The effect of water on the heterogeneously catalyzed selective oxidation of acrolein: An isotope study

Article abstract of DOI:10.1002/cctc.201400099

The effect of water on the selective gas phase oxidation of acrolein to acrylic acid on a Mo/V/W mixed oxide catalyst was studied by performing steady-state isotopic transient kinetic analysis experiments with H ₂¹⁸O. Experiments were performed in the temperature range of 90-345C at ambient pressure. It could be shown that acrolein exchanges its carbonylic oxygen with oxygen from water even at low temperatures (<200C), at which no acrolein oxidation occurs. At higher temperatures (>200C), the oxygen atoms of the water molecules incorporate into all oxidation products such as acrylic acid, carbon monoxide, and carbon dioxide.

Full text of DOI:10.1002/cctc.201400099

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201400099
The Effect of Water on the Heterogeneously Catalyzed
Selective Oxidation of Acrolein: An Isotope Study
Tina Petzold, Nina Blickhan, Alfons Drochner, and Herbert Vogel*[a]
Dedicated to Professor Robert Schlçgl on the occasion of his 60th birthday
The effect of water on the selective gas phase oxidation of
acrolein to acrylic acid on a Mo/V/W mixed oxide catalyst was
studied by performing steady-state isotopic transient kinetic
analysis experiments with H218O. Experiments were performed
in the temperature range of 90–3458C at ambient pressure. It
could be shown that acrolein exchanges its carbonylic oxygen
with oxygen from water even at low temperatures (<2008C),
at which no acrolein oxidation occurs. At higher temperatures
(>2008C), the oxygen atoms of the water molecules incorpo-
rate into all oxidation products such as acrylic acid, carbon
monoxide, and carbon dioxide.
Introduction
Acrylic acid is produced from propylene in a two-step process,
with an annual production of more than 4 million tons.[1]
Mixed oxides such as Mo/V/W oxides are used as catalysts in
the second reaction step. Their composition is empirically de-
termined and found to be complex. Many attempts have been
made to understand the mechanism of the partial oxidation of
acrolein with molecular oxygen as well as the role and cooper-
ation of different metal components. However, for an entire
understanding of the mechanism, additional aspects have to
be considered: one is the effect of water, which is a natural by-
product in the industrial process and affects the mechanism of
acrolein oxidation significantly.
5) The desorption of acrylic acid is promoted in the presence
of water; therefore, the total oxidation of acrylic acid is re-
duced.[9–12]
6) Water increases the adsorption sites on the catalyst surface;
therefore, more acrolein can be adsorbed on the catalyst
surface.[13–16]
To obtain more mechanistic information, isotopic investiga-
tions are required. Only a few experiments, performed under
reaction conditions, with H218O on oxidic catalyst systems are
mentioned in the literature.
For example, Suprun et al. investigated the oxidation of
1-butene to acetic acid on the VOx–TiO2 system. The conclu-
sions drawn from this study confirm the assumption that
water promotes the desorption of the partial oxidation prod-
uct on the catalyst surface. Acetic acid is probably chemically
bonded to the catalyst surface as an acetate species; water hy-
drolyses the acetate to the free acid. There is evidence that
water has an active role in the reaction mechanism because
the oxygen atom from the water molecule incorporates into
the oxidation products faster than does the molecular oxygen
from the gas phase.[12]
The effect of water on the structure and performance of
mixed oxide catalysts has been studied earlier. In the presence
of additionally fed water, both the conversion and the selectivi-
ty towards acrylic acid are increased. The addition of 7.5 vol%
of water leads to a much higher conversion at lower tempera-
tures. A selectivity of more than 90% is achieved at 80% con-
version.[2]
In general, the effect of water on oxidic systems is discussed
in the literature:
1) Physical effects are discussed, such as the improvement in
heat removal.[3]
The aim of our study was to clarify the role of water in the
selective oxidation of acrolein to acrylic acid on Mo/V/W mixed
oxides. The effect of water on the selective oxidation on these
mixed oxides was investigated by performing temperature-
programmed experiments. We also performed isotope ex-
change [steady-state isotopic transient kinetic analysis
(SSITKA)] experiments with H218O at several temperatures to
follow the reaction pathway of the oxygen originating from
water into educts and oxidation products. Herein, a Mo/V/W
mixed oxide was used, which shows up to 70% selectivity at
80% conversion.[17,18]
2) Water prevents coke formation; more active sites are avail-
able.[4]
3) Water accelerates the re-oxidation of the catalyst.[5,6]
4) Water increases the active sites of the catalyst.[7,8]
[a] T. Petzold, N. Blickhan, Prof. Dr. A. Drochner, Prof. Dr. H. Vogel
Ernst-Berl-Institut fꢀr Technische und Makromolekulare Chemie
Technische Universitꢁt Darmstadt
Alarich-Weiss-Straße 8, 64287 Darmstadt (Germany)
Fax: (+49)6151-16-3465
E-mail: vogel@ct.chemie.tu-darmstadt.de
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2053 – 2058 2053

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
All isotopologues of the combustion products (carbon mon-
oxide and carbon dioxide) are formed in a similar manner.
The next question is whether the oxygen exchange between
water and acrolein occurs only on the catalyst or even in the
gas phase. For this purpose, a blind experiment (SSITKA experi-
ment) without any catalyst was performed, which is described
below (Figure 3).
SSITKA experiments with H216O/H218O
Concentration profiles
SSITKA experiments with H218O were performed in a tempera-
ture range of 90–3458C. By using this method, it is possible to
follow the reaction pathway of the oxygen atom from water.
Under steady-state conditions at 908C, H216O was replaced
stepwise by H218O (Figure 2). The resulting concentration pro-
Figure 1. Concentration profiles of the temperature-programmed reaction
experiment: a) acrolein, oxygen, and water; b) acrylic acid, carbon monoxide,
and carbon dioxide. Feed: 5 vol% acrolein, 10 vol% oxygen, and 7.5 vol%
H218O; heating rate: 10 Kminꢀ1
.
Results and Discussion
Temperature-programmed reaction experiment with H218O
Figure 2. SSITKA experiment with H218O at 908C. Transient responses of acro-
lein, oxygen, and water after the stepwise change from H216O to H218O in the
feed gas at t=t0 (5 vol% acrolein, 10 vol% oxygen, 7.5 vol% water in
helium, 50 mg of Mo8V2W1.5Ox).
The concentration profiles of the educts and products ob-
tained during a temperature-programmed reaction experiment
with H218O are shown in Figure 1. The concentration profiles of
the educts acrolein, oxygen, and water are shown in Figure 1a.
Surprisingly, 18O-labelled acrolein could be detected even at
temperatures below 2008C, at which no oxidation occurs. In
addition, the formation of H216O could be verified, which indi-
cates that water and acrolein can exchange their oxygen
atoms. H216O could be formed through an exchange reaction
between H218O and 16Acr. Below 2008C, this is the only possible
way to form H216O. At higher temperatures, H216O could also
be formed by the combustion of acrolein or acrylic acid with
the start of the oxidation reactions. In addition to the ex-
change reaction between water and acrolein, an oxygen ex-
change between H218O and 16O-labelled acrylic acid can occur
at higher temperatures. 1618O2 and 1818O2 could not be detect-
ed. This finding leads to the conclusion that once gas phase
oxygen incorporates into the catalyst, there is no recombina-
tion of the oxygen atoms (within the resolution of the used
online mass spectrometer).
files reveal that acrolein exchanges more than 50% of its car-
bonyl oxygen with 18O-labelled oxygen from water. During
these experiments, the concentration of the molecular oxygen
from the gas phase remains constant. This finding confirms the
overall steady state of the experiment and that no exchange
between molecular oxygen from the gas phase and oxygen
from water or acrolein occurs.
To ensure that this effect is due to the presence of the cata-
lyst, a reference measurement without any catalyst was per-
formed (Figure 3). At 1608C, acrolein can exchange 40% of its
oxygen with oxygen from water (Figure 3a). In contrast, the
reference measurement without any catalyst indicates that no
labelled acrolein appears (Figure 3b). This observation reveals
that the exchange of oxygen occurs only on the catalyst sur-
face (Scheme 1). The gas phase oxygen does not participate di-
rectly in the exchange reaction.
However, all three isotopologues of acrylic acid (1616AA,
1618AA, and 1818AA) are formed. Their maxima appear at differ-
ent temperatures (1616AA: 3758C; 1618AA: 3508C; 1818AA: 3258C).
The sum of all acrylic acid species corresponds to the acrylic
acid concentration of a temperature-programmed reaction ex-
periment with the similar amount of H216O in the feed. Hence,
no kinetic isotope effect could be evaluated.
Scheme 1. Illustration of the oxygen exchange between water and acrolein.
ChemCatChem 2014, 6, 2053 – 2058 2054
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 3. SSITKA experiment with H218O at 1608C. Transient responses of
acrolein, oxygen, and water after the stepwise change from H216O to H218O
in the feed gas at t=t0. a) Exchange on the catalyst. b) Exchange in an
empty reactor (5 vol% acrolein, 10 vol% oxygen, 7.5 vol% water in helium,
50 mg of Mo8V2W1.5Ox).
To investigate this oxygen exchange under reaction condi-
tions, we performed a SSITKA experiment at 2708C. The re-
sponse of the system after the exchange from H216O to H218O
is shown in Figure 4. Again, the concentration of oxygen re-
mains constant during the experiment. At this temperature,
acrolein converts to acrylic acid and combustion products. In
this regard, the conversion is approximately 44%. The labelled
oxygen of water appears in every product, as shown in Fig-
ure 4a–c. For both acrylic acid and carbon dioxide, mixed la-
belled and fully substituted isotopologues were detected. The
concentration of the mixed labelled species is always higher
than that of the fully substituted species.
Figure 4. SSITKA experiment with H218O at 2708C. Transient responses of
a) acrolein, oxygen, and water, b) acrylic acid, and c) carbon monoxide and
carbon dioxide after the stepwise change from H216O to H218O in the feed
gas at t=t0 (5 vol% acrolein, 10 vol% oxygen, 7.5 vol% water in helium,
50 mg of Mo8V2W1.5Ox).
Table 1. Degree of labelling of acrolein, acrylic acid, carbon dioxide,
carbon monoxide, and water.
Component
Acrolein
Degree of labelling, h
Degree of labelling
Acrylic acid
Firstly, the degree of labelling of the reactants during a SSITKA
experiment with H218O is discussed for the temperatures at
which no reaction occurs (measurements at 90, 120, and
1608C). Here, we observe an interaction only between acrolein,
water, and the catalyst. Then, the degree of labelling of all re-
actants is discussed for the entire temperature range.
Carbon dioxide
Carbon monoxide
Water
The degree of labelling of the individual components is cal-
culated by using the volume fraction f (Table 1).
The temperature-dependent degree of labelling of acrolein
and water 10 min after switching from H216O to H218O is depict-
ed in Figure 5. Up to 1608C, no acrolein oxidation occurs and
the interaction between water and acrolein can be described
without superposition of the oxidation reaction. Diffuse reflec-
tance infrared Fourier transform spectroscopy (DRIFTS) investi-
gations on a similar catalyst[19] reveal that the band intensities
of the hydroxyl groups on the catalyst surface decreases with
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2053 – 2058 2055

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
as acrylic acid, acrolein, and water above 3158C. In contrast,
the degree of labelling of carbon dioxide increases with in-
creasing temperatures. This difference between carbon monox-
ide and carbon dioxide is an indicator of different formation
pathways.
Conclusions
The results of steady-state isotopic transient kinetic analysis ex-
periments with H218O on Mo8V2W1.5Ox indicate the participation
of water in the acrolein partial oxidation reaction. At low tem-
peratures (908C), a fast oxygen exchange between water and
acrolein occurs. Under the chosen conditions, more than 50%
of acrolein exchanges its carbonylic oxygen with 18O-labelled
oxygen from water. This exchange occurs only on the catalyst
surface. Without any catalyst, no exchange reaction occurs. Re-
garding the degree of labelling, the situation changes at tem-
peratures above 1608C, at which acrolein is converted to acryl-
ic acid and combustion products.
Figure 5. Degree of labelling of acrolein and water versus temperature (after
10 min in the presence of 5 vol% acrolein, 10 vol% oxygen, and 7.5 vol%
H218O).
The labelled oxygen is recovered in every oxidation product
owing to the oxygen exchange between acrolein and water
(Scheme 2).
Figure 6. Degree of labelling after 10 min in the presence of 7.5 vol% H218O,
5 vol% acrolein, and 10 vol% oxygen at 12 temperatures.
Scheme 2. Illustration of the incorporation of oxygen originating from H218O
into acrolein.
increasing temperature. This is most likely the explanation for
the decreasing degree of labelling of acrolein. With increasing
temperature, the number of labelled hydroxyl groups on the
catalyst surface decreases and acrolein has therefore less possi-
bility to exchange its carbonyl oxygen.
The formation of double-labelled acrylic acid involves differ-
ent pathways. Owing to the adsorption and desorption of
water, 18O-labelled oxygen permanently incorporates into the
catalyst (MꢀOꢀM+H2OÐ2MꢀOH). The adsorption and desorp-
tion of acrolein on labelled hydroxyl groups results in labelled
acrolein. The next step towards the formation of acrylic acid is
the incorporation of oxygen. Therefore, the degree of labelling
of acrylic acid depends on the oxygen isotope (16O or 18O),
which incorporates into the acrolein surface complex. 1616AA or
1618AA could be formed on the basis of 16Acr and 1618AA, and
1818AA could be formed on the basis of 18Acr. Without doubt,
the formation of 1818AA could also proceed via an adsorption
and desorption equilibrium of acrylic acid and labelled hydrox-
yl groups (formal reaction: H218O+1618AAÐH216O+1818AA). An-
other way to form double-labelled acrylic acid was discovered
earlier. In previous works, we could show through 18O2-SSITKA
experiments with 1616AA that in addition to the combustion re-
action the formation of 1618AA and 1818AA occurs,[20–22] which
could only be due to an oxygen exchange between acrylic
acid and the catalyst surface. If there was no exchange, only
a stationary trend of 1616AA would be expected regarding the
The oxidation reaction starts at temperatures above 1608C,
and the degree of labelling of acrolein increases with increas-
ing activity (Figure 6). With increasing conversion, more gas
phase 16O2 takes part in the reaction. The catalyst is re-oxidized
with unlabelled oxygen from the gas phase. Hence, this
oxygen is available for exchange with oxygen from acrolein
and thus incorporates into all oxidation products. Thus, the
degree of labelling of acrolein decreases with increasing con-
version. As a result, the degree of labelling of water decreases
above 1608C. At 2508C and above, the degree of labelling of
acrolein, acrylic acid, and water is identical.
The degree of labelling of carbon monoxide and carbon di-
oxide is plotted only in the temperature range 300–3458C.
Below 3008C, the concentrations of these components are too
small to give reliable values.
The degree of labelling of carbon monoxide decreases with
increasing temperature and has the same isotopic distribution
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2053 – 2058 2056

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
results of SSITKA experiments with 18O2. The molecular oxygen
from the gas phase does not exchange directly its oxygen
atoms with those from water or any other component.
The variation in temperature leads to the conclusion that
the degree of labelling of acrolein depends on the concentra-
tion of the hydroxyl groups on the catalyst surface. Lower tem-
peratures lead to a higher hydroxyl group density on the sur-
face.[19] Therefore, an increased oxygen exchange with oxygen
from acrolein occurs. In contrast, the surface hydroxyl group
density decreases with increasing temperature through dehy-
dration reactions.
Temperature-programmed reaction experiments
Temperature-programmed reaction experiments were conducted
with H218O in the feed to evaluate the effect of water on the per-
formance of mixed oxides in dependence on temperature.
The calcined catalyst was pretreated in the reactor in an oxidative
atmosphere (4008C, 60 min, flow rate: 20 mLminꢀ1
oxygen in helium). Subsequently, three consecutive temperature-
programmed reaction cycles were performed (heating rate:
10 Kminꢀ1, final temperature: 4808C). The feed consisted of acrole-
in (5 vol%), oxygen (10 vol%), and water (7.5 vol%). Each tempera-
ture-programmed reaction cycle was followed by a re-oxidation
process (10 vol% oxygen at 4008C for 60 min). By repeating this
method, we could assess the activation and deactivation processes
as well as the long-term stability of the catalyst under thermal
stress.
, 10 vol%
In addition, the degree of labelling of all components,
except for carbon dioxide, decreases at higher temperatures
because with increasing conversion the catalyst re-oxidizes
with unlabelled oxygen from the gas phase. This unlabelled
oxygen is then available for the exchange with oxygen from
acrolein (or acrylic acid). On the one hand, H218O forms fewer
hydroxyl groups on the surface at higher temperatures and
thus acrolein (or acrylic acid) has less opportunity to exchange
its oxygen with labelled oxygen from water. On the other
hand, more unlabelled oxygen species (1616O2) incorporate into
the catalyst through re-oxidation and are exchanged with
oxygen from acrolein and incorporate into all oxidation prod-
ucts.
SSITKA experiments
Isotope exchange experiments were performed to obtain informa-
tion about the oxygen dynamics. The technique involved the re-
placement of a reactant by one of its isotopologues in the gas
flow (here H216O against H218O), but the overall reaction remained
at steady state.
To abbreviate the start-up phase of the catalyst system under reac-
tion conditions, a pretreatment was performed as follows: A 1 h
treatment (10 vol% oxygen, 4008C) followed by two temperature-
programmed reaction cycles (10 vol% oxygen, 5 vol% acrolein,
7.5 vol% water, heating rate: 10 Kminꢀ1) and consecutive re-oxida-
tion of the catalyst (30 min, 4008C, 10 vol% oxygen) were per-
formed.
Experimental Section
All experiments were performed under the continuous flow of the
reaction gas mixture (5 vol% acrolein, 10 vol% oxygen, 7.5 vol%
water, flow rate: 20 mLminꢀ1). Under steady-state conditions, after
setting the target value to the appropriate temperature the switch
to H218O (>97%, Sigma-Aldrich) occurred. After another 10 min,
H218O was switched back to H216O.
The response of the system after the water exchange was moni-
tored, and the rate of oxygen exchange and the isotopic distribu-
tion of all components were analyzed. Measurements were per-
formed at 12 temperatures ranging from 90 to 3458C.
Catalyst preparation
The catalyst with the general formula Mo8V2W1.5Ox (23ꢁxꢁ33.5)
was prepared by using the preparation strategy developed by
Kunert et al.; details are described in previous works.[16,18,23,24] An
aqueous solution containing ammonium heptamolybdate, ammo-
nium metavanadate, and ammonium metatungstate with the de-
sired metal ratio of the solid catalyst was adjusted to a pH value of
5 and was boiled under reflux for 90 min. The cooled solution was
spray dried (600kPa air, 2608C). The dried precursor was calcined
in nitrogen atmosphere at 3258C for 4 h.
Keywords: acrolein · heterogeneous catalysis · mixed oxides ·
oxidation · isotopic exchange
Reactor apparatus
[1] Ullmann’s Encyclopedia of industrial chemistry, 7th ed., Wiley-VCH, Wein-
heim, 2009.
[2] T. Jekewitz, A. Drochner, H. Vogel, Catal. Commun. 2012, 20, 25–28.
[3] H.-J. Arpe, Industrielle Organische Chemie, 6th ed., Wiley-VCH, Weinheim,
2007.
[4] J.-S. Yu, L. Kevan, J. Phys. Chem. 1991, 95, 6648–6653.
[5] H. Redlingshçfer, O. Krçcher, W. Bçck, K. Huthmacher, G. Emig, Ind. Eng.
Chem. Res. 2002, 41, 1445–1453.
[6] L. B. Levy, P. B. DeGroot, J. Catal. 1982, 76, 385–392.
[7] R. Recknagel, L. Riekert, Chem. Tech. 1994, 46, 324–331.
[8] E. K. Novakova, E. G. Derouane, J. C. Vedrine, Catal. Lett. 2002, 83, 177–
182.
Isotope exchange experiments were conducted in a setup de-
scribed elsewhere.[23,24] An arrangement of several mass flow con-
trollers and two-stage gas saturators allowed a flexible dosage of
gaseous and liquid components. A quartz U tube served as the re-
actor, in which the catalyst was fixed between two quartz wool
stoppers. The reactor was placed in an electrically heated oven,
which was temperature controlled. The reaction gas was analyzed
online with a mass spectrometer (GAM 400, InProcess Instruments,
Germany).
A prerequisite for a SSITKA experiment with H218O was a nearly
ideal switch between H216O and H218O without disturbing the over-
all flow. A syringe pump was used for this purpose. It consisted of
a step motor driving a screw that moved two plungers simultane-
ously into two cylinders: one was filled with water (H216O) and the
other one with 18O-labelled water (H218O). With use of a heating
block system, water was continuously evaporated and was carried
to the feed with inert gas.
[9] T. V. Andrushkevich, Catal. Rev. Sci. Eng. 1993, 35, 213–259.
[10] E. M. Erenburg, T. V. Andrushkevich, G. Y. Popova, A. A. Davydov, V. M.
Bondareva, React. Kinet. Catal. Lett. 1979, 12, 5–11.
[11] Y. A. Saleh-Alhamed, R. R. Hudgins, P. L. Silveston, J. Catal. 1996, 161,
430–440.
[12] W. Y. Suprun, D. P. Sabde, H.-K. Schꢁdlich, B. Kubias, H. Papp, Appl. Catal.
A 2005, 289, 66–73.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2053 – 2058 2057

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
´
[13] J. Tichy, A. A. Davydov, Collect. Czech. Chem. Commun. 1976, 41, 834–
[20] P. Kampe, Dissertation, Technische Universitꢁt Darmstadt, 2007.
[21] A. Drochner, P. Kampe, N. Blickhan, T. Jekewitz, H. Vogel, Chem. Ing.
Tech. 2011, 83, 1667–1680.
838.
´
[14] J. Tichy, J. Svachula, J. Machek, N. C. Allachverdova, React. Kinet. Catal.
Lett. 1986, 31, 159–166.
[15] M. M. Bettahar, G. Costentin, L. Savary, J. C. Lavalley, Appl. Catal. A 1996,
145, 1–48.
[16] J. Kunert, A. Drochner, J. Ott, H. Vogel, H. Fueß, Appl. Catal. A 2004, 269,
53–61.
[17] S. Endres, P. Kampe, J. Kunert, A. Drochner, H. Vogel, Appl. Catal. A
2007, 325, 237–243.
[22] A. Drochner, P. Kampe, N. Menning, N. Blickhan, T. Jekewitz, H. Vogel,
Chem. Eng. Technol. 2014, 37, 398–408.
[23] A. Drochner, P. Kampe, J. Kunert, J. Ott, H. Vogel, Appl. Catal. A 2005,
289, 74–83.
[24] R. Bçhling, A. Drochner, M. Fehlings, D. Kçnig, H. Vogel, Chem. Ing. Tech.
1999, 71, 226–230.
[18] P. Kampe, L. Giebeler, D. Samuelis, J. Kunert, A. Drochner, F. Haaß, A. H.
Adams, J. Ott, S. Endres, G. Schimanke, T. Buhrmester, M. Martin, H.
Fuess, H. Vogel, Phys. Chem. Chem. Phys. 2007, 9, 3577–3589.
[19] K. Krauß, A. Drochner, M. Fehlings, J. Kunert, H. Vogel, J. Mol. Catal. A
2000, 162, 413–422.
Received: January 30, 2014
Revised: March 25, 2014
Published online on May 30, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 2053 – 2058 2058