Sunteți pe pagina 1din 5

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/231365880

Kinetics of rhodium-catalyzed methanol


carbonylation
ARTICLE in INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH NOVEMBER 1992
Impact Factor: 2.59 DOI: 10.1021/ie00011a011

CITATIONS

READS

26

3 AUTHORS, INCLUDING:
Lech Nowicki

Stanislaw Ledakowicz

Lodz University of Technology

Lodz University of Technology

64 PUBLICATIONS 431 CITATIONS

243 PUBLICATIONS 2,265 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,


letting you access and read them immediately.

SEE PROFILE

Available from: Stanislaw Ledakowicz


Retrieved on: 21 March 2016

Ind. Eng. Chem. Res. 1992,31, 2472-2475

2472

Ch = pollutant concentration in liquid at outlet of column,


mol/m3 or ppmv
Co, = oxygen concentration in gaa phase, mol %
CRHi= reactant Concentration in gas phase for species i, ppmv
C O R E = inlet reactant concentration in gas phase for species
i, P P ~ V
Em = percent removal of BTX, %
G = superficial gas mass velocity, kg/(m2.s)
G/L = gas to liquid ratio, w/w
k = reaction rate constant, l/s or m3/(kmol.s)
kl,kz,k3 = surface reaction rate constant in Mars-van Krevelen mechanism, l/s
L = superficial liquid mass velocity, k g / ( m 2 d
RHi = hydrocarbon species i
t = space time, s
VHSV = hourly volume space velocity, m3 of feed gas/(m3
of &&h)
vi = stoichiometric coefficient for deep oxidation of species
1

X = conversion of BTX, %
Registry No. C&,

71-43-2;PhMe, 108883;xylene, 133CL20-7.

Literature Cited
Andereon, R. B.; Stein, K. C.; Feenan, J. J.; Hofer, L. F. J. Catalytic
Oxidation of Methane. Znd. Eng. Chem. 1961,53(lo), 8Oi+812.
Cheng, 5.; Chuang, K. T. Simultaneous Methanol Removal and
Destruction from Wastewater in a Trickle-Bed Reactor. Can. J.
Chem. Eng. 1992,in press.
Evaldeson, L.; Lowendahl, L.; W r s t e d t , J. E. Fibrillar Alumina an
a Wash-Coat on Monoliths in the catalytic Oxidation of Xylene.
Appl. Catal. 1989,55,123-136.
Gangwal, S.K.; Mullins, M. E.; Spivey, J. J.; Caffrey, P. R.; Tichenor,
B. A. Kinetics and Selectivity of Deep Catalytic Oxidation of

n-Hexane and Benzene. Appl. Catal. 1988,36,231-247.


Golodets, G. I. Heterogeneous Catalytic Reactions Involving Molecular Oxygen. In Studies in Surface Science and Catalysis;
Ross,J. R. H., Ed.; Elsevier: New York, 1983;Vol. 15,Chapters
X and XIV.
Kittikul, P.; Veenstra, A,; Akolade, J. N.; Weinert, M. A. Study of
High Water Temperature Effects on Air Stripping of Volatile
Organics from Water. Forty-fourth Purdue Industrial Waste
Conference Proceedings; Le*.
Chelsea, MI, 1990;pp 435452.
Kosusko, M.; Mullins, M.E.; Ramanathan, K.; Rogers, T. N. Catalytic Oxidation of Groundwater Stripping Emission. Enuiron.
Prog. 1988,7 (2),136-142.
Lamarre, B. L.; MaCarry, F.; Stover, E. L. Design, Operation and
Results of Pilot Plant for Removal of Contaminants from
Groundwater. Roceedings of the Third National Symposium on
Aauifer Restoration and Groundwater Monitoring,
-. Columbus,
OH, May 1983.
Sklyarov, A. V.; Tretyakov, I. I.; Shub, B. R.;Roginakii, S. Z.Dokl.
Akad. Nauk SSSR 1969,189,1302.
Spivey, J. J. Complete Catalytic Oxidation of Volatile Organics. Znd.
Eng. Chem. Res. 1987,26,2165-2180.
Suppiah, S.;Chuang, K. T.; Rolston, J. H. Diffusional Effects in
Wetproofed Catalysta for Isotopic Exchange Between Hydrogen
Gas and Water Vapor. Can. J. Chem. Eng. 1987,65, 256-263.
Tretyakov, I. 1.; Sklyarov, A. V.; Shub, B. R. Kinet. Katal. 1971,12,
996.
Vassileva, M.; Andrew, A.; Dancheva, S.; Kilser, N. Complete Catalytic Oxidation of Benzene Over Supported Vanadium Oxides
Modified by Palladium. Appl. Catal. 1989,49,125-141.
Volter, J.; Lietz, G.; Spindler, H.; Lieske, H. Role of Metallic and
Oxidic Platinum in the Catalytic Combustion of n-Heptane. J.
Catal. 1987,104, 375-380.
Received for review March 4,1992
Revised manuscript received July 9,1992
Accepted August 4,1992

Kinetics of Rhodium-Catalyzed Methanol Carbonylation


Lech Nowicki, Stanislaw Ledakowicz,* and Roman Zarzycki
Faculty of Process and Environmental Engineering, L6di Technical University, W61czahka 175,
90-924 U d i , Poland

The kinetic data of the rhodium-catalyzed carbonylation of methanol have been critically analyzed.
It has been found that the discrepancy between the reaction rate constants is due to ignoring the
effect of vapor-liquid equilibrium of the iodide promoter. The distribution coefficients describing
the vapor-liquid equilibrium of CH31 were determined and employed in the recalculation of the
reaction rate data. A very close agreement of our corrected rate constanta with those obtained from
the literature was achieved.
Introduction
A common and easy way to run a kinetic study or check
the activity of the employed catalyst in a gas-liquid (or
gas-liquid-solid) system is to follow the pressure drop of
the gaseous reactant over the liquid (or slurry) phase. A
set of such experiments taken over suitable ranges of
temperature and reactant concentrations forms the database from which the kinetic model can be also derived.
The problem is that sometimes the model will be seriously
in error if the gas-liquid or vapowliquid equilibrium of
the system is not taken into account in a proper way. We
encountered such a situation when investigating the
well-known process of methanol carbonylation.
The carbonylation of methanol using a homogeneous
rhodium catalyst provided an effective route for the
manufacture of acetic acid (Eby and Singleton, 1983).

* To whom correspondence should be addressed.

Many investigations on this subiect have been reDorted


in the literature (Paulik and Rot&196&Roth et al.; 1971;
Forster, 1976,1979;Brodzki et al., 1976;Hjortkjaer and
Jensen, 1976,1977;Smith et al., 1987;Dake et al., 1989).
The kinetics of the reaction has been studied by Roth et
al. (1971)and Hjortkjaer and Jensen (1976,1977).The
experimental rate expression developed for the reaction
confirms first order with respect to the catalyst and the
iodide promoter. There waa no direct effect of varying
methanol concentration above 0.5 kmol/m3 and partial
pressure of carbon monoxide above 0.2 MPa on the reaction rates. Moreover, the iodide compound and the catalyst can be charged to the reaction mixture in different
forms without changing the rate of methanol carbonylation. In these studies kinetics of the reaction was determined by measuring the pressure drop of carbon monoxide
in time.
Recently, Dake et al. (1989)proposed new rate equations
for the methanol carbonylation in aqueous and acetic acid

0888-5885/92/2631-2472$03.00/00 1992 American Chemical Society

Ind. Eng. Chem. Res., Vol. 31, No. 11, 1992 2473
Table 11. Exmrimental Data of Methanol Carbonvlation
[CHaIlo, 1@[Rhlo,
[CHJ],, l@r, kmol/
run no. kmol/m3 kmol/m3 T,K KI kmol/ma
(ma e)
23
0.406
3.28
456 0.92 0.211
0.585
24
465 1.02 0.196
3.21
0.397
0.781
26
3.30
456 0.92 0.165
0.316
0.393
27
3.29
456 0.92 0.184
0.353
0.522
28
3.36
456 0.92 0.171
0.327
0.472
0.332
3.36
443 0.77 0.187
29
0.292
0.323
0.707
469 1.06 0.156
3.24
30
3.31
456 0.92 0.171
31
0.328
0.467
32
0.398
0.622
3.28
456 0.92 0.207
3.15
456 0.92 0.319
33
0.613
0.922
456 0.92 0.258
3.22
34
0.496
0.773
456 0.92 0.210
4.26
35
0.403
0.805
36
0.398
0.442
2.33
466 0.92 0.207
0.561
0.594
3.22
448 0.83 0.306
51
3.16
457 0.93 0.287
52
0.555
0.829
53
0.556
0.667
3.18
453 0.88 0.295
0.551
1.062
54
4.90
456 0.92 0.287
0.555
1.111
55
4.88
456 0.92 0.289
1.132
57
4.88
453 0.88 0.295
0.556
3.20
449 0.84 0.304
0.538
0.617
71
0.546
0.689
74
3.09
453 0.88 0.285
~

2 10

230

220

2 40

1mn. 1K

Figure 1. Comparison of ow results with the literature data in


Arrheniue diagram: (-- -) Hjortkjaer and Jensen (1976);(+) ow raw
data correlated by eq 1; (A) our recalculated data correlated by eq
2; (V)Smith et al. (1987); (A) Dake et al. (1989).

Table I. Owrating Conditionr for Methanol Carbonvlation


range
o p e r a t a parameter
443-470
temperature, K
0.00234.0043"
catalyst, kmol/m3
0.324.61"
methyl iodide, kmol/m3
1.70-3.2"
methanol, kmol/ma
8.1"
water, kmol/m3
2.2-5.0
preseure, MPa

"Based on the amount introduced to the reactor.

media. In their experimenta a reservoir of CO was used


and the reaction rate was determined by measuring CO
uptake at a constant CO pressure in the reactor. The
reactant concentrations in the liquid phase were measured
by gas chromatography. The authors confirmed agreement
of the reaction orders with t h w reported in the literature.
However, they did not point to the significant differences
in the values of preexponential factors in the kinetics
equations presented by them and by Hjortkjaer and Jeneen
(1976,1977).Figure 1 shows the dependence of reaction
rate constanta on temperature (Arrhenius diagram) for
both seta of experimental data (additionally, one experimental point reported by Smith et al. (1987)for a similar
water concentration is also presented). As can be seen
from this figure, the k values calculated from original data
presented by Dake et al. (1989)are more than 3 times
higher than those of Hjortkjaer and Jensen (1976). Thus,
the questions arise which kinetic data are correct, and what
ia the reason of the differences. The purpose of this paper
was to anewer these questions on the basis of our experimental data.

Exprimental Section
Carbonylation reactions were carried out batchwise in
a mechanically stirred 2W-cm3stainless steel autoclave.
The reactor was maintained at constant temperature by
using an electrical heating bath outside the reactor. The
reactor was equipped with a magnetically driven stirrer,
an automatic temperature control, a gas feed line, and a
highly resolving pressure sensor coupled with a sensitive
recorder.
The experimental procedure was similar to that described by Dake et al. (1984). We did not analyze samples
of the reaction mixture taken during the run. In order to
determine the product distribution, the gas chromatography analyses of the reaction mixture were performed
aftar completion of the run and cooling of the reactor.
These analyses were ueed for maas balance checking.
Rhodium nitrate as a catalyst and methyl iodide distilled
prior to use as a promoter were employed in the experimenta. The carbon monoxide was from BASF (Germany)

0.00

0.20

0.ao

Mrhn 1cOa CCICENTRATIm. ,

0.60

--

a BO

krnllrn'

Figure 2. Effect of catalyat and promoter concentrations on methanol carbonylation rata (A)catalyst ([CH31]o= 0.55 kmol/m3); (V)
iodide promoter charged to the reactor ([I&], = 3.2 X l0.g km01/m3);
(v)iodide promoter in the liquid phase. Temperature = 466 K.

with a minimum purity of 99.9%. In all runs, 80 w t %


aqueous acetic acid was used as a solvent. The reaction
conditions employed in our study are shown in Table I.
The rate of carbonylation was measured by registering
the pressure drop as a function of time. This pressure drop
is the only correct measure of the rate when no other
reactions producing or consuming gases are run.

Results and Discussion


A total 21 runs were made to study the kinetics of the
carbonylation reaction. The raw data are presented in
Table XI. The material balance based on CO consumption
was found to agree to the extent of 95% on average and
was never worse than 90%.
The carbonylation rate dependence on the rhodium
catalyst and iodide concentrations is shown in Figure 2.
As shown, the reaction rate was a linear function of both
the catalyst and the promoter concentration. Therefore,
it is evident that the reaction rate in the range of the
operating conditions may be expressed by the commonly
accepted equation:
r = k[Rh]o[CHSI]o
(1)
where k is the rate constant that is assumed to follow the
Arrhenius equation
k = ko exp(-E/RTl

2474 Ind. Eng. Chem. Res., Vol. 31, No. 11,1992


Table 111. Kinetio Parameters for Methanol Carbonvlation
Hjortkjaer Smith Dake
this work & Jeneen et al. et al. this work
(1987) (1989) cor data
parameter
raw data
(1976)
IO'kO, m3/(kmol*s) 4.5
3.5
170
158.8
E , kJ/mol
61.2
61.7
62.4
72.3
72.2

A nonlinear least squares regression was used to determine


the best values for ko and E parameters which fit all the
raw data presented in Table 11. The values of kinetic
parameters are shown in Table 111, and agreement of the
identified kinetic equation with experimental data is also
shown in Figure 1.
Table 111shows also a comparison of the preexponential
factors ko and the activation energies E calculated in the
present work with those found in the literature. The activation energy determined in our work is very close to the
valuea given by Hjortkjaer and Jensen (1976). The similar
value of activation energy E = 62.4 kJ/mol was given by
Smith et al. (1987),however, for a much smaller water
concentration-2.3 kmol/m3.
In spite of good agreement of the activation energy, there
is a large discrepancy between the preexponential factors.
As it may be seen from Figure 1, our rate constants are
30% higher than those obtained by Hjortkjaer and Jensen
(19761,but still much lower than those calculated from the
data reported by Dake et al. (1989).
The above calculation of the rate constants was based
on the amount of methyl iodide and catalyst introduced
to the reactor. However, the actual iodide concentration
in the liquid phase is always lower because due to its
volatility part of this compound must exist in the gas phase
over the liquid reaction mixture. This also refers to the
other reactants except for rhodium as a nonvolatile component. CHJ is the only reactant directly affecting the
reaction rate. Upon taking into account the vapol-liquid
equilibrium, the rate equation (1)should be rewritten into
the following form:
r = k[Rh]o[CH31]1

(2)

The relationship between the actual CH31concentration


in the liquid phase [CH3II1and the iodide concentration
calculated on the basis of the amount of CHJ introducted
to the reactor [CH3IlOis as follows:
(3)

After simple rearrangements, the general expression for


the distribution coefficient is given by
(7)

Unfortunately, the activity coefficient y i for the system


is not available in the literature, so additional experiments
were performed to estimate the K value for CH31.
The experiments were carried out in the same reactor
under the operating conditions of methanol carbonylation,
but in the absence of the catalyst. The pressure of the gas
phase in the reactor was measured for the system with and
without CH31 after the vapor-liquid equilibrium was
reached. The difference between these two values may be
assumed to be the partial pressure of CH31pI, provided
that the addition of CH31did not affect the partial pressure
of other components. This assumption seems to be justified as the molar fraction of CHJ was very small, about
1?% The distribution coefficient of CH31may be calculated directly from

The distribution coefficients KI determined for the condition of each run are shown in Table 11.
As expected, the KIvalues are not zero; they are in the
range of unity, which means that the amount of methyl
iodide introduced to the reactor was almost equimolarly
distributed between the gaseous and liquid phases. On the
basis of the determined KIvalues the actual concentrations
of CHJ in the liquid phase were recalculated and presented in Table 2 as [CH31],. It is worth noticing that the
linear dependence of the reaction rate on the actual liquid
concentration of CHJ has not been changed at all as
demonstrated by the upper line in Figure 2. The kinetic
parameters were recalculated, and their new values are
compared with the literature in Table I11 and in Figure
1. Our corrected values of ko and E are very close to those
obtained by Dake et al. (1989).
The experimental points in Figure 1indicate near random error distribution around the upper line representing
the corrected kinetic equation
r = 158.8 X lo6 e~p(-72200/RT)[Rh]~[CH,I]~
(9)

where KI is the distribution coefficient defined by


KI = (ndg/(%)I

(4)

The subscript I was used for convenience,instead of CHJ.


This distribution coefficient describing the vapor-liquid
equilibrium is not strictly a thermodynamic quantity; it
depends on pressure, temperature, composition, and volume of the gas phase. Assuming no chemical reaction in
the gas phase and its ideal behavior, the equation of vapor-liquid equilibrium for each component may be written
as
pi = yip&

or

(5)

Summary
Good agreement of our experimental data calculated
taking into account the vapor-liquid equilibrium with
those of Dake et al. (1989)was found. Thus we can conclude that the application of the simple method for CO
pressure recording in kinetic investigations of the reaction
leads to an efficient and quick estimation of the kinetic
parameters unless the vapor-liquid equilibrium is accounted for in a proper way. This is also an indication that
the vapor-liquid equilibrium cannot be forgotten when
designing gas-liquid reactors.
Acknowledgment

We are grateful to Prof. W.-D. Deckwer (GBF Braunschweig,Germany) for providing the rhodium catalyst and
for encouragement. We also thank the Alexander von
Humboldt Foundation, Germany, for donation of a Setra
pressure sensor and a CO gas cylinder.

Ind. Eng. Chem. Res. 1992,31,2475-2483

Nomenclature
E = activation energy, kJ/mol
k = rate constant, m3/(kmol-s)
ko = preexponential factor, m3/(kmold
Ki = distribution coefficient of ith component
KI= distribution coefficient of CH31
m = number of components
ni = number of moles of component i
p i = partial pressure of component i, MPa
Pi = saturated vapor pressure for pure i, MPa
r = rate of carbonylation, kmol/(m3-s)
R = gas constant, 8.314 J/(mol.K)
T = temperature, K
V , = volume of the gas phase, m3
x i = liquid-phase mole fraction of species i
r i = activity coefficient of species i
Subscripts
g = gas phase
i = species index
j = species index
I = methyl iodide
1 = liquid phase
0 = initial value

Literature Cited
Brodzki, D.; Lecleve, C.; Denise, B.; Pannetier, G. Catalytic Properties of the Noble Metal Complexes of Methanol Carbonylation
in Acetic Acid by Rhodium Compounds. Bull. SOC.Fr. 1976,l-2,
61-65 (in French).

2476

Dake, S. B.; Kolhe, D. S.; Chaudhari, R.V. Carbonylation of Ethanol


Using Homogeneous Rh Complex Catalyst: Kinetic Study. J.
Mol. Catal. 1984,24,98-113.
Dake, S. B.;Jaganthan, R;Chaudhari, R.V. New Trends in the Rate
Behavior of Rhodium-Catalyzed Carbonylation of Methanol. Znd.
Eng. Chem. Res. 1989,28,1107-1110.
Eby, R.T.; Singleton, T. C. Methanol Carbonylation to Acetic Acid.
In Applied Industrial Catalysis; Academic Press: New York,
1983;pp 275-296.
Forster, D. On Mechanism of a Rhodium Complex Catalyzed Carbonylation of Methanol to Acetic Acid. J. Am. Chem. SOC.1976,
98,846-848.
Forster, D. Mechanistic Pathways in the Catalytic Carbonylation of
Methanol by Rhodium and Iridium Complexes. Adv. Organomet.
Chem. 1979,17,255-267.
Hjortkjaer, J.; Jeneen, V. W. Rhodium Complex Catalyzed Methanol
Carbonylation. Znd. Eng. Chem. R o d . Res. Deu. 1976,15,46-49.
Hjortkjaer, J.; Jensen, V. W. Rhodium Complex Catalyzed Methanol
Carbonylation. Effects of Medium and Various Additives. Znd.
Eng. Chem. R o d . Res. Dev. 1977,16,281-285.
Paulik, F. E.; Roth, J. F. Catalysts for the Low Pressure Carbonylation of Methanol to Acetic Acid. J. Chem. SOC.,Chem. Commun. 1968,1578-1582.
Roth, J. F.; Craddock, J. H.; Hershman, A.; Paulik, F. E. Low
Pressure Process for Acetic Acid vis Carbonylation of Methanol.
Chem. Technol. 1971,600403.
Smith, B.L.; Torrence, G.P.; Murphy, M. A.; Aguilo, A. The Rhodium Catalysed Methanol Carbonylation to Acetic Acid at Low
Water Concentrations: The Effects of Iodide and Acetate on
Catalyst Activity and Stability. J. Mol. Catal. 1987,39,116136.
Received for review February 25, 1992
Revised manuscript received August 18, 1992
Accepted August 31, 1992

Ceric Sulfate Oxidation of p -Methoxytoluene: Kinetics and Reaction


Results
Theodore Tzedakis and Andre J. Savall*
Laboratoire de G6nie chimique et Electrochimie, URA 192 CNRS, Universit6 Paul Sabatier, 118 Route de
Narbonne, 31 062 Toulouse, France

The chemical oxidation in the indirect electrochemical process for oxidizing 4-methylanisole (MA)
by a sulfuric aqueous solution of ceric sulfate was studied. The dissolution of MA into the aqueous
phase precedes the pure homogeneous oxidation consecutive steps: Cmethylanisole anise alcohol
anisaldehyde anisic acid. The kinetic parameters for each of the steps were deduced from
the measurement of the rate of reaction by absorption spectroscopy. The MA oxidation in the
two-phase medium shows that the aldehyde yield reachea 80% at 50 O C when the initial concentration
of cerium(1V) is less then 40 mol m-3. Anise alcohol may be obtained with a selectivity as high as
50%. The alcohol yield decreases to 15% for a conversion of MA of 8.5%. It is shown that it is
possible to achieve a MA conversion, in a liquid-liquid CSTR fed with a 300 mol m-3 C e ( W solution,
of more than 50% with an aldehyde selectivity of 75 ?%

I. Introduction
The indirect oxidation of substituted toluenes into aldehydes by electrochemicalmeans may be achieved by the
cations, at the highest valency, of redox couples such as:
Co(II)/Co(III), Mn(II)/Mn(III), Ce(III)/Ce(IV). Numerous authors present the Ce(III)/Ce(IV) couple as a
mediator providing valid results concerning the selectivity
of the oxidation reaction of the methyl group of the toluenes (e.g. Syper, 1966;Ho,1973;Ibl et al., 1979; Kramer
et al., 1980; Torii et al., 1982;Kreysa and Medin, 1986;
Kreh et al., 1987;Wendt and Schneider, 1986).

* Author to whom correspondence should be addressed.

The Mn3+ion in sulfuric acid has been used (Wendt and


Schneider, 1986)for two-phase medium oxidation of substituted toluenes: p-chlorotoluene, p-xylene, and pnitrotoluene, for example. By operating in continuous
mode, these authors were able to extract the aqueous phase
and distill the organic phase. Mn3+regeneration yields
obtained on a polished platinum electrode bordered on
95%. A complete kinetic study was also performed and
assessed in order to anticipate the temporal distribution
of oxidation products.
Ceric sulfate was used in sulfuric aqueous solution to
oxidize 4-methylanisole (MA) in solution in methylene
chloride (Kreysa and Medin, 1986). These authors give
an overall p-anisaldehyde (AA)yield of approximately 90%

0888-5885/92/2G31-2475$03.00/00 1992 American Chemical Society

S-ar putea să vă placă și