Ammoxidation of 2,6-dichloro toluene - From first trials to pilot plant studies

Article abstract of DOI:10.1016/j.cattod.2010.01.037

The scaling-up of the gas phase catalytic ammoxidation of 2,6-dichloro toluene (DCT) to 2,6-dichloro benzonitrile (DCBN) over a promoted vanadium phosphate (VPO) catalyst from first lab-scale experiments to pilot plant runs is reported. First experiments in a row of conversions of isomeric dichloro toluenes using simple, non-promoted VPO catalysts only show poor yield and selectivity. In particular, DCT ammoxidation is hindered due to bulky chlorine substituents probably preventing a sufficient interaction of the methyl group and lattice oxygen and/or N-containing surface species. Improved synthesis of VPO catalyst with the addition of promoters and γ-alumina or titania leads to significant increase in DCT conversion and DCBN yield. A Cr containing vanadyl pyrophosphate catalyst admixed with titania (anatase) showed conversion up to 97% with DCBN yields of ca. 80%. The same catalyst was also used for pilot plant runs, usually in the form of 5 mm × 3.5 mm shaped tablets that were prepared from a larger batch of solid synthesis. The scaling-up of the process using 100 ml of catalyst was investigated both by catalytic experiments and reactor simulations. The results showed that the temperature control will be crucial in scaling-up. Validation of simulation results with that of experimental results was also checked and a good agreement between measured and simulated results is observed.

Full text of DOI:10.1016/j.cattod.2010.01.037

Catalysis Today 157 (2010) 275–279
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Ammoxidation of 2,6-dichloro toluene—From first trials to pilot plant studies
A. Martin *, V.N. Kalevaru, Q. Smejkal
Leibniz-Institut f u¨ r Katalyse e. V. an der Universit a¨ t Rostock, Albert-Einstein-Str. 29a, D-18059 Rostock, Germany
A R T I C L E I N F O
A B S T R A C T
Article history:
The scaling-up of the gas phase catalytic ammoxidation of 2,6-dichloro toluene (DCT) to 2,6-dichloro
benzonitrile (DCBN) over a promoted vanadium phosphate (VPO) catalyst from first lab-scale
experiments to pilot plant runs is reported. First experiments in a row of conversions of isomeric
dichloro toluenes using simple, non-promoted VPO catalysts only show poor yield and selectivity. In
particular, DCT ammoxidation is hindered due to bulky chlorine substituents probably preventing a
sufficient interaction of the methyl group and lattice oxygen and/or N-containing surface species.
Available online 18 February 2010
Keywords:
Ammoxidation
VPO catalyst
Scale-up
Improved synthesis of VPO catalyst with the addition of promoters and
g-alumina or titania leads to
Pilot plant experiments
Reactor simulation
VPO catalyst
significant increase in DCT conversion and DCBN yield. A Cr containing vanadyl pyrophosphate catalyst
admixed with titania (anatase) showed conversion up to 97% with DCBN yields of ca. 80%. The same
catalyst was also used for pilot plant runs, usually in the form of 5 mm  3.5 mm shaped tablets that
were prepared from a larger batch of solid synthesis. The scaling-up of the process using 100 ml of
catalyst was investigated both by catalytic experiments and reactor simulations. The results showed that
the temperature control will be crucial in scaling-up. Validation of simulation results with that of
experimental results was also checked and a good agreement between measured and simulated results
is observed.
2
2
,6-Dichloro toluene
,6-Dichloro benzonitrile
Temperature profile
ß 2010 Elsevier B.V. All rights reserved.
1
. Introduction
The transfer of a chemical reaction from lab-scale to commer-
double bond of olefinic, aromatic or hetero aromatic hydrocarbons
to produce corresponding nitriles in a continuous process (e.g.
[5,6]). The heterogeneously catalyzed ammoxidation of various
hydrocarbons to synthesize wide range of industrial important
nitriles has been the subject of great interest in recent times
because the nitriles are very useful basic chemicals (e.g. acryloni-
trile by the SOHIO process) and organic intermediates to prepare a
good number of value added fine chemicals [5,6]. The reaction runs
via H-abstraction from such a methyl group leading to an allylic
intermediate that reacts with bulk oxygen to an oxygen-containing
intermediate through Mars–van Krevelen mechanism (e.g. [7,8]).
In general the industrial reaction is carried out in the gas phase
using fixed bed or fluidized bed reactors. Mainly transition metals
such as vanadium, antimony, molybdenum, etc. either in
supported form or as bulk materials are used as catalysts.
Furthermore, ammoxidation is a very ‘‘green’’ reaction because
oxygen or air is used as oxidant and in general, water is the only by-
product.
The present study was focused on the catalytic ammoxidation
of 2,6-dichloro toluene (DCT) to 2,6-dichloro benzonitrile (DCBN)
over vanadium phosphate (VPO) catalysts that are well known
from large scale conversion of n-butane to maleic anhydride (e.g.
[9]). Ammoxidation of DCT to DCBN in a single step is an
industrially important reaction as the target product provides an
easy access to prepare many effective herbicides and fungicides
[10,11]. DCBN is also an interesting intermediate for producing
various special kinds of engineering plastics. The results presented
cial dimension is pooled in scale-up studies (e.g. [1]). Such studies
are necessary to ensure the establishment of a reliable chemical
process for industrial application. This includes the move from the
lab-scale test set-up to commercial size equipment and the
development of a solid catalyst in a heterogeneously catalyzed
reaction. In such a move, a large number of tasks and questions
have to be taken into account for example, optimization of reaction
conditions, effect of heat and mass transfer, condensation and
separation of products, catalyst synthesis and shaping, catalyst
deactivation vs. long-term stability and attrition behavior (e.g.
[
1,2]). A pilot plant with a scale-up factor of ca. 25, if correctly
designed, allows detecting the critical aspects of the process and, at
the same time, points out solutions that can appreciably decrease
various risks in larger size equipment. These studies ensure that
the reactor model developed from lab studies can be related and
extrapolated to the design and performance of a commercial
reactor [3,4].
Ammoxidation is a partial oxidation reaction with selective
insertion of nitrogen into a methyl group (also aldehyde or
alcoholic groups are able to react) that is located in
a-position to
*
Corresponding author. Tel.: +49 381 1281 246; fax: +49 381 1281 51246.
E-mail address: andreas.martin@catalysis.de (A. Martin).
0920-5861/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2010.01.037

2
76
A. Martin et al. / Catalysis Today 157 (2010) 275–279
Table 1
Summary of used vanadium phosphate catalysts.
Catalyst
Precursor
Oxide
Promoter
Shape
VPPaq
VHPaq
–
–
g
TiO
TiO
–
1–1.25 mm sieve fraction
1–1.25 mm sieve fraction
1–1.25 mm sieve fraction
1–1.25 mm sieve fraction
5 mm  3.5 mm tablets (graphite)
VPPorg
VHPorg
VHPorg
VHPorg
VHPorg
–
VPPorg/S
MVPPorg/S
VPPpp
-Al
2
O
3
, TiO
2
–
2
2
Mo, Co, Fe, Cr
Cr
in this communication include first lab tests directed to the
conversion of isomeric dichloro toluenes over VPO catalysts
obtained by an aqueous synthesis route [12]. In addition, some
catalyst improvements by means of changed synthesis route using
Finally, the catalysts were shaped to larger tablets by hydraulic
press and calcined at 723 K for 3 h under weak oxidizing
2 2
atmosphere (0.5% O in N ). The resulting promoted and supported
VPO final catalyst consists of 25 wt.% VPP(M) with P:V = 0.95 and
M:V molar ratio = 0.05 (M = Co, Cr, Fe, and Mo). All catalysts used in
lab-scale tube reactor were synthesized in 100 g scale, crushed
again after pressing and sieved to the required size fraction (1–
1.25 mm).
organic reductants, admixture of oxides such as
g-alumina or
titania and addition of promoters are reported [13]. Furthermore,
details on the assembling of a pilot plant unit and processing of
kinetic studies and optimized reaction runs are also given [14,15].
A MVPPorg/S catalyst (VHPorg precursor (25 wt.%), M = Cr
(Cr:V = 0.05), P:V = 0.95, S = titania) was selected for pilot plant
tests (VPPpp) and synthesized in an amount of ca. 2.5 kg. After
catalyst synthesis, shaping of catalyst powder into the form of
tablets for pilot plant runs was conducted. Various trials with
graphite and Mg-stearate as auxiliaries were carried out.
Catalytic results proved that graphite is the best suited. The
catalyst powder was then industrially shaped to 5 mm  3.5 mm
tablets using 5% graphite as pressing auxiliary. The tablets were
finally calcined under similar conditions as above (i.e. 723 K, 3 h,
2
. Catalysts
Recently, it was shown that VPO derived catalysts are suitable
to catalyze ammoxidations of methyl aromatics and hetero
aromatics [5,16]. Additionally, these solids again revealed some
potential effects in improving their performance in ammoxidations
due to some synthesis variations and addition of various promoters
and co-components [13].
The layered vanadyl hydrogenphosphate hemihydrate
VOHPO
after calcination (in general 723 K, 3 h, 0.5% O
VO) (VPP) phase. For example, the VHP precursor compound
can be obtained according to an aqueous preparation route (aq)
using V , H PO and a reducing agent (e.g. oxalic acid) or an
organic synthesis route (org) using benzyl alcohol and 2-butanol as
reduction means. Various P:V ratios in the final VPP solid in a range
of 0.5–2 were adjusted. Table 1 gives a short summary on the used
catalysts. However, VHP compound can also be used directly as
catalyst precursor, which is then in situ transformed during
ammoxidation reaction into an ammonium ion-containing VPO
4
Á0.5 H
2
O (VHP) served as precursor for all catalysts; which
2 2
0.5% O in N ) [10].
2
in N ) gave mainly
2
(
2
P
2
O
7
3. Catalytic tests and scale-up of DCBN synthesis
2
O
5
3
4
3.1. First trials – ammoxidation of isomeric dichloro toluenes
at lab-scale
In an early stage of the investigations various isomeric dichloro
toluenes were converted to their corresponding nitriles mainly to
discover the effect of the position of the bulky chlorine
substituents on the activity and selectivity, details can be found
elsewhere [12]. VPPaq and VHPaq (transformed into AVP phase on-
stream) were used as catalysts and precursor, respectively. Fig. 1
shows a nitrile yield vs. dichloro toluenes conversion plot. It is clear
from the figure that the conversions and nitrile yields achieved are
higher in case of reactants, where the substituents are located
more remote from the reaction center, i.e. from the methyl group.
For instance, in case of 3,4-dichloro benzonitrile, a yield of ca. 55%
was reached at conversions close to 90–95%. On the other hand,
DCT, having substituents close to the reaction centre, the
conversion obtained was considerably low (between 50% and
4 2 3 2 7 2
compound ((NH ) [(VO) (P O ) ] – AVP) that is gradually trans-
formed into VPP, depending on the reaction conditions and time;
details on such studies can be found elsewhere [12,13]. The first
trials for ammoxidation of DCT described below were conducted
using VPPaq and VHPaq that was transformed during reaction first
to AVP and then slowly to VPP as well [12].
The preparation of ‘‘supported’’ VPO catalyst (VPPorg/S) was
carried out by solid–solid wetting method using the VHPorg
precursor (25 wt.%) and
g-alumina or titania (anatase). In this
solid–solid wetting procedure, the desired amount of VHPorg and
the support powder were mixed thoroughly in a porcelain mortar
for about 20 min until the color of the mixture is perfectly uniform.
Calcination as described above gives the VPPorg/S solids [13].
The preparation of promoted and ‘‘supported’’ catalyst
(MVPPorg/S) was carried out via (i) the addition of suitable
promoter to VHPorg precursor and (ii) by solid–solid wetting of the
promoted precursor with the support compound [13]. The
beneficial behavior of such promoters is recently discovered in
n-butane oxidation research [17–20]. The required amount of
promoter source (M:V molar ratio = 0.05; Co and Fe as acetates, Cr
as nitrate and Mo as ammoniumheptamolybdate) was dissolved in
aqueous ethanol (exclusively water was used in case of Mo) and
warmed up to ca. 343 K; the calculated amount of VHPorg
precursor in powder form was added. Then, the slurry was
evaporated to dryness and the resulting solid was oven dried at
Fig. 1. Nitrile yield vs. isomeric dichloro toluenes conversion: (NH
and (VO) used as catalysts (both aqueous route), dichloro toluene:
:NH :H O = 1:5:8:25, reaction temperatures of 688–728 K, 2,6-, 2,5-, 2,4-, 2,3-,
3,4-DCT.
4 2 3 2 7 2
) (VO) (P O )
3
93 K for 16 h. Subsequently, the desired amount of carrier (TiO
anatase) or -Al ) was taken for manufacture of MVPPorg/S
samples in a solid–solid wetting procedure as described above.
2
2 2 7
P O
(
g
2 3
O
O
2
3
2

A. Martin et al. / Catalysis Today 157 (2010) 275–279
277
Fig. 3. DCBN yield vs. DCT conversion: over transition metal promoter containing
vanadyl pyrophosphate catalysts (MVPPorg/TiO ) with M = Fe, Co, Mo, and Cr
(M:V = 0.05); DCT:H O:NH :air = 1:15:4:21, T = 673 K (open symbol marks
performance on non-promoted supported catalyst (VPPorg/TiO ) for comparison).
Fig. 2. DCBN yield vs. DCT conversion: over vanadyl pyrophosphate catalysts
VPPorg) with varying P:V ratio: DCT:H O:NH :air = 1:15:6:50, T = 713 K.
2
(
2
3
2
3
2
7
0%) compared to others and at the same time, DCBN yield was also
additional sorption sites for ammonia but also for Lewis bound
aromatic reactant molecules helping to speed up the reaction.
In a last stage of the catalyst development, additional transition
metal promoters such as Cr, Co, Fe or Mo were added during
catalyst synthesis in a M:V ratio of 0.05. The admixture of such
promoters revealed a further beneficial effect both on catalyst
activity and selectivity. As a result, the yield of DCBN increased
considerably, the results are plotted in Fig. 3. The catalytic tests
were carried out under comparable conditions and the use of Cr as
promoter revealed the highest conversion of DCT of 97% and a yield
of DCBN of 79%, which is in accordance with a DCBN selectivity of
ca. 81%. These results revealed that incorporation of promoters
play a key role in enhancing the performance. Most of the
promoted catalysts of this study displayed good potentiality giving
high yields of DCBN (70–80%) at high conversion levels (90–97%)
compared to their parent bulk VPO solids. After successful
achievement of DCBN yields close to 80%, the further investigations
were focussed on scale-up studies, which are described below in
more detail.
rather poor (around 20%). Mainly destruction products such as
carbon oxides were detected as reaction products in this case.
3.2. Lab-scale progresses – catalyst improvements and optimization
of reaction conditions
In a next step, catalyst performance had to be improved and
reaction conditions need to be optimized for obtaining enhanced
conversion of DCT with good selectivity towards DCBN. To achieve
such objective, the BET surface area rich VPPorg catalysts were
prepared and used for the target reaction. In the first step, bulk
VPPorg samples showing a wide range of P:V ratios from 0.5 to 2
were checked. Fig. 2 illustrates DCBN yield vs. DCT conversion
depending on the P:V ratio. It can be seen from the figure that the
solids with a P:V ratio of 0.7 and 0.95, respectively, reveal the
highest conversion and product yield around 90–95% and 60%,
respectively. This corresponds to ca. 65% selectivity of DCBN. For
further development, catalyst with P:V ratio of 0.95 was selected;
in particular due to its larger P content compared to the P:V = 0.7
sample. Interestingly, the yield obtained from VPPorg solids is
almost double compared to the one obtained with VPPaq or AVP
catalyst. The reason for such a dramatic increase in performance is
mainly due to the increased BET surface area of the VPPorg catalyst
and its morphology (e.g. [21]). Additionally, the preparation of
3.3. Scale-up – pilot plant set up and catalytic test runs
The pilot plant experiments were carried out in a flow
apparatus equipped with a stainless steel fixed bed reactor (i.d.:
0.016 m, length: 1 m; catalyst amount up to 100 ml), feeding
section, cooling and product separation units (for separating liquid
and solid products from gaseous mixture) including online-GC.
Gases and liquids were fed using mass flow controllers and HPLC
pumps, respectively. The gaseous stream was preheated and then
introduced to the reactor along with vaporized liquid feed (DCT
and water). The reactor outlet was heated to 623–673 K to avoid
the condensation of product components (DCBN, melting point:
416–419 K), and also to prevent deposition of by-products (e.g.
2 5 3 4
present VPO catalysts from V O , organic alcohols and o-H PO
leads to initial formation of VHP precursor phase with organic
alcohols trapped between the layers of phosphate structures,
which provoke some kind of disorderness in the catalyst structure.
During the successive calcination, VHP transforms into vanadyl-
pyrophosphate phase with the corresponding disorder in the plane
of layer stacking [18,21–23]. This disorder in the catalysts in turn
influence (i) the catalytic activity, (ii) redox properties and (iii) the
reducibility of the catalysts. Due to all these effects, the catalysts
prepared through organic route exhibited better performance
compared to the ones prepared by an aqueous route.
4 2 3 4 4
(NH ) CO or NH Cl formed by DCT total oxidation, (NH Cl
sublimation point is ca. 613 K!)) that can easily block the reactor
outlet. In general, catalyst was used in pure form or diluted with
glass beads (1:5) in order to study heat transfer effects and hot-
A further improvement was reached by mixing VPO and oxidic
3
‘‘supports’’ in a solid–solid wetting procedure. Titania (anatase)
spot behavior. The volumetric flow rate was set to 34–68 m /h
3
and -alumina were used for the preparation of those catalysts
g
(STP) resulting in a residence time of 700–4600 kgcat s/m (STP)
containing 25 wt.% VPO proportion, in general. The results showed
that a conversion of DCT around 90% and a DCBN yield up to ca. 70%
could be obtained using supported catalysts. This result clearly
indicates that a further increase in DCBN selectivity of ca. 20% is
successfully achieved compared to the performance of bulk VPOs.
The improvement of the catalytic performance seems to be caused
by acidic properties of the admixed oxides; acidic properties
depending upon catalyst amount used. The reaction conditions
also determined the fluid dynamics in the reactor depending on
volume flow rate and composition of the feed. Based on that the
dimensionless numbers can be derived (e.g. Re, Nu).
The product stream from the catalytic reactor consisted of
unconverted feed components (i.e. DCT, oxygen, and ammonia),
diluting agent (e.g. neon that was used sometimes to determine
(
Lewis and Brønsted sites) of aluminas are well known but also the
undesired N
products such as DCBN, (NH
minor amounts of other trace products. Analytically detected
concentration (GC) of those components (including NO ) was
2
formation caused by NH
3
oxidation) and the reaction
applied titania showed acidic properties due to remaining sulfate
stemming from the use of titanyl sulfate as a precursor for the
titania synthesis. Such additional acidity (acidic sites) can supply
4
)
2
CO , NH
3
4
Cl, CO, CO , water and
2
x

2
78
A. Martin et al. / Catalysis Today 157 (2010) 275–279
lower than 1 mass% and can be neglected in further discussion. C-
balance was always above 95% and close to 100% in most runs. The
gaseous product mixture from the reactor outlet is subsequently
cooled both externally and internally by quenching from reaction
temperature 623–673 K to 303–313 K depending upon the product
composition using an internal water spray tower (mainly for
dissolving salts and preventing blockages). The gas outlet from the
quenching column was connected to online-GC equipped with two
capillary columns such as PoraPlot Q & molecular sieve and TCD
and FID detectors.
Various test runs were conducted according to a pre-designed
experimental plan to study the influence of the partial pressure of
3 2
the reactants (DCT, NH , O and water). The tests revealed that
with increasing reaction temperature, the conversion of DCT and
yield of DCBN continuously increased and, however, total
oxidation product proportion too. Interestingly, a significant raise
in nitrogen formation occurred, which is undoubtedly caused by
accelerated ammonia oxidation. Furthermore, the results showed
that with increasing proportion of DCT in the feed, the DCBN
concentration in the product rises. However, in parallel to this
observation a slight increase in nitrogen formation due to
ammonia combustion and an increased total oxidation is also
observed [14]. The increase in the amount of ammonia in the feed
leads to decreasing conversion of DCT due to competitive
adsorption between the two reactant molecules for the same
sites. Therefore, a dropping amount of DCT adsorption sites is
Fig. 5. Influence of reaction temperature on DCT conversion and product selectivity;
feed mol composition DCT: 2.4%; ammonia: 10%; water: 20%; inert (Ne): 54.6%;
oxygen: 13% (full line: degree of ammonia combustion towards nitrogen).
investigated and presented in Fig. 5. Moreover, the degree of
ammonia combustion to nitrogen is plotted. It is obvious that
higher reaction temperature improves the conversion of oxygen
and thereby increased selectivity of undesired total oxidation
products. In addition, the temperature increase also leads to raised
ammonia oxidation. As expected, the higher selectivity of DCBN is
obtained in the lower temperature range between 633 and 643 K.
This observation is in a good agreement with long-term pilot plant
runs, where a process optimum was found under comparable
reaction conditions and temperature (643 K).
3
possible especially at higher partial pressures of NH . This situation
in turn causes a decline in DCBN formation. On the other hand,
nitrogen formation is significantly increased due to progressive
increase in ammonia combustion. Furthermore, oxygen concen-
tration was also varied with the aim to study its influence on the
The results from kinetic observations were consequently used
in the scale-up strategy. The best results obtained in the pilot plant
showed a DCT conversion close to 100% and a DCBN selectivity of
2
rate of formation of CO, CO , DCBN and nitrogen. The oxygen
partial pressure was related to the inert gas concentration in the
feed. The concentration of the oxygen portion was set between 6.8
and 13.5 mol%. Two results of these kinetic experiments are given
in Fig. 4. The composition of the product with increasing oxygen
concentration significantly changes with changing experimental
conditions. DCT conversion increases whereas oxygen conversion
decreases; DCBN yield is more or less constant. Under the chosen
conditions increase in oxygen concentration leads to both
increased total oxidation and ammonia combustion only.
The dependence of the reaction temperature on the product
selectivity and reaction yield is one of the most important issues in
every design work. In the case of exothermal oxidation of
aromatics, the temperature influence is much more crucial for
process safety. Keeping this aspect in mind, the influence of the
temperature on the reactant conversion and product selectivity is
ca. 83–84%. Total oxidation leads to the formation of CO and CO
2
and the sum selectivity of these two unwanted products amounts
to ca. 15%. The oxygen conversion reached ca. 93% [14]. Table 2
summarizes some reaction characteristics.
4. Comparison of catalytic run results and process simulation
After successful completion of lab-scale tests, an extended
effort has been further focussed on the development of kinetics
and process description of the presented ammoxidation system.
The best pilot plant runs were fitted by process model, based on
overall kinetics description. The kinetic study performed in lab-
scale showed strong non-isothermicity. The non-ideal behaviour of
the reaction system was confirmed in the pilot plant runs.
Therefore, it was necessary for evaluation of kinetic data to apply
a non-isothermal reactor model. For this purpose, a steady-state
pseudo-homogeneous fixed bed tubular reactor model with
negligible axial and radial diffusion was used. The details of
kinetics model can be found elsewhere [14,24,25].
The results of process simulation and experimental results
3
obtained at 653 K and at a residence time of 4600 kgcat s/m (STP)
are compared in Fig. 6. A very good agreement between measured
and simulated results could be observed. It is also evident that the
yield of main product DCBN around 80% is achieved at almost total
conversion of both oxygen and DCT. It is worth mentioning that
Table 2
Reaction characteristics.
Parameter
Dimension
Value
Reaction heat of ammoxidation of DCT
kJ/mol
K
À517
a
Adiabatic temperature increase
85–350
70–120
Fig. 4. Dependence of conversion of DCT and O
on the oxygen concentration (CO (mol%)) in the feed; T = 630 K; feed mol
composition DCT: 2.4%; ammonia: 10%; water: 20%; inert: (67.6 À x)%; oxygen: x%.
2 2 2
and yield of DCBN, CO, CO and N
2
Overall heat transfer coefficient
J/(s K m )
2
a
Due to reaction conditions at chosen catalyst dilution.

A. Martin et al. / Catalysis Today 157 (2010) 275–279
279
in the feed and temperature of reaction showed significant
influence on the product distribution. Finally, a good agreement
between simulated and experimental results is obtained.
Acknowledgements
The authors gratefully acknowledge Mrs. P. R o¨ ssler and Mrs. J.
Kubias for experimental assistance and Tessenderlo Chemie S.A.
(
Belgium) for financial support.
References
[1] A. Bisio, R.L. Kabel, Scaleup of Chemical Processes: Conversion from Laboratory
Scale Tests to Successful Commercial Size Design, John Wiley & Sons Inc., 1984,, p.
175.
Fig. 6. Comparison of pilot plant experiment and process simulation; Optimized
[2] V.R. Gangwal, J. van der Schaaf, B.F.M. Kuster, J.C. Schouten, J. Catal. 232 (2005)
feed composition DCT: 2.85%; ammonia: 12%; water: 42.1%; inert: 29.85%; oxygen:
432.
1
3.2% (catalyst: VPPpp).
[3] L. Ying, V. Raman, R.O. Fox, A.D. Harvey, Chem. Eng. Sci. 59 (2004) 5167.
[
[
[
[
[
4] K.B. Shyamal, A.K. Dalai, S.K. Maity, Catal. Today 64 (2001) 333.
5] A. Martin, B. L u¨ cke, Catal. Today 57 (2000) 61.
6] F. Cavani, F. Trifir o` , Chim. e l’Ind. 70 (1988) 58.
7] Y. Zhang, A. Martin, H. Berndt, B. L u¨ cke, M. Meisel, J. Mol. Catal. A 118 (1997) 205.
8] P. Mars, D.W. van Krevelen, Chem. Eng. Sci. 3 (1954) 41.
these results have been reproduced in another long term pilot
plant run, which gave quite comparable results. Finally, the
catalyst developed and the process set up was successfully applied
in the industrial scale.
[9] G. Centi (Ed.), Vanadyl pyrophosphate catalysts Catal. Today 16 (1993) 1.
10] A. Martin, K.V. Narayana, B. L u¨ cke, D. van Deynse, M. Belmans, F. Boers, WO Patent
[
[
03/101939 A2 (2003) to Tessenderlo Chemie S.A., Belgium.
11] Z. Qiong, H. Chi, X. Guangyong, X. Chongwen, C. Yuanyin, Synth. Commun. 29
(1999) 2349.
5
. Conclusions
[
[
12] A. Martin, B. L u¨ cke, G.-U. Wolf, M. Meisel, Catal. Lett. 33 (1995) 349.
13] K.V. Narayana, A. Martin, B. L u¨ cke, M. Belmans, F. Boers, D. van Deynse, Z. Allg.
Anorg. Chem. 631 (2005) 25.
Various VPO based catalyst compositions were explored and
optimum composition and suitable promoters could be identified.
The VPOs prepared through organic route are found to exhibit
superior performance compared to the ones prepared by aqueous
route. The efforts to enhance the conversion of DCT and the yield of
DCBN were successful. P:V ratio is a crucial factor on improving the
performance. Nature of promoter has a clear influence on the
catalytic properties of VPO solids. Among different promoters
applied, Cr is observed to exhibit better performance giving the
yield of DCBN close to 80%. The objective of pilot plant set up,
product separation, and the scale-up studies from first trials to
pilot plant runs were also quite successful. Oxygen concentration
[14] N. Dropka, V. Narayana Kalevaru, A. Martin, D. Linke, B. L u¨ cke, J. Catal. 240 (2006) 8.
[15] N. Dropka, Q. Smejkal, V.N. Kalevaru, A. Martin, Appl. Catal. A: Gen. 349 (2008)
125.
[
16] A. Martin, B. L u¨ cke, H. Seeboth, G. Ladwig, Appl. Catal. 49 (1989) 205.
[17] G.J. Hutchings, R. Higgins, J. Catal. 162 (1996) 153.
[18] G.J. Hutchings, C.J. Kiely, M.T. Sananes-Schulz, A. Burrows, J.C. Volta, Catal. Today
40 (1998) 273.
[
[
19] A. Tuel, J.C. Volta, C.J. Kiely, G.J. Hutchings, N. J. Chem. 25 (2001) 125.
20] P.L. Gai, K. Kourtakois, Science 267 (1995) 661.
[21] A.A. Rownaghi, Y.H. Taufiq-Yap, F. Rezaei, Ind. Eng. Chem. Res. 48 (2009) 7517.
[
[
[
22] F. Cavani, G. Centi, F. Trifir o` , G. Poli, J. Therm. Anal. Calorim. 30 (1985) 1241.
23] P. Cˇ apkov a´ , K. Mel a´ nova, L. Bene sˇ , H. Schenk, J. Mol. Model. 6 (2000) 9.
24] K. Hansen, Chem. Eng. Sci. 28 (1973) 723.
[25] L. Forni, Appl. Catal. A: Gen. 37 (1988) 305.