Ru-C-ZnO composite catalysts for the synthesis of methyl isobutyl ketone via single step gas phase acetone self-condensation

Article abstract of DOI:10.1007/s10562-014-1253-1

Ruthenium/activated charcoal (Ru-C) was modified by a solid-solid interaction method with synthesized nano-zinc oxide (n-ZnO). Three different ratios of Ru-C:n-ZnO (1:2, 1:1 and 3:2) were used to prepare Ru-C-ZnO composite catalysts. These were used in the gas-phase, one-step self-condensation of acetone to methyl isobutyl ketone (MIBK). The composite catalyst (1:1 ratio) contained 2.5 wt% Ru showed superior conversion of acetone and selectivity for MIBK. Furthermore, this catalyst showed good consistency for MIBK formation for 100 h without any deactivation. Characterization of the catalysts revealed that balanced hydrogenation and acid-base functional character is crucial to obtain high catalytic performance.

Full text of DOI:10.1007/s10562-014-1253-1

Catal Lett
DOI 10.1007/s10562-014-1253-1
Ru–C–ZnO Composite Catalysts for the Synthesis of Methyl
Isobutyl Ketone via Single Step Gas Phase Acetone
Self-Condensation
•
Abdulaziz A. Bagabas Mohamed Mokhtar
•
•
Vagif M. Akhmedov Katabathini Narasimharao
•
•
Sulaiman N. Basahel Abdulrahman Al-Rabiah
Received: 16 January 2014 / Accepted: 29 March 2014
Ó Springer Science+Business Media New York 2014
Abstract Ruthenium/activated charcoal (Ru–C) was
modified by a solid–solid interaction method with synthe-
sized nano-zinc oxide (n-ZnO). Three different ratios of
Ru–C:n-ZnO (1:2, 1:1 and 3:2) were used to prepare Ru–
C–ZnO composite catalysts. These were used in the gas-
phase, one-step self-condensation of acetone to methyl
isobutyl ketone (MIBK). The composite catalyst (1:1 ratio)
contained 2.5 wt% Ru showed superior conversion of
acetone and selectivity for MIBK. Furthermore, this cata-
lyst showed good consistency for MIBK formation for
100 h without any deactivation. Characterization of the
catalysts revealed that balanced hydrogenation and acid–
base functional character is crucial to obtain high catalytic
performance.
1 Introduction
The conversion of inexpensive organic compounds, such
as acetone, to high-value chemicals is highly relevant for
industrial applications [1, 2]. One example of such a
chemical is methyl isobutyl ketone (MIBK), which is
widely used in a number of applications including: in
paints, as a liquid–liquid extraction solvent, as a solvent
for vinyl, epoxy and acrylic resin production, for the
removal of paraffin from mineral oils, in the synthesis of
rubber chemicals, and in fine chemistry applications [3,
4]. The global demand for MIBK is estimated to be
300,000 tons per year. Historically, MIBK was produced
via a three-step, homogeneous catalytic process from
acetone [5] (Scheme 1). The first step is the self-con-
densation of acetone to diacetone alcohol (DA), the
second step is the dehydration of DA to mesityl oxide
(MO) and the final step is the selective hydrogenation of
the carbon–carbon double bond of MO to form MIBK.
This process thus involves several unit operations,
resulting in high capital and operating costs and inherent
inefficiencies. Some unwanted side reactions also occur;
the most commonly observed are over-condensation and
non-selective hydrogenations. Consequently, it is desir-
able to produce MIBK from acetone in a single stage
and a number of catalytic systems have been proposed
that aim to reduce the number of steps and side
reactions.
Keywords Ruthenium-activated charcoal ꢀ Nano-ZnO ꢀ
Acidic/basic site concentration ratio ꢀ Ru-loading ꢀ Acetone
self-condensation ꢀ Methyl isobutyl ketone
A. A. Bagabas (&)
Petrochemicals and Refining Center, Petrochemicals Research
Institute, King Abdulaziz City for Science and Technology,
P.O. Box 6086, Riyadh 11442, Kingdom of Saudi Arabia
e-mail: abagabas@hotmail.com
M. Mokhtar ꢀ K. Narasimharao ꢀ S. N. Basahel
Department of Chemistry, Faculty of Science, King Abdulaziz
University, P.O. Box 80203, Jeddah 21589, Kingdom of Saudi
Arabia
A significant study has investigated the use of sup-
ported transition metals to catalyze the condensation of
acetone (dimethylketone, DMK) to a,b-unsaturated
ketone, followed by another in situ catalytic reduction
step. It has been reported that MIBK can be produced
from acetone and hydrogen in one step in a catalytic
distillation reactor [6].
V. M. Akhmedov
Department of Chemistry, Baku State University, Zahid Xalilov
St. 23, Baku, Azerbaijan Republic
A. Al-Rabiah
Department of Chemical Engineering, College of Engineering,
King Saud University, Riyadh 11451, Kingdom of Saudi Arabia
123

A. A. Bagabas et al.
OH
O
O
O
-H2O
+acetone
-acetone
+H2
O
+H2O
MO
DA
MIBK
acetone
+
a
1
c
,
e
6
-
t
o
n
-
H
2
a
l
e
d
O
o
-
l
H
i
z
OH
O
OH
O
a
2
O
t
i
o
n
-2H2O
PH
triacetone dialcohol
Mesitylene (M)
O
O
diisobutyl ketone (DIBK)
OH
OH
IPH
Isopropanol
methyl isobutyl carbinol (MIBC)
OH
+H2
diisobutyl carbinol (DIBC)
propene
propane
Scheme 1 Reaction pathway of conventional process for the production of MIBK
Bi-functional catalysts, such as Pd- or Pt-supported on
sulfonated resins, which have condensation, dehydration
and hydrogenation capabilities, are used for this reaction
[7], which is undertaken under high pressures and thus adds
significant cost to the overall procedure. The MIBK also
needs to be extracted from the effluent in the reactor and
purified, as its concentration is usually lower than 30 wt%.
It is, therefore, of great interest to develop novel catalysts,
which can offer comparable or better MIBK yields at lower
temperatures and under atmospheric pressure. Ideally, the
catalyst should be versatile so that it can be used in the
production of a variety of aliphatic low molecular weight
ketones.
in the gas-phase at ambient pressure. A Pd/CuO/Al2O3/
SiO2 catalyst offered 30 % acetone conversion with
60 % MIBK selectivity at 230 °C under atmospheric
pressure. Several studies have also been focused on the
application of different supports with acid–base surface
functionality; metal catalysts supported on activated
carbon (C) provided high activity and selectivity for
MIBK from acetone [17]. Recently, we reported the
addition of n-ZnO to Pd metal catalysts and how
the composition of metal and n-ZnO is influencing the
activity and selectivity of acetone condensation reaction
[18]. It was determined that 1.0 wt% Pd supported on
n-ZnO and 1.0 wt% Pd supported on a n-ZnO and CrO3
composite material were highly active and selective
catalysts for a one-step, gas-phase synthesis of MIBK
from acetone at atmospheric pressure. To achieve better
MIBK selectivity in the self-condensation of acetone, it
seems that the catalyst should possess basic sites of
medium strength along with hydrogenation activity [9].
It is known that Ru–C is an effective hydrogenation
catalyst [19] and ZnO possesses both acidic and basic
sites [20]. To develop a catalyst which possesses
hydrogenation and acid–base properties, we chose to
prepare a composite of Ru–C and n-ZnO.
Talwalkar and Mahajani [8] summarized the conver-
sion rates and selectivities of the most promising cata-
lysts reported to date for MIBK synthesis from acetone.
Group VIII or IB metals, such as Pd, Pt, Ni and Cu, on
acidic or basic supports such as zeolites, mixed transition
metal oxides and ion exchange resins, have been exten-
sively investigated. Most notably, Pd [9–11] and Ni [12,
13] supported on Mg–Al or Zn–Cr mixed oxides, alu-
mina, H-ZSM-5, Cu/MgO [14], Cu–Mg alloy powders
[15] and Pd–Zr–P multifunctional Pd-polyoxometalate
catalytic system [16] are reported to catalyze this process
123

Ru–C–ZnO for the Synthesis of Methyl Isobutyl Ketone
In this study, we have investigated the one-step gas-
phase self-condensation process of acetone using n-ZnO as
the substrate, which possesses both acid and base proper-
ties. Our working hypothesis is that activated charcoal,
with its characteristically high surface area, can enhance
the dispersion of ZnO nanoparticles. Moreover, it is
anticipated that the hydrophobicity of this charcoal can
hinder the interaction between the active sites and water
molecules. The results reported here indicate that the
combination of Ru, ZnO, and activated charcoal (Ru–C–
ZnO) creates a unique catalyst system for the gas-phase
condensation reaction. To investigate the interactions of
the three components in the Ru–C–ZnO catalysts, several
characterization techniques (XRD, TEM, N2-adsorption,
chemisorption and TPD/TPR) were utilized. We also
studied the effect of reaction temperature, concentration of
acid–base sites and Ru-loading on catalytic performance.
To best of our knowledge, the application of Ru–C–ZnO
composite catalysts for the aforementioned reaction has not
previously been reported.
1:1 wt% (CAT-III) and 3:2 wt% (CAT-IV). For compari-
son, the pure n-ZnO (CAT-I) and pure Ru–C (CAT-V)
catalysts were also studied.
2.3 Characterization of Catalysts
The phase identification for the calcined catalysts were
performed using a Philips ‘X’ pert pro’ diffractometer,
operated at 40 kV and 40 mA, using CuKa radiation, in the
2h range from 2° to 100° in steps of 0.02°, with a sampling
time of 1 s/step. The average crystallite size was calculated
using Scherrer’s equation:
Dhkl ¼ Kk=ðBhkl cos hÞ
where Dhkl is the average crystallite size of the phase under
ð1Þ
investigation, K is the Scherrer’s constant (0.89), k is the
˚
wavelength of the X-ray beam used (1.54056 A), Bhkl is the
full width at half maximum (FWHM) of the analyzed
diffraction peak (101) and h is the diffraction angle.
A Philips CM200FEG microscope, 200 kV, equipped
with a field emission gun was used for transmission elec-
tron microscopy (TEM) analysis. The coefficient of
spherical aberration was Cs = 1.35 mm. The information
limit was better than 0.18 nm. High-resolution images with
a pixel size of 0.044 nm were taken with a CCD camera.
The BET-surface area of the prepared catalysts was
estimated from N2 adsorption at -196 °C using a
Quantachrome Autosorb-1-MP instrument. The acidic and
basic properties of each catalyst was determined by the
temperature-programmed desorption (TPD) technique with
ammonia (NH3) and carbon dioxide (CO2) respectively,
using a TPD/R/O 1100 thermo electron instrument. Typi-
cally, calcined sample of 50 mg was first pre-treated in
helium gas at 150 °C for 2 h, and then cooled to 30 °C.
Subsequently the sample was saturated with NH3 or CO2 at
this temperature for 1 h. The initial flushing with helium
gas was done in order to remove any excess NH3 or CO2.
The TPD profiles were recorded with a thermal conduc-
tivity detector at a heating rate of 283 K min-1 from room
temperature to 500 °C under a continuous flow of helium
gas. Quantachrome CHEMBET 3000 was used to perform
the TPR analysis of catalysts. The experiments were car-
ried out using a gas mixture of 5 % H2 in Ar at a flow rate
2 Experimental
2.1 Chemicals
Acetone (Winlab HPLC grade, 99.8 %), hydrogen gas
(99.9 %), sodium pyruvate (Sigma-Aldrich, C99 %),
hydroxylamine hydrochloride (Carlo-Erba, 99 %), sodium
hydrogen carbonate (Fluka, C99.5 %), zinc sulfate hepta-
hydrate (BDH, 99.5 %), and ruthenium 5 wt% on activated
charcoal catalyst (Ru–C, Winlab) were purchased and were
used without further purification.
2.2 Preparation of Catalysts
n-ZnO was prepared following the procedure reported by
Bagabas et al. [21]. An appropriate amount of aqueous
solution of ZnSO4ꢀ7H2O was combined with an aqueous
solution of sodium pyruvic acid oxime monohydrate
[Na(PAO)H2O], which was prepared from sodium pyru-
vate, hydroxylamine hydrochloride, and sodium hydrogen
carbonate in aqueous medium. The white precipitate of
Zn(PAO)2(H2O)2 was filtered off, then dried under vacuum
overnight, and subsequently calcined for approximately
12 h at 400 °C in a muffle furnace to obtain pure n-ZnO.
A solid–solid interaction method was adopted to syn-
thesize the composite of Ru–C–ZnO catalysts. Ru/AC and
n-ZnO in different ratios were mixed thoroughly using a
pestle and mortar then the mixture was pulverized at room
temperature using a pulverizing machine and subsequently
the composites were calcined at 400 °C for 12 h. The ratios
of Ru–C:n-ZnO used were as follows: 1:2 wt% (CAT-II),
of 30 mL min-1 with
a linear heating ramp of
278 K min-1 up to 600 °C. The amount of H2 consumed
was monitored as a function of temperature on a thermal
conductivity detector (TCD).
The metal surface area and dispersion were determined
by H2 pulse chemisorption experiments using a micro-
metrics AutoChem 2910 instrument. Approximately 50 mg
of catalyst was treated at 150 °C for 16 h under vacuum
prior to starting the experiment. The catalyst was loaded
into the sample tube and the sample heated to 400 °C at the
123

A. A. Bagabas et al.
rate of 10 °C min-1 under a mixed H2/Ar (5/95 vol%) gas
flow of 10 mL min-1. The sample was kept under these
conditions for 1 h, and then it was flushed for 1 h by
switching the flow to helium gas. After completion of the
reduction process, the temperature of the sample was
decreased to 70 °C and pulses of H2 were initiated to
determine the volume of H2 uptake. The metal surface area
and dispersion were determined from the hydrogen con-
sumption by the sample.
CAT-V
CAT-I
500
400
300
200
100
0
CAT-II
CAT-III
CAT-IV
2.4 Catalytic Activity Measurements
All catalysts were activated by reducing the catalyst at
400 °C under a H2 gas flow of 60 mL min-1 over night,
prior to catalytic testing. The self-condensation of acetone
was performed at reaction temperatures of 250, 300, 350
and 375 °C respectively, under hydrogen flow, at atmo-
spheric pressure, in a glass tubular fixed-bed micro reactor
(ID 0.9 cm). The reactor was placed in a vertical tubular
furnace and the catalyst sample (*250 mg) was loaded
into the reactor between two layers of pyrex glass wool and
pieces of ceramic. The H2/acetone molar ratio was main-
tained at 1.5, 3, 4.5 or 6. Acetone was introduced to the
reactor by a Hitachi HPLC pump (Model L-2100/2130),
connected to a heated tube set to 65 °C for its delivery as a
vapor mixed with H2 before entering the reaction zone at
the top of the catalyst bed. A K-type thermocouple was
placed in the center of the catalyst to monitor the reaction
temperature. The liquid products were condensed at
-40 °C and collected every hour, and subsequently ana-
lyzed off-line using a Varian CP-3800 gas chromatograph,
equipped with a flame ionization detector (FID) and a
50 9 0.32 9 1.2 mm3 CP-Wax 58 CB column. The gas-
eous downstream flow was analyzed online by a refinery
gas analyzer Varian CP-3800 GC, equipped with an FID
and TCD. The products were identified by GC using
standard samples.
10
20
30
40
50
60
70
80
2Θ (degrees)
Fig. 1 XRD patterns of all the investigated catalysts
Table 1. In this instance, CAT-I was found to have an
average crystallite size of 20 nm. The preparation proce-
dure is known to have great influence on the phase and
particle size of ZnO. The typical particle size of nano-sized
ZnO varies from 10 to 40 nm. There is a little difference in
crystallite size of CAT-II (23 nm) and CAT-I (20 nm).
However, a more pronounced change was observed by
further increasing the wt% of Ru–C; CAT-III and CAT-IV
showed an average crystallite size of 25 and 30 nm
respectively.
The crystallite size of Ru metal in Ru–C (CAT-V)
sample was not determined because of its amorphous
nature. Previously, Navale et al. [22] reported structural
parameters of crystalline Ru-doped ZnO materials. They
observed an increase in the average crystallite size of the
materials with the increase in Ru content and pretreatment
temperature. An increase of the crystallite size from 16 to
36 nm was observed when the Ru content changed from
1.0 to 3.0 wt%. During the calcination step, there is a
clear possibility for oxidation of Ru to ruthenium oxide,
however diffraction peaks corresponding to ruthenium
oxide were not observed for any of the composite mate-
rials probably due to very fine dispersion of those
particles.
3 Results and Discussion
3.1 Powder X-ray Diffraction
The crystalline phases, as well as the average crystallite
size of the catalysts, were determined from the powder
XRD patterns shown in Fig. 1. The CAT-V sample was
shown to have an amorphous nature, whereas CAT-I was
highly crystalline as depicted by the diffraction peaks
corresponding to the wurtzite phase of ZnO [JCPDS
36-1451]. It is clear that wurtzite is the dominant phase in
all the composite catalysts, similar to the pure n-ZnO
sample. The average crystallite size of each of the catalysts
was determined using Eq. 1 and the values are shown in
3.2 TEM Analysis
TEM analysis was performed (Fig. 2) to obtain more
detailed information about the average crystallite size and
structure of the catalysts. The TEM image for pure n-ZnO
(CAT-I) powder shows hexagonal ZnO crystallites of
approximately 3–6 nm in diameter. Whereas, the pure
123

Ru–C–ZnO for the Synthesis of Methyl Isobutyl Ketone
b
Table 1 Results from XRD and
N2 adsorption measurements of
all the investigated catalysts
Catalyst
Major
phasea
Crystallite
size (nm)a
Ru
loading
(wt%)
SBET
(m2 g-1
SRu
(m2 g-1
Ru
dispersionb
(%)
)
)
CAT-I
wurtzite
wurtzite
wurtzite
wurtzite
amorphous
20
23
25
30
–
0.0
1.67
2.5
3.0
5.0
37
300
289
278
327
–
–
a
Powder XRD measurements
CAT-II
CAT-III
CAT-IV
CAT-V
24.36
30.70
25.73
22.43
5.00
6.31
5.29
4.61
b
H2-chemisorption
measurements; catalysts were
treated at 150 °C for 16 h under
vacuum before starting
measurements
Fig. 2 TEM images of the representative catalysts
3.3 N2 Adsorption–Desorption
Ru–C sample (CAT-V) showed a sponge-like amorphous
structure and black spots, which are observed on the sur-
face of activated charcoal, indicate dispersed Ru nano
particles. A TEM image of a selected composite catalyst
(CAT-III) can also be seen in the Fig. 2c. It can be con-
cluded from the image analysis of CAT-III that the
ZnO crystallite size is not uniform. Furthermore, the
amorphous nature of carbon has not changed after n-ZnO
incorporation.
BET-surface areas (SBET), calculated from the nitrogen
adsorption/desorption isotherms of composite catalysts, are
given in Table 1. CAT-I was determined to have a surface
area of 37 m2 g-1, which is large for a highly crystalline
oxide such as ZnO. The micro-crystalline ZnO shows a
surface area in the range of 5–10 m2 g-1 [23]. Nanometer-
sized particles of ZnO have very different physical and
chemical properties compared with the bulk material.
When used as catalysts, their catalytic activity is expected
to be enhanced not only because of their increased surface
area, but also because of the changes in surface properties,
such as surface defects. The CAT-V sample was found to
Navale et al. [22] calculated the average crystallite size
using Scherrer’s equation and the size is in the range from
20 to 130 nm for Ru–ZnO samples. In our work, we have
not observed such a large change in the crystallite size of
the calcined composite samples. This difference is thought
to be due to the different method of preparation used and
that the composite catalysts consist of amorphous C as one
of the components.
have a high surface area of 327 m2 g-1
.
It is known that the textural properties of commercially
available Ru–C samples depend upon the manufacturing
123

A. A. Bagabas et al.
process, as well as the nature of the carbon used [15]. The
major contribution to the surface area of the Ru–C catalyst
results from the surface structure of activated charcoal. The
surface areas of the composite catalysts are lower than that
of the pure Ru–C, therefore, mixing of Ru–C and n-ZnO
leads to a decrease in surface area. We compared the sur-
face area of the composite catalysts with theoretical Ru
wt% loading (Table 1) and found that the surface area
decreases with increasing Ru metal loading. Other com-
posite catalysts showed similar behavior, where a decrease
of surface area is observed upon increasing Ru loading.
Furthermore, as described in the literature [24], it is pos-
sible that agglomerated particles of Ru–ZnO could partially
block the pores of active carbon leading to a decrease of its
available surface for N2 adsorption.
500
400
300
200
100
0
A
CAT-III
CAT-IV
CAT-II
CAT-V
CAT-I
50
100
150
200
250
300
350
400
450
500
Temperature (oC)
500
400
300
200
100
0
B
3.4 H2 Chemisorption
The Ru-metallic surface area, as well as its degree of dis-
persion over the calcined composite catalysts, was deter-
mined using H2 pulse chemisorption at 70 °C (Table 1).
Bachiller-Baeza et al. [25]. studied the reducibility of
mono- and bi-metallic Ru–C catalysts. They observed a
two-stage reduction process for both type of catalyst; the
first reduction was at a low temperature (*225 °C) and
the other at higher temperature (*400 °C). For this reason,
we reduced the catalysts at 400 °C to achieve the complete
reduction of Ru and ZnO present in the composite mate-
rials. The trend of both metallic surface area and the degree
of dispersion of Ru metal in all the composite catalysts was
similar. An increased Ru loading resulted in an increase of
metallic surface area and degree of dispersion, except in
the case of the CAT-III sample, for which the apparent
maximum values of both metallic surface area and degree
of dispersion were obtained. This is thought to be due to an
optimum loading effect of n-ZnO, which interacts with the
Ru metal. The CAT-I sample adsorbed the H2, but the
volume of H2 uptake is relatively low compared to any of
the other composite catalysts. The partial reduction of ZnO
at low temperature has been reported in the literature for
the bimetallic supported system, Ru–Zn/SiO2 [26], and the
H2 consumption peaks were assigned to H2 chemisorption
on ZnO (ZnO ? H2 ? ZnH–OH). Neri et al. [27] rejected
this pathway, and explained that the peak at low temper-
ature could be due to a spill-over mechanism by which
hydrogen, dissociatively adsorbed on metallic Ru, migrates
to the ZnO support and reduces it, yielding metallic Zn and
water.
CAT-III
CAT-IV
CAT-II
CAT-I
CAT-V
50
100
150
200
250
300
350
400
450
500
Temperature (oC)
Fig. 3 TPD patterns of all the investigated catalysts a NH3 and
b CO2
metal oxides has been described as being affected by the
presence of a noble metal and in general the reduction
temperature of the oxide becomes lower [25–27]. It has
also been reported that heating, in the presence of H2 gas,
induces the formation of alloy phases (thermodynamically
favorable) between metallic Zn and active noble metals
such as Ru or Pt even at low temperatures [28, 29].
3.5 Acid–Base Properties
The acid–base properties of the calcined catalysts were
measured by NH3 and CO2-TPD experiments. Figure 3a
shows the NH3-TPD profiles of the n-ZnO and composite
materials. For the CAT-I sample, a single desorption peak
centered at 95 °C can be attributed to the physically
adsorbed NH3 on the surface of n-ZnO. In contrast, CAT-V
(Ru–C) sample showed two desorption peaks at 100 °C and
290 °C. These two peaks can be attributed to physically
adsorbed NH3 and moderate acid sites, respectively. The
CAT-II (33.3 wt% of Ru–C) sample showed a similar
desorption pattern as CAT-V. An additional desorption
CAT-IV chemisorbed a higher amount of hydrogen than
Ru–C, which clearly indicates that there are more Ru and
Zn surface atoms accessible to hydrogen in this composite
catalyst. This could be explained by the larger interactive
particle size, detected by XRD. The reducibility of surface
123

Ru–C–ZnO for the Synthesis of Methyl Isobutyl Ketone
peak at 340 °C can be observed for CAT-III (50 wt% Ru–
C) and CAT-IV (60 wt% Ru–C). It is worth noting that the
intensity of the additional peak is greater in the case of
CAT-III than for CAT-IV. It is apparent that the moderate
acid sites on the composite samples increases with the
increase of Ru–C wt% and the appearance of the new peak
indicates the development of relatively strong acid sites in
the CAT-III and CAT-IV composite samples.
temperature desorption peaks due to presence of finely
dispersed ruthenium oxide species.
The distribution of the acidic and basic sites for each of
the samples is quite different. The quantities of NH3 and
CO2 desorbed per gram of catalyst and acid/base site
concentration values are summarized in Table 2. The
amounts of CO2 desorbed were estimated by integrating
areas under the peaks centered at 280 and 350 °C. The
concentration of the acidic and basic sites is measured from
the amount of NH3 and CO2 desorbed, respectively. The
composite catalyst (CAT-III) has the largest concentration
of acidic and basic sites when compared to the other
composite catalysts; and this may be due to an optimum
ratio of n-ZnO and Ru–C composition in the sample.
However, the acidic/basic site concentration is the highest
for the CAT-V sample, which is thought to be due to the
acidic functional groups present on the surface of activated
charcoal. The addition of n-ZnO to Ru–C resulted in an
increase in the basic site concentration of the catalysts.
The CO2-TPD patterns of the materials are presented in
Fig. 3b. A small desorption peak centered at 110 °C can be
observed for the CAT-V sample, which can be attributed to
physically adsorbed CO2. In other hand, CAT-I (n-ZnO)
showed a large desorption peak centered at 290 °C along
with the small peak at 110 °C. The peak at 290 °C can be
attributed to weak basic sites. All the composite samples
showed similar CO2 desorption patterns as n-ZnO. The
peak shift to lower energy in CO2-TPD patterns of com-
posite catalysts indicates that these catalysts possessed
relatively weak basic sites than the pure ZnO. An addi-
tional desorption peak at 350 °C appeared in all the com-
posite samples and the intensity of this peak is in the order
of CAT-III [ CAT-IV [ CAT-II. It can be seen that the
strength of the base sites of the composite samples is not
dependent on either Ru–C or n-ZnO content. The prepared
composite samples show a similar CO2-TPD pattern as n-
ZnO, but a larger proportion of moderate and strong base
sites are found in the prepared composite samples (in
particular CAT-III and CAT-IV). Relatively stronger acid
and base sites are observed in CAT-III (50 % Ru–C and
50 % n-ZnO), which indicates that an optimum ratio of
oxides may be needed to generate the acid and base sites.
Li et al. [30] reported that Ru–ZnO catalysts showed
NH3-desorption peaks in temperature ranges of
150–240 °C (specifically at 170 °C) and 250–425 °C
(specifically at 400 °C). The authors also observed that
with more Ru content, the position of high temperature
peak shifts to lower temperature. This is an indication of
lowering of strength of surface acid sites with high Ru
content. It was also observed that when the catalyst con-
tained 5 % of Ru, the intensity of NH3 desorption peak at
low temperature increased significantly. The authors also
claimed that the extra surface acidity of the catalysts could
be due to the generation of new Ru species, which are not
reduced to the metal state Ru completely, i.e. Run? species
in the catalyst. In addition, Yasumoto [31] reported that
there is a wide temperature range 100–350 °C for chemi-
sorbed NH3 and CO2; the temperature shifts to a lower
range, corresponding to the ‘qa’ values of 80 and 60 kJ/mol
for NH3 and CO2 respectively. The observation made by
Yasumoto clearly indicating that NH3 and CO2 can be
adsorbed on ZnO surface relatively at low temperature. In
our opinion, a similar phenomenon is happening in case of
calcined Ru–C–ZnO composites and showing low
3.6 Self-Condensation of Acetone
It is known that some side reactions during the synthesis of
MIBK from acetone do occur (Scheme 1). Besides the
aldol condensation of acetone to form DA, acetone can be
also hydrogenated over metallic sites to form isopropyl
alcohol (IPA). The selectivity towards IPA decreases
enormously with increasing temperature because of the
exothermic nature of the hydrogenation process of acetone
to IPA (DH = -54.4 kJ/mol) [20]. The conversion of
acetone and selectivity for MIBK over all the catalysts at
different reaction temperatures are shown in Fig. 4a, b
respectively. As the reaction temperature was increased
from 250 to 375 °C, a decrease in the conversion of ace-
tone was observed for all the catalysts. All the composite
catalysts showed better conversion than n-ZnO or Ru–C
alone. The Ru–C catalyst produced 20 % conversion, while
n-ZnO only showed 10 %. The composite CAT-III showed
the maximum conversion of 47 % at 250 °C. The selec-
tivity for MIBK followed a different trend; an increase in
the temperature from 250 to 350 °C also resulted in an
enhanced selectivity for MIBK but increasing the temper-
ature further to 375 °C caused a decrease in selectivity. We
repeated the first set points of catalytic test series at the end
of each run and observed a decrease of acetone conversion
by 5–8 %. The decrease of catalytic activity is probably
due to formation of the water in the reaction of acetone
condensation, which negatively influences the metallic
centers catalytic activity. These results were in accordance
with the results reported in the literature [8].
It is interesting to note that all the catalysts offered the
high selectivity to MIBK at 350 °C. The n-ZnO again
offered less selectivity to MIBK (highest 13 %) than Ru–C
123

A. A. Bagabas et al.
Catalysta CO2 ads.
(mmol g-1 b
NH3 ads.
(mmol g-1 b
Acidic/
basic site of acetone
Conversion Selectivity (%)
Table 2 TPD of NH3, CO2 and
catalytic activity results of all
the investigated catalysts
)
)
MIBK IPA MO DIBK Others
ratio
(%)
CAT-I
0.37
0.40
0.03
0.23
0.46
0.33
0.21
0.08
0.57
1.02
0.80
5.25
5
9
8
60
70
54
15
22
10
3
45
3
8
12
15
18
10
17
15
10
12
10
CAT-II
a
Calcined at 400 °C for 12 h in
CAT-III 0.45
CAT-IV 0.41
30
24
9
2
air
b
13
59
3
Catalysts pre-treated in
helium gas at 150 °C for 2 h
CAT-V
0.04
6
oxides and again CAT-III offered the maximum MIBK
selectivity of 60 % at 350 °C. Bombos et al. [32] studied
the influence of ZnO incorporation to the Pd/H-ZSM-5
catalyst for the synthesis of MIBK. The authors reported
that the addition of ZnO caused a decrease in the activity
and selectivity for MIBK and concluded that the decrease
of acidic centers of the total catalyst was the reason for
such behavior.
50
45
40
35
30
25
20
15
10
5
A
CAT-I
CAT-II
CAT-III
CAT-IV
CAT-V
The TPD results also indicate that Ru–C catalyst have
more acidic sites than the n-ZnO. The high acidity and
activity of Ru–C compared to n-ZnO could be due to the
presence of surface acid functional groups on the activated
carbon [17]. The superior conversion and selectivity of
composite catalysts could be linked to synergistic effect
between Ru–ZnO and enhancement in acidic and basic
functionality of these catalysts. The maximum conversion
was attained at 250 °C, whilst the maximum selectivity for
MIBK was reached at 350 °C. The selectivity towards IPA
decreases upon increasing the reaction temperature for all
the catalysts. The condensation and dehydrogenation pro-
ducts catalyzed by acidic sites are favored above 250 °C,
while direct hydrogenation products are favored below this
temperature (data not presented). It was reported in the
literature that metal oxide based catalysts promote the self
condensation of acetone by both acid- and base- catalyzed
mechanisms [8–10], however deactivation rate would be
closely related to the surface basic properties. It was also
observed that activity declines due to blockage of both acid
and base active sites by a carbonaceous residue formed by
secondary reactions. The observations from the literature
and our experimental results indicating that balanced acid–
base sites, low crystallite size (resistant to coke deposition)
are the responsible factors for the observed catalytic
activity.
250
275
300
325
350
375
Reaction temperature (oC)
70
60
50
40
30
20
10
CAT-I
B
CAT-II
CAT-III
CAT-IV
CAT-V
250
275
300
325
350
375
Reaction temperature (oC)
Fig. 4 Catalytic activity of all the investigated catalysts at different
reaction temperatures a conversion of acetone, b selectivity to MIBK;
catalyst amount: 250 mg, H2/acetone molar ratio: 4.5, GHSV:
3,600 h-1
The effect of variation of H2/acetone molar ratios on
acetone conversions and MIBK selectivity at 350 °C are
shown in Fig. 5a and b respectively. The conversion
increased with increasing H2/acetone up to a molar ratio
equal to 4.5 for all the catalysts. A decrease of conversion
of acetone was observed with an increase of H2/acetone
molar ratio from 4.5 to 6. The selectivity towards MIBK
formation followed a very similar trend as conversion.
(highest 28 %) at all the reaction temperatures. Di-isobutyl
ketone (DIBK), isopropyl alcohol (IPA) and mesityloxide
(MO) are the major products found alongside MIBK. We
also found small amounts of DA, mesitylene, propane and
propene when all the catalysts were used. The composite
catalysts offered better selectivity to MIBK than the pure
123

Ru–C–ZnO for the Synthesis of Methyl Isobutyl Ketone
50
1.2
1.0
0.8
0.6
0.4
0.2
Con. of acetone
Sel. to MIBK
70
60
50
40
30
20
10
CAT-I
CAT-II
CAT-III
CAT-IV
A
45
40
CAT-V
35
30
25
20
15
10
5
CAT-I
CAT-II
CAT-III
CAT-IV
CAT-V
Catalyst
1.5
3.0
4.5
6.0
H2/acetone molar ratio
Fig. 6 Correlation between acid/base site ratio and catalytic perfor-
mance of the catalysts
CAT-I
70
60
50
40
30
20
10
B
CAT-II
CAT-III
CAT-IV
CAT-V
direct hydrogenation of the acetone carbonyl group. The
multifunctional composite catalyst with Ru/C:n-ZnO ratio of
1:1 (wt/wt) (CAT-III) exhibited the highest MIBK selectivity.
Many literature reports suggest that the formation of MIBK
requires the cooperation between metallic and acidic or basic
sites, but the hydrogenation of acetone only needs metallic
sites [10]. Figure 6 shows the comparison between CAT-I,
CAT-II, CAT-III, CAT-IV and CAT-V, and the effect of their
acidic/basic site ratio on the acetone conversion and MIBK
selectivity at optimum reaction conditions. A direct correla-
tion can be observed for all the catalysts except CAT-V. The
activity and selectivity increases steadily with increasing
acidic/basic site ratio. CAT-III showed the highest acetone
conversion and MIBK selectivity. The high acidic/basic site
ratio for CAT-V is due to the acidic functional groups pre-
sented on the surface of the AC, but just acidity alone does not
play a vital role in the acetone self-condensation reaction.
These results showed that the higher activity, the more con-
densation products were obtained with catalysts that have a
balance of acid/base sites.
1.5
3.0
4.5
6.0
H2/acetone molar ratio
Fig. 5 Effect of H2/acetone molar ratio on the catalytic performance
of the catalysts. a Conversion of acetone, b selectivity to MIBK.
Catalyst amount: 250 mg, reaction temperature: 350 °C, GHSV:
3,600 h-1
CAT-III showed the highest selectivity towards MIBK at a
H2/acetone molar ratio of 4.5. Kozhevnikova et al. [10]
observed similar behavior in the case of Pd supported on
Zn–Cr mixed oxide catalysts. Thus, a reaction temperature
at 350 °C and a H2/acetone molar ratio of 4.5 were selected
as the optimum conditions for MIBK formation selectivity
over the CAT-III catalyst.
It is reported that any catalyst that contains a metal–
metal oxide interface, where electron-deficient sites such as
oxygen vacancies exist in close proximity to metallic
hydrogenating sites, can perform better in the self-con-
densation reaction of acetone [11]. This metal–metal oxide
interface can be created either by a partially reduced metal
oxide or as a consequence of a metal–support interaction
like the formation of a solid solution. It is also known that
when an n-type semiconductor particle, such as ZnO, is
exposed to air, oxygen molecules can be adsorbed onto the
surface of the particle and form O2-, O22-and O2- ions by
capturing electrons from the conduction band, which in
turn produces an electron-depleted space charge layer in
the surface region of the particle [33]. Even though oxygen
is not presented in the reactants, ZnO could still be exposed
to oxygen due to in situ formation of H2O in the
The conversion of acetone and selectivity for MIBK as a
function of Ru-loading is presented in Table 2. Both the
conversion of acetone and MIBK selectivity increased with
increasing the Ru loading up to 2.5, further increase of Ru
loading caused a decrease in conversion and selectivity. The
pronounced increase in MIBK selectivity is attributed to the
presence of balanced acidic/basic bi-functional sites (hydro-
genation and condensation) and optimum crystallite size.
IncreasingRuloadingabove 2.5wt% resultedinanincreasein
IPA selectivity, indicating that the high Ru loading favors the
123

A. A. Bagabas et al.
40
35
30
25
20
15
10
70
65
60
55
50
45
40
selectivity to MIBK. Time-on-stream analysis also
showed that this catalyst offers good durability for the
gas-phase, one-step synthesis of MIBK. TEM analysis of
the composite catalysts provided the information that a
synergistic effect existed between Ru–ZnO and TPD
results indicated the enhancement in acidic and basic
centers in these catalysts. Superior performance of the
CAT-III composite catalyst could be directly linked to its
balanced hydrogenation and acid–base functional
character.
Acknowledgments The authors are grateful to Felicity Sartain, Bio-
Nano Ltd., UK for her support in revising English, W. Schwieger,
Alexandra Inyat, Amer Inyat, Hallah Alyousef, Dirk Enk, University
of Erlangen-Nuernberg, Germany for their technical support in cata-
lyst characterization, Abdulaziz AL-Ghashem, Muhamad Ashinqiti,
Muhamad Ashammeri and Faez Al-Otaibi, King Abdulaziz City for
Science and Technology (KACST) for their support in reactor oper-
ation. KACST is acknowledged for funding this work under project
No. 29-280.
0
10
20
30
40
50
60
70
80
90
Reaction time (h)
Fig. 7 Time on stream analysis over CAT-III catalyst; catalyst
amount: 250 mg, H2/acetone molar ratio: 4.5, GHSV: 3,600 h-1
condensation reaction. The carbonyl group of acetone
could be activated by attachment to the surface of the
oxides through coordination of the O-atom to the Mn?
surface site. The surface O2- ion can act as a basic site to
extract a methyl proton resulting in the formation of a
nucleophile species [34]. This reacts with a neighboring co-
ordinated carbonyl centre in the presence of adsorbed
hydrogen atoms on metallic sites to form a molecule of
MIBK and water. It is reported that both the acidic and
basic sites of the catalysts are involved in the aldol con-
densation step [35]. In other words; formation of products
like MIBK, MO or DIBK is controlled by the acidic/basic
bi-functional nature of the catalysts.
References
1. Zhang G, Hattori H, Tanabe K (1988) Appl Catal 36:189
2. Reichle WT (1980) J Catal 63:295
3. Zahn H, Krasowski A (2002) Ullmann’s encyclopedia of indus-
trial chemistry, 6th edn (CD ROM).Wiley-VCH, Weinheim
4. Reith W, Dettmer M, Widdecke H, Fleischer B (1991) Stud Surf
Sci Catal 59:487
5. Weissermel K, Arpe HJ (1997) Industrial organic chemistry.
Wiley, Germany
6. Lawson KH, Nkosi B (1999) US Patent 6,008,416
7. Mitschker A, Wagner R, Lange PM (1988) In: Guisnet M (ed)
Heterogeneous catalysis and fine chemicals I. Elsevier,
Amsterdam
8. Talwalkar S, Mahajani S (2006) Appl Catal A 302:140
9. Yang SM, Wu YM (2000) Appl Catal A 192:211
10. Kozhevnikova EF, Kozhevnikov IV (2006) J Catal 238:286
11. Martinez-Ortiz MJ, Tichit D, Gonzalez P, Coq B (2003) J Mol
Catal A 201:199
12. Unnikrishnan R, Narayanan S (1999) J Mol Catal A 144:173
13. Unnikrishnan R, Narayanan S (1996) Appl Catal A 145:231
14. Chikan V, Molnar A, Balazsik K (1999) J Catal 184:134
15. Molnar A, Varga M, Mulas G, Mohai M, Bertoti I, Lovas A,
Cocco G (2001) Mater Sci Eng A 304–306:1078
16. Hetterley RD, Kozhevnikova EF, Kozhevnikov IV (2006) Chem
Commun 7:782
Time-on-stream studies with the composite catalysts
show no appreciable loss in activity for a period of 100 h.
Figure 7 shows the conversion of acetone and selectivity
for MIBK response for our best-performing catalyst (CAT-
III). The catalyst reaches a steady state in approximately
1 h and the conversion and MIBK selectivity are main-
tained constant for 100 h. These results indicate the suit-
ability of a Ru–C–ZnO catalyst for industrial application
for the acetone condensation process.
17. Waters G, Richter O, Czarnetzki BK (2006) Ind Eng Chem Res
45:6111
18. Bagabas AA, Akhmedov V, Mostafa MM, Al-Rabiah A (2011)
US Patent 7,951,976, 31 May 2011
4 Conclusions
19. Galvagno S, Capannelli G (1991) J Mol Catal 64:237
20. Tanabe K, Shimazu K, Hattori H, Shimazu K (1979) J Catal
57:35
21. Bagabas AA, Apblett A, Shemsi A, Seddigi Z (2008) Main Group
Chem 7:65
22. Navale SC, Ravi V, Mulla IS (2009) Sens Actuators B 139:466
23. Hosono E, Fujihara S, Kimura T, Imai H (2004) J Sol-Gel Sci
Tech 29:71
24. Baeza BB, Ruiz AG, Ramos IR (2000) Appl Catal A 192:289
25. Baeza BB, Ruiz AG, Wang P, Ramos IR (2001) J Catal 204:450
The Ru–C–n-ZnO composite catalysts are efficient for the
synthesis of MIBK by reductive condensation of acetone
in the gas phase. The best MIBK selectivity from one
step, acetone self-condensation over Ru–C–n-ZnO was
obtained at 350 °C, H2/acetone = 4.5 and Ru load-
ing = 2.5 %. A catalyst with Ru–C:n-ZnO ratio is 1:1
(wt%/wt%) showed the highest acetone conversion and
123

Ru–C–ZnO for the Synthesis of Methyl Isobutyl Ketone
26. Wang JQ, Wang YZ, Zhe SH, Qiao MH, Li HX, Fan KN (2004)
Appl Catal A 272:29
27. Neri G, Mercadante L, Donato A, Visco AM, Galvagno S (1994)
Catal Lett 29:379
28. Fernandez EVR, Ferreira AFP, Escribano AS, Kapteijn F, Rein-
oso FR (2008) J Catal 258:52
29. Wiberg KB, Crocker LS, Morgan KM (1991) J Am Chem Soc
113:3447
31. Yasumoto I (1984) J Phys Chem 88:4041
32. Bombos D, Bozga G, Bombos M, Stefan A, Stanciu I (2000)
Chem Pap 54:171
33. Morrison SR (1987) Sens Actuators B 12:425
34. Idriss H, Kim KS, Barteau MA (1993) J Catal 139:119
35. Di Cosimo JI, Diez VK, Apesteguia CR (1996) Appl Catal A
137:149
30. Li B, Hu G-S, Jin L-Y, Hong X, Lu J-Q, Luo M-F (2013) J Ind
Eng Chem 19:250
123