methyl isobutyl ketone; Bifunctional catalyst; Hydrogenation; Ion exchange resin; Acetone; Mesityl oxide
1. Introduction
Cation exchange resins are popular solid acid catalysts for
liquid phase reactions conducted under relatively mild
conditions. Some excellent reviews on catalysis by cation
exchange resin have appeared in the past [1,2]. However, not
much information is available on catalysis with a bifunctional
ion exchange resin catalyst, as it is a relatively new field.
Multiple reactions involving acid catalysts followed by
hydrogenation or dehydrogenation and vice versa are com-
monly encountered in many industrial processes. If the
conditions for both the reactions are overlapping, one can
advantageously perform them in a single step with a
bifunctional catalyst. Aldol condensation followed by hydro-
genation of dehydrated aldol is an important class of these
reactions [3–5] for which such a bifunctional catalyst can be a
potential candidate. An industrially important reaction of aldol
condensation of acetone, followed by hydrogenation to methyl
isobutyl ketone, has been studied in the present work.
Methyl isobutyl ketone is a widely used solvent in the
pharmaceutical, coating and mining industries. It is also used in
the manufacture of rubber antiozonants. Its synthesis from
* Corresponding author. Tel.: +91 22 25767246; fax: +91 22 25726895.
E-mail address: [email protected] (S. Mahajani).
0926-860X/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2006.01.004
acetone in the presence of acidic and hydrogenation catalyst
consists of three reaction steps. They are as follows:
� R
eaction 1:Reaction 2:
�Reaction 3:
�Several references in the form of communications,
research articles and patents are available in the literature
on the synthesis of methyl isobutyl ketone from acetone.
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148 141
Nomenclature
ai activity of component i
E8 activation energy for reaction (kJ/mol)
k8 rate constant for reaction (kmol/(s gcat))
kwtr adsorption coefficient of water
kact adsorption coefficient of acetone
MCAT mass of the catalyst (g)
ni number of moles of component i
nAc,0 initial number of moles of acetone
nAc,f number of moles of acetone at any time t
r rate of reaction (kmol/s)
S selectivity towards methyl isobutyl ketone
T Temperature (K)
xi liquid mole fraction of component i
X conversion of acetone
Greek letters
yi stoichiometric coefficient for component i
F objective function for optimisation
However, the use of bifunctional catalyst for one-step
synthesis is gaining importance and substantial work has
been reported with different metals from Groups VIII and
IB of the Periodic Table like Pd, Cu and Pt on acid supports.
Table 1 gives a review on the previous studies on this system
with bifunctional catalysts. It is clear from Table 1 that
Pd is the most widely used metal for this system with almost
all types of acid/base catalytic supports like alumiosi-
licates, zeolites, metal oxides and ion exchange resins.
Different catalysts, like Pd-ZSM-5 [24], Pd–C–Nb2O5
[12,13] and Pd–CuO/MgO/SrO [8] in liquid phase and Pd-
H-ZSM-5 [37], Cu–MgO [26] and Pd-CS-H-ZSM-5 [16] in
gas phase, have been studied. They show conversions of
acetone in the ranges 20–40% and 40–60% and selectivity
towards methyl isobutyl ketone of �90% and 30–80%,
respectively. It is reported that the catalyst used in the
present work gives a conversion as high as 50% with 90%
selectivity towards methyl isobutyl ketone in a liquid phase
reaction [45]. The supports other than ion exchange resins
need pre-treatment; indeed, the performance of these
supports is sensitive to the method and conditions of the
pre-treatment. Moreover, the reaction conditions are more
severe than those with ion exchange resins without much
improvement in conversions and selectivities. Ion exchange
resins are more popular than traditional supports for low
temperature operations due to the ease of separation, less
pre-treatment and high acid strength. A commercial
bifunctional ion exchange resin, Amberlyst CH28, has been
designed for such performance and has been investigated in
the present work.
Ion exchange resin can be loaded with desired metal ions by
contacting an aqueous solution of the metal ion with the
hydrogen form of the cation exchange resin in a batch or
continuous mode. Typically, the metal ion will be provided in
the form of a metal salt, such as chlorides, bromides, nitrates,
sulphates and acetates. The detailed procedure for preparation
for such catalysts can be found elsewhere [9].
In spite of a lot of literature on the present system, very little
work is reported on commercial metal-ion exchange resins as a
catalyst. Nicol and du Toit [45] have successfully performed
reactions with the same catalysts in a laboratory-scale trickle
bed reactor. A couple of patents from Catalytic Distillation
Technologies [50–53] claim the use of catalytic distillation for
enhanced conversion of acetone to methyl isobutyl ketone
preferably catalysed by metal-ion exchange resin as a catalyst.
Systematic kinetic studies and a reliable kinetic model are
necessary to design an industrial reactor; the present work is
undertaken to provide inputs in this regard.
2. Experimental
2.1. Materials
Acetone (99.5%), methyl isobutyl ketone (99%) and
methyl ethyl ketone (99.5%) were obtained from Merck Ltd.,
India. Rohm and Haas, France, supplied the catalyst
Amberlyst CH28. The properties of catalyst are given in
Table 2. Before its use, the catalyst was dried in an oven at
353 K for 3 h.
2.2. Apparatus and procedure
A stainless steel autoclave from Parr Instrument Company,
USA, with a capacity of 3 � 10�4 m3, equipped with an online
temperature monitoring and control facility was used for
conducting all the batch reactions. Before feeding to the
reactor, each catalyst was washed with acetone in order to
remove the moisture present, if any. The desired quantities of
the catalyst and reactants (100 g of acetone in all runs) were
charged to the reactor. The reactor and gas lines were flushed
with hydrogen in order to remove the air present in the empty
space in the reactor and lines. The reaction mixture was heated
up to the desired temperature with slow stirring. As the
reaction temperature was reached, the speed of agitation was
increased up to the desired level and the corresponding time
was regarded as the zero reaction time. The samples were
withdrawn at different time intervals to study the kinetics of
the reaction. In all the runs, the reaction volume was about
8 � 10�5 m3.
2.3. Analysis
The reactants and products were analysed using a gas
chromatograph (GC-MAK-911) equipped with a flame ionisa-
tion detector (FID). A 25 m long capillary column BP-1 (SGE,
Australia) was used to separate the different components in the
reaction mixture using methyl ethyl ketone as an external
standard. The column temperature was maintained at 373 K
isothermally. The various components in the reaction mixture
and the separated products were characterized by authentic
samples.
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148142
Table 1
Previous work for one-step synthesis of methyl isobutyl ketone from acetone
Catalyst Pressure (atm) T (K) Conversion (%) Selectivity (%) Reference
Pd-KUZ 20 393 50 94.5 [6]
Pd–CuO/Al2O3/SiO2 1 423–503 30 60 [7]
Pd–CuO/MgO/SrO 20 433 38.5 93.6 [8]
Pd–Nb2O5–Al2O3 – – 28 90 [9]
Pd-oxides of Ti, Zr, Cr 10 413 33.9 92.3 [10]
Pd-KUZFPP 30 353–388 – – [11]
Pd–C–Nb2O5 20 413 39.5 92.5 [12]
Pd–C–Nb2O5 10 413 30 91.7 [13]
Pd-oxides of Ce, Hf, Ta 10 413 33 90.2 [14]
Pd-IER 40 363–393 - – [15]
Pd-CS-H-ZSM-5 1 523 41.9 82.4 [16]
Pd–Nb2O5 – – 41.8 93.5 [17]
Pd-oxides of Zr – – 27 94.9 [18]
Pd-KS-IER 50 373–403 - 90 [19]
Pd-ZSM-5 50 443 40.25 95.36 [20]
Pd-IER – – – – [21]
Ni, Cu, Co-g-a-Al2O3 – 453 53.9 37.1 [22]
Ni–Al2O3 1 373 – 95 [23]
Pd-ZSM-5 SiO2/Al2O3 = 30 6–60 433 41.24 90.98 [24]
Ni–CaO – 473 70–80 60 [25]
Cu–MgO 1 653 60–80 60–75 [26]
Pt-HMF 1 433 – – [27]
Ni–CaO – – 60–80 50–60 [28]
Pd–AIPO4-II, SAPO4 – – – – [29]
Ni–Al phosphate – – – – [30]
Pd-(Zn)-H-ZSM-5 5 408–483 55 83–94 [31]
Pd–Nb2O5–SiO2 – – 30–35 88–92 [32]
Ni–CaO – – 60–70 70 [33]
Pd–Al2O3 40–90 413–473 – – [34]
Pt-H-ZSM-5 1 433 – – [35]
Pt-NaX – 713 – 70 [36]
Pd-H-ZSM-5 1 473 47.3 30.7 [37]
Pd-IER 40–70 353–373 – – [38]
Cu–MgO–Al2O3 – – 65.26 57.49 [39]
Pd-zeolite – – 70 87.5 [40]
Pd-IER – 403 43.2 98.2 [41]
Pt-Sn-H-ZSM-5 1 433 – – [42]
Pd–C 1-20 333 – – [43]
Pd–Ca–Al2O3 - - – – [44]
Amberlyst CH28 30 403–423 25–50 70–90 [45]
Pd-IER 5–15 373–453 – – [46]
Pd-MCM-56 – – 33.5 81.2 [47]
Pd-MCM-49 – – 35.6 85 [48]
Cu–MgO–Al2O3 1 513 71.7 50.1 [49]
Table 2
Physical properties of Amberlyst CH28 [59]
Property Value
Matrix Styrene DVB polymer
Physical form Spherical beads
Total capacity (mequiv./g dry basis) 4.8
Palladium loading (% dry basis) 0.7
Specific surface area (m2/gm) 36
Pore diameter (A) 260
Temperature stability (K) 403
Harmonic mean size (mm) 0.85–1.05
Moisture content in wet form (%) 52–58
2.4. Calculations for conversion and selectivity
The conversion of acetone was calculated as
X ¼ nAc;0 � nAc;fnAc;0
(1)
while the selectivity towards methyl isobutyl ketone was
calculated as
S ¼ 2nMIBK
nAc;0 � nAc;f(2)
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148 143
Fig. 2. Effect of speed of agitation on reaction kinetics. Temperature = 413 K;
catalyst loading: 5% w/w of acetone; pressure = 30 bar; speed in rpm.
3. Results and discussion
3.1. General course of the reaction
As mentioned earlier, we define the zero reaction time as the
time at which the desired temperature is attained. Hence, in all
the kinetic runs, we observe a small extent of reaction that
occurred before the desired temperature was reached. The
extent of reaction during this heat-up period was found to be
relatively higher at higher catalyst loadings and temperature.
Fig. 1 shows typical profiles of mole fractions of acetone and
methyl isobutyl ketone with respect to time. The concentration
of water is the same as the concentration of methyl isobutyl
ketone and is not shown in Fig. 1. It should be noted that the
concentrations of intermediates, diacetone alcohol and mesityl
oxide, were below the detectable limits.
3.2. Mass transfer effects
In order to ensure that there is no significant external mass
transfer resistance across the solid–liquid and gas–liquid
interfaces, the reactions were performed at different stirrer
speeds over a range 600–1700 rpm. It was observed that the
reaction kinetics is insensitive to agitation above 1000 rpm, as
shown in Fig. 2, and hence all the reactions were performed
above 1000 rpm.
3.3. Comparisons of the rates of individual steps
In all the experiments, we have observed very little or no
intermediate products, i.e., mesityl oxide and diacetone
alcohol. Therefore, it was assumed that rapid dehydration of
diacetone alcohol leads to formation of mesityl oxide, which in
turn instantaneously gets converted to methyl isobutyl ketone in
the presence of hydrogen. To compare the rates of mesityl oxide
Fig. 1. General course of the reaction. Temperature = 393 K; catalyst loading:
5% w/w of acetone; pressure = 30 bar.
and methyl isobutyl ketone formation, we performed a run in
the absence of hydrogen for 50 min. The samples were
collected up to 50 min and hydrogen was introduced to the
reactor at this time at 30 bar. Fig. 3 shows that, as soon as
hydrogen was fed to the reactor, the concentrations of mesityl
oxide and diacetone alcohol dropped instantaneously to almost
zero and the methyl isobutyl ketone concentration builds up
sharply. This suggests that the hydrogenation is much faster
than both the intermediate reactions and practically no or very
little intermediate components were observed in the range of
parameters for which the kinetics of methyl isobutyl ketone
formation from acetone was generated on the bifunctional
Amberlyst CH28.
Fig. 3. Speed of hydrogenation reaction. Temperature = 393 K; catalyst load-
ing: 5% w/w of acetone; hydrogen injection at 3000 s at 30 bar.
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148144
Therefore, the overall reaction with the present catalyst is
given by
Fig. 3 also indicates that the rate of mesityl oxide formation is
very low even at low conversion levels, indicating that the
reaction rates are highly affected by equilibrium. Thus, the
simultaneous hydrogenation of mesityl oxide may help in
shifting the reaction in the forward direction by keeping the
mesityl oxide concentration at low level at any given time in the
reactor. Thus, complete conversion of acetone was expected in
the present work because of the removal of mesityl oxide in
hydrogenation reaction to of methyl isobutyl ketone. However,
low reaction rates were observed at higher conversion levels.
This may imply that the hydrogenation reaction is equilibrium-
limited and only equilibrium conversions can be obtained. How-
ever, Winter et al. [43] showed that 100% conversion mesityl
oxide in liquid phase could be possible on Pd/CNF catalyst. This
confirms that the observed low reaction rates onAmberlyst CH28
at higher conversion level are due to catalyst deactivation and not
because of reaction equilibrium. The overall reaction is thus a
pseudo equilibrium type as a result of deactivation of catalyst by
the moisture formed. More evidence of the same can be found in
the literature [54]. We confirmed this through the catalyst
deactivation experiments described in the next section.
3.4. Cause of catalyst deactivation
As discussed above, the catalyst gets deactivated during the
course of reaction, as a result of which low rates were observed
Fig. 4. Cause of catalyst deactivation. Temperature = 393 K; catalyst loading:
5% w/w of acetone; pressure = 30 bar.
at higher conversion levels of acetone. To find out the cause of
catalyst deactivation, the catalyst was separated from the
reaction mixture and dried in an oven at 353 K for 3 h and the
same catalyst was reused in the next run. This procedure was
repeated for next few runs. As shown in Fig. 4, the performance
of the used catalyst was almost identical to that of the fresh
catalyst. This indicates that the deactivation of catalyst is
reversible and that is caused by the presence of polar
components like water [1,2,55], which is produced in the
dehydration of aldol. Polar compounds like water have strong
affinity towards catalytic sites. In a relatively non-polar
environment, water molecules are known to adsorb on the
resin surface and cover the catalytic sites, thereby causing an
‘‘inhibition effect’’. As mentioned before, it should be noted
that the reaction becomes very slow in this region and the
kinetics, as shown in Fig. 1, looks similar to that of a reversible
reaction approaching equilibrium. Since in this case, true
thermodynamic reaction equilibrium is not attained, we call
this state a ‘‘pseudo-equilibrium’’. A detailed investigation of
the same has been reported in the literature for the reaction of
acetone to mesityl oxide on a mono-functional ion exchange
resin [54].
3.5. Effect of moisture concentration
In order to ascertain the exact role of water in the kinetics,
we performed experiments with and without water. As shown in
Fig. 5, we found that the presence of water in the reaction
Fig. 5. Effect of water addition. Temperature = 413 K; catalyst loading: 5% w/
w of acetone; pressure = 30 bar water concentration in wt.%.
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148 145
Fig. 6. Effect of temperature on reaction kinetics. Temperature in K; catalyst
loading: 5% w/w of acetone; pressure = 30 bar.
Fig. 7. Effect of temperature on selectivity towards methyl isobutyl ketone.
Temperature in K; catalyst loading: 5% w/w of acetone; pressure = 30 bar.
system is the true cause of the observed reversible catalyst
deactivation; hence, low rates of reaction are observed for the
reaction with initial moisture. This implies that the rate of
formation of mesityl oxide is affected due to inhibition of
sulfonic acid groups by polar components like water. This
phenomenon is well known for many reactions on ion exchange
resins. However, mesityl oxide formation is still followed by its
instantaneous hydrogenation to methyl isobutyl ketone. In
short, the rate of formation of methyl isobutyl ketone is equal to
the rate of formation of mesityl oxide as in the case without
moisture. This suggests that simultaneous removal of mesityl
oxide along with the removal of water by some means is
necessary to enhance the reactor performance further. There-
fore, one can regard this system to be a good candidate for
multifunctional reactor system like reactive distillation or
reactive adsorption. As mentioned before, except for few
Fig. 8. Effect of catalyst loading on (a) reaction kinetics (b) initial rate. Tem
references [50–53] not much work has been reported in the
literature on this aspect.
3.6. Effect of temperature
The effect of temperature was studied over a range of
373–413 K. Fig. 6 shows the conversion of acetone at different
temperatures. As expected, the conversion of acetone increases
with an increase in temperature at the cost of reduced selectivity
towards methyl isobutyl ketone, as shown in Fig. 7. At
relatively high temperatures, mesityl oxide is likely to react
with acetone to form heavy components like isophorone
and other condensation products. Nicol and du Toit [45]
evaluated the thermal stability of the present catalyst for
the same reaction, and reported that the catalyst is stable
up to 423 K.
perature = 393 K; catalyst loading: in w/w of acetone; pressure = 30 bar.
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148146
3.7. Effect of catalyst loading
The reactions were performed at different catalyst loadings
(2–12% w/w of acetone) and it was observed that the rate of
reaction increases when the catalyst loading was increased from
2% to 5%w/w of acetone. However, the rate of reaction reaches
saturation above the catalyst loading of 5% w/w of acetone, as
shown in Fig. 8. Moreau et al. [56] have reported such non-liner
influence of catalyst loading on the rate of hydrolysis of acetals
over H-montmorillonite and strong cation exchange resin. It is
expected that the number of catalyst sites above a particular
catalyst loading becomes excessive and hence the rate becomes
insensitive to the further increase in the catalyst loading. We
believe this effect is not because of the diffusional limitations as
variables such as stirrer speed etc. do not influence the overall
rate of reaction under these conditions. In our opinion, such an
effect can be explained only through the rigorous thermo-
dynamic models of adsorption. A theoretical explanation for
the same is lacking in the literature though some evidences have
been reported [56].
3.8. Effect of hydrogen pressure
The solubility of hydrogen varies with pressure and hence
it may influence the reaction kinetics. In order to study the
effect of hydrogen pressure, we performed reactions over a
range of 15–45 bar. Interestingly, it was observed that there is
no effect of hydrogen pressure on the reaction kinetics, as
shown in Fig. 9. This finding again confirms that the catalyst
activity for hydrogenation is sufficient enough and that it is
the formation of mesityl oxide that governs the overall rate of
the reaction.
Fig. 9. Effect of hydrogen pressure on the reaction kinetics. Tempera-
ture = 393 K; catalyst loading 5% w/w of acetone; pressure in bar.
4. Kinetic modeling
O’Keefe et al. [57] proposed a mechanism for hydrogenation
of mesityl oxide to methyl isobutyl ketone. The mechanism
consists of five steps. In the first step, mesityl oxide adsorbs on
two adjacent sites to form a di-adsorbed species. In the second
step, hydrogen adsorbs dissociatively on two adjacent sites. The
adsorbed hydrogen atoms are then added stepwise to the di-
adsorbed mesityl oxide to give methyl isobutyl ketone, which
then desorbs from the catalyst surface to the bulk. The authors
have confirmed this mechanism by pulse adsorption of mesityl
oxide, which was monitored by diffuse reflectance infrared
spectroscopy and they proposed a model based on the
Langmuir–Hinshelwood approach. We assume that a similar
mechanism holds good for the present system. Acetone gets
adsorbed on two adjacent active acid sites and condenses to
give diadsorbed dimer, which then dehydrates to form di-
adsorbed mesityl oxide to give methyl isobutyl ketone in an
adsorbed state in a similar way to that discussed above. As
discussed before, we observed that the reaction is pseudo-zero
order with respect to the hydrogen concentration over the range
of pressure studied. The rate equation may be given by
r ¼ ka2actð1þ kwtrawtr þ kactaactÞ4
(3)
The adsorption term for the methyl isobutyl ketone was omitted
from the model, as O’Keefe et al. [57] found that there is no
effect on reaction kinetics when methyl isobutyl ketone was
added to the reaction mixture at time equal to zero. Also, the
parameter fitting with inclusion of this term gave us very low
values for the adsorption coefficient of methyl isobutyl ketone.
5. Parameter estimation
The mole balance for the batch reactor for all the
components can be written as
dnidt
¼ MCATyir (4)
The temperature dependency of the rate constant can be
expressed by the Arrhenius equation
k ¼ k0 exp
��E�
RT
�(5)
The objective function to be minimized is given as
minF ¼X
All samples
ðxi;calculated � xi;measuredÞ2 (6)
For optimisation, a SQP (Sequential Quadratic Programming)
approach from NAG library was used in the DIVA [58]
simulation environment. The non-idealities of liquid phase
were described using the UNIQUAC method and the thermo-
dynamic parameters were obtained from the data bank of the
commercial package ASPEN. Hydrogen solubility was pre-
dicted by using Henry’s law and interaction parameters to
calculate Henry’s constant were also taken from the commer-
cial package ASPEN. The estimated parameters for the kinetic
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148 147
Table 3
Kinetic parameters for proposed model
k8 (kmol/(g s)) 5.8714 � 107
E8 (kJ/mol) 97.60
Kwtr 4.22
Kact 2.58
f 2.3 � 10�3
Fig. 10. Comparison between modelled and measured conversion of acetone at
different temperatures. Temperature in K; catalyst loading: 5% w/w of acetone;
pressure = 30 bar.
Fig. 11. Comparison between modelled and measured conversion of acetone at
different values of catalyst loadings. Temperature = 393 K; catalyst loading: in
% w/w of acetone; pressure = 30 bar.
Fig. 12. Model predictions for effect of moisture conversion of acetone.
Temperature = 413 K; catalyst loading: 5% w/w of acetone; pressure = 30 bar,
moisture concentration in wt.%.
model are given in Table 3. As shown in Figs. 10 and 11, the
model captures the effect of parameters like temperature and
catalyst loading. It should be noted that the model is applicable
for catalyst loading up to 5% w/w of acetone, beyond which the
rate change is non-liner with catalyst loading (see Fig. 8). The
model was thus used to independently predict the performance
of a batch reactor with initial moisture present. The predicted
results are compared with the experimental ones in Fig. 12. The
model slightly over-predicts the rate in the presence of initial
moisture content. Another approach to capture the effect of
moisture by considering Freundlich adsorption of water [55]
was also applied. However, the fitting was worse by the latter
approach.
6. Conclusions
Kinetics of the one-pot synthesis of methyl isobutyl ketone
from acetone was studied. The effects of various parameters
like temperature and catalyst loading were investigated in
detail. It was observed that catalyst undergoes reversible
deactivation due to the formation of water in the condensation
reaction. As a result, lower rates are observed at higher
conversion levels of acetone, showing pseudo-equilibrium,
which was confirmed by independent reactions. About 45%
conversion of acetone with more than 95% selectivity towards
methyl isobutyl ketone was realised in about 3.5 h at 393 K.
This is promising, considering the performance of different
reported catalysts, like Pd-ZSM-5 [24], Pd–C–Nb2O5 [12,13],
Pd–CuO/MgO/SrO [8], Pd-H-ZSM-5 [37], Cu–MgO [26] and
Pd-CS-H-ZSM-5 [16] with 20–60% conversion of acetone and
30–90% selectivity towards methyl isobutyl ketone under
similar or even more harsh conditions. The pseudo-equilibrium
due to adsorption of water may be considered as a possible
S. Talwalkar, S. Mahajani / Applied Catalysis A: General 302 (2006) 140–148148
limitation of ion exchange resin catalyst. However, the
simultaneous removal of water during the course of reaction
will help to overcome this drawback. A kinetic model that
considers the adsorption of water on the catalyst is proposed
which satisfactorily explains the experimental observations.
References
[1] A. Chakrabarti, M.M. Sharma, React. Polym. 20 (1993) 1–45.
[2] M.M. Sharma, React. Funct. Polym. 26 (1995) 3–23.
[3] S. Veibel, J.I. Nielsen, Tetrahedron 23 (1967) 1723–1733.
[4] L. Barfacker, D. Tom, P. Eilbracht, Tetrahedron Lett. 40 (1999) 4031–
4034.
[5] F. King, G.J. Kelly, Catal. Today 73 (2002) 75–81.
[6] S. Guseva, A. Grigor’eva, T. Prozorovskaya, Neftepererabotka i Nefte-
khimiya 6 (1978) 30–32.
[7] S. Wen, Z. Li, Z. Qian, X. Mao, Jilin Daxue Ziran Kexue Xuebao 1 (1981)
99–108.
[8] Y. Too, T. Nakayama, JP62258335 (1987).
[9] Y. Higashio, K. Hamada, EP87-308156 (1987).
[10] T. Maki, T. Yokoyama, Y. Sumino, JP63096147 (1988).
[11] A.A. Grigor’ev, S.I. Guseva, E.A. Katsman, Khimicheskaya Promyshlen-
nost 7 (1988) 392–394.
[12] T. Maki, T. Yokoyama, Y. Sumino, JP63068538 (1988).
[13] T. Maki, T. Yokoyama, Y. Sumino, JP63068539 (1988).
[14] T. Maki, T. Yokoyama, Y. Sumino, JP63096146 (1988).
[15] Y. Qian, G. Ding, H. Wu, Y. Zheng, J. Jiang, Huaxue Fanying Gongcheng
Yu Gongyi 4 (1988) 89–93.
[16] H. Chin, K. Shu, Y. Chin, N. Cho, B. Hayashi, S. So, JP02069430 (1990).
[17] Y. Too, T. Nakayama, K. Shioda, JP03093743 (1991).
[18] T. Ushikubo, Y. Kakuno, H. Tsujiura, EP458216 (1991).
[19] J. Pasek, J. Hajek, J. Koubek, S. Zicha, P. Pavlas, P. Kralik, A. Kospach, A.
CS273524 (1992).
[20] L. Chu, Z. Yao, CN1069674 (1993).
[21] K. Bian, L. Ding, J. Gao, H. Gu, Ziran Kexue 31 (1995) 63–67.
[22] K. Tao, X. Qiu, Y. Wang, CN1108151 (1995).
[23] S. Narayanan, R. Unnikrishnan, Appl. Catal. A: Gen. 145 (1996) 231–236.
[24] P. Chen, S. Chu, K. Wu, W. Lin, US Patent 5,684,207 (1997).
[25] B. Coh, J. Hur, H. Lee, Chem. Lett. 7 (1998) 583–584.
[26] V. Chikan, M. Arpad, B. Katalin, J. Catal. 184 (1999) 134–143.
[27] L. Melo, G. Giannetto, L. Cardozo, A. Llanos, L. Garcia, P. Magnoux, M.
Guisnet, F. Alvarez, Catal. Lett. 60 (1999) 217–222.
[28] B. Coh, J. Hur, H. Lee, Stud. Surf. Sci. Catal. 121 (1999) 453–456.
[29] S. Yang, Y. Wu, Appl. Catal. A: Gen. 192 (2000) 211–220.
[30] M. Montes, M. Martin, A. Diaz, J. Benitez, J. Odriozola, Mater. Sci.
Forum (2000) 325–326.
[31] D. Bombos, G. Bozga, M. Bombos, A. Stefan, I. Stanciu, Chem. Papers 54
(2000) 171–176.
[32] Y. Chen, B. Liaw, H. Tan, K. Shen, Appl. Catal. A: Gen. 205 (2001) 61–69.
[33] H. Lee, B. Coh, Proceedings of the 221st ACS National Meeting, 2001.
[34] Z. Lu, W. Ai, L. Song, L. Xu, H. Zhu, Faming Zhuanli Shenqing Gongkai
Shuomingshu 9 (2001).
[35] L. Melo, A. Llanos, M. Mediavilla, D. Moronta, J. Mol. Catal. A: Chem.
177 (2002) 281–287.
[36] L. Mattos, F. Noronha, J. Monteiro, J. Catal. 209 (2002) 166–176.
[37] K. Lin, A. Ko, J. Chin. Chem. Soc. 49 (2002) 935–942.
[38] Z. Lu, F. Ai, L. Song, W. Liang, CN1331071 (2002).
[39] Y. Shen, J. Wu, Z. Zhang, Shenyang Huagong Xueyuan Xuebao 16 (2002)
261–265.
[40] D. Bombos, T. Juganaru, A. Stefan, M. Bombos, E. Juganaru, RO117154
(2001).
[41] M. Vandersall, R. Weinand, EP321450 (2003).
[42] R. Morales, L. Melo, J. Brito, A. Llanos, D. Moronta, L. Albornoz, E.
Rodriguez, J. Mol. Catal. A: Chem. 203 (2003) 277–286.
[43] F. Winter, A. van Dillen, K. de Jong, J. Mol. Catal. A: Chem. 219 (2004)
273–281.
[44] X. Sun, D. Lu, X. Bu, Z. Wang, H. Liu, D. Zhou, Jilin Huagong Xueyuan
Xuebao 21 (2004) 22–23.
[45] W. Nicol, E. du Toit, Chem. Eng. Process. 43 (2004) 1539–1545.
[46] F. Ai, Z. Lu, L. Song, H. Wang, S. Wang, Shunnong, CN1488617 (2004).
[47] P. Yang, J. Yu, Z.Wang,M. Xu, Q. Liu, X. Yang, T. Wu, Catal. Commun. 6
(2005) 107–111.
[48] P. Yang, Z. Wang, J. Yu, M. Xu, Q. Liu, S. Jing, T. Wu, Polish J. Chem. 79
(2005) 121–127.
[49] Y. Shen, H. Zhang, J. Wu, Z. Zhang, Huagong Shengchan Yu Jishu 12
(2005) 9–11.
[50] K. Lawson, B. Nkosi, WO9965851 (1999).
[51] N. Saayman, G. Lund, WO0222542 (2002).
[52] N. Saayman, G. Lund, US2003065227 (2003).
[53] K. Lawson, MXPA00011793 (2002).
[54] E. du Toit, R. Schwarzer, W. Nicol, Chem. Eng. Sci. 59 (2004) 5545–5550.
[55] E. du Toit, W. Nicol, Appl. Catal. A: Gen. 277 (2004) 219–225.
[56] C. Moreau, R. Durand, P. Geneste, S. Mseddi, J. Mol. Catal. 112 (1996)
133–141.
[57] W.K. O’Keefe, M. Jiang, F.T.T. Ng, G.L. Rempel, Chem. Eng. Sci. 60
(2005) 4131–4140.
[58] M. Mangold, A. Kienle, E. Gilles, K. Chem. Eng. Sci. 55 (2000) 441.
[59] http://www.rohmhaas.com/ionexchange/IP/ch28_typical.htm, seen on 20
August 2005.