Sulfur Poisoning of Iron Ammonia Catalyst Probed by ... - Springer Link

Jointly published by Akadémiai Kiadó, Budapest and Kluwer Academic Publishers, Dordrecht

React.Kinet.Catal.Lett. Vol. 74, No. 1, 143-149 (2001)

RKCL3864 SULFUR POISONING OF IRON AMMONIA CATALYST PROBED BY POTASSIUM DESORPTION Andrzej Kotarbaa*, Jaromir Dmytrzyka, Urszula Narkiewiczb and Andrzej

%DUDVNLc

a

Faculty of Chemistry, Jagiellonian University, Ingardena 3, 30-060 Krakow, Poland, b Technical University of Szczecin, 3XáDVNLHJR ± 3RODQG c Regional Laboratory of Physicochemical Analysis and Structural Research, Jagiellonian University, Ingardena 3, 30-060 Krakow, Poland Received February 1, 2001 In revised form July 4, 2001 Accepted July 16, 2001

Abstract The surface of an unpoisoned and sulfur-poisoned industrial iron ammonia catalysts is investigated by K, K+ thermal desorption. The K+ desorption energy increases while the K energy decreases upon poisoning. Presence of sulfur also suppresses the potassium desorption in electronically excited states. Keywords: Potassium desorption, iron catalyst, sulfur poisoning

INTRODUCTION Alkali promoter additives are used to improve the catalyst performance in many heterogeneous catalysts. For example, the ammonia synthesis iron catalyst is promoted efficiently by potassium [1]. The chemical state of the K promoter in the catalyst has been the subject of a number of investigations [1,2,3]. However, the role of potassium promotion in ammonia synthesis is still debated. __________________________________________ *

Corresponding author: Fax No. +48 12 6340515, E-mail: [email protected] 0133-1736/2001/US$ 12.00. © Akadémiai Kiadó, Budapest. All rights reserved.

144

KOTARBA et al.: POTASSIUM DESORPTION

The new approach of investigating the behavior of alkali metals at catalytic surfaces consists in the application of a species resolved (parallel detection of alkali atoms, ions and excited states) thermal alkali desorption method (SR-TAD). It has proved its usefulness applied in alkali promoter surface states investigations in several commercial and model catalysts. Since the K, K+ desorption activation energies can serve as fingerprints, they were used for characterization of potassium states in iron ammonia catalyst. The modification of the catalyst surface during catalyst preparation and industrial usage reveals large changes in potassium ionic flux desorption parameters [4]. The deactivation by industrial exploitation [5] and water treatment [6] was followed by changes in the activation energies of potassium atoms as well as the changes of the work function value. Sulfur is strongly chemisorbed by the iron surface, which makes it the most common permanent poison of ammonia synthesis catalysts [7]. In general, the contamination of fused iron catalysts by sulfur can occur either during catalyst preparation or in the reactor under the operating conditions. Since the catalyst precursor is often an iron ore with promoter additives, natural contamination by sulfates is always present [8]. Therefore, the simplest way to mimic such contamination without introducing additional cations is that with sulfuric acid. In the industrial reactor the catalyst is the subject of the hazard of sulfur impurities contained in the gas feed. This can be achieved in model investigations by introducing a controlled amount of SO2 into the ammonia synthesis gas. Since the chemical status induced by sulfur contaminants and its impact on the catalyst may be different in both cases, this problem has to be elucidated. The addition of potassium decreases the harmful effect of sulfur [9]. In the present study changes in the iron catalyst surface induced by sulfur poisoning were investigated by thermal desorption of potassium ions (K+), atoms (K) and electronically excited states (K*). EXPERIMENTAL The samples used in the present study were commercial fused iron catalysts for ammonia synthesis KM–I and BASF. Both catalysts were investigated in unpoisoned and sulfur-poisoned states. The sulfur was introduced in two different ways. In the case of KM–I sulfur was added by liquid H2SO4 impregnation into its precursor. After that the sample was polythermally reduced under atmospheric pressure in the flow of a N2:3H2 mixture and then catalytic tests were performed. The BASF catalyst was first reduced (under the same conditions as KM–I) and then poisoned by introduction of SO2 into the reactor with the mixture of N2:3H2.


145

The activity tests in ammonia synthesis were performed in a plug flow isothermal integral reactor in the temperature range 350–500°C at a pressure of 10 MPa with space velocity of 25,000 h-1. The concentration of ammonia was measured acidimetrically and by interferometry. As a measure of activity the reaction rate constant k2 from the Temkin–Pyzhev equation was used (α = 0.5) [9]. The potassium thermal desorption experiments were carried out in a vacuum apparatus with a background pressure of 10-7 mbar. The samples were heated from room temperature to 650°C. The only ions leaving the catalyst surface are K+, as revealed by QMS screening. Since the sample was held at a positive potential of 100 V, K+ formed at the surface was accelerated by the applied electric field towards the collector. The K, K+ and K* fluxes were measured, respectively, by alkali surface ionization detector [5], directly as an ionic current, and by field ionization detector [10]. Since the work function of the iron catalyst surface can be easily determined [4] using the Richardson–Duschman method [11], thermionic electron emission from the sample was measured in the temperature range of 450 – 700°C with a negative voltage of 400 V applied to the sample. RESULTS AND DISCUSSION The results of potassium desorption from KM-I and BASF catalysts for both unpoisoned and sulfur-poisoned, plotted in Arrhenius-like co-ordinates are presented in Figs 1 and 2 for atoms and ions, respectively. Since the plots exhibit a linear character (correlation coefficients higher than 0.995) the potassium ions and neutrals activation energies for desorption can be reliably determined. Their values are listed in Table 1. Table 1 Desorption activation energies (eV) for atoms E(K) and ions E(K+) and work function values (eV) from unpoisoned and sulfur-poisoned iron catalysts Catalyst KM-I

E(K) E(K+) Φ

BASF

unpoisoned

270 ppm S

unpoisoned

2600 ppm S

2.55±0.01 2.54±0.02 ~4.0

0.95±0.01 3.32±0.02 3.3

2.22±0.01 2.21±0.02 —

2.06±0.01 3.03±0.02 —

146


•

Fig. 1. Arrhenius plots for potassium atoms for unpoisoned ( ) and sulfurpoisoned ( ) (KM–I — 270 ppm, BASF — 2600 ppm) iron ammonia catalysts

{

Comparison of the data from Table 1 shows that the activation energies for atomic and ionic fluxes are equal within the experimental error for both unpoisoned catalysts. This means that the K+ and K desorption takes place from the same surface state. Sulfur poisoning drastically changes the values of K and K+ desorption activation energies for both catalysts. The most spectacular changes were observed in the case of atoms, where the decrease of desorption energy approaches 1.6 eV (for KM-I catalyst). For potassium ions the energy increases by ~ 0.8 eV upon poisoning. Since the emission of highly excited potassium species (K*) from the industrial iron ammonia catalyst were experimentally proved [12], we checked an influence of sulfur poisoning on the K* emission. The amount of 270 ppm of sulfur causes the excited states signal to decrease approximately by a factor of three in case of KM-I catalyst. Ten times greater amount of sulfur added to BASF catalyst causes this signal to drop to zero (Fig. 3). Additionally, from the measurements of the thermoemission of electrons from KM-I catalyst surface it was found that sulfur decreases the work function value (see Table 1). This fact


147

can be related to the exchange of oxygen with the less electronegative sulfur, since a less polarized bond between K and S favors desorption of K mainly as atoms. According to the Saha-Langmuir equation [11], the potassium desorption from iron catalyst is dominated by an atomic flux because the catalyst work function (Table 1) is lower than the K ionization potential (4.3 eV). Thus sulfur poisoning is apparently responsible for the lower stability of the potassium promoter modifying the electronic properties of the catalyst.

•

Fig. 2. Arrhenius plots for potassium ions for unpoisoned ( ) and sulfurpoisoned ( ) (KM–I — 270 ppm, BASF — 2600 ppm) iron ammonia catalysts

{

The changes in the catalyst surface induces by sulfur are reflected not only in potassium desorption but also in catalytic reactivity. The ammonia synthesis tests revealed that sulfur poisoning obviously decreases the activity by 30 % and 70 % for KM–I and BASF, respectively. In the industrial reactor for ammonia synthesis the first layers of the catalyst bed are the most exposed to a hazard of sulfur poisoning. Indeed the K and K+ desorption energies for the spent catalyst taken from different heights of the 15metre-long ammonia reactor confirm this fact [13]. The lowest stability of potassium promoter was observed for the sample taken from the top of the catalyst bed.

148


Fig. 3. The comparison of the variation of the excited potassium species signal with sample voltage for BASF catalysts. According to [14] the two maxima can be assigned to potassium clusters (K*n) at ~ 9 V and isolated atoms (K*) — at ~ 26 V

To conclude, the species resolved potassium thermal desorption can be a useful tool to evaluate the sulfur poisoning of industrial iron ammonia catalysts. The sulfur chemisorption reduces potassium surface stability (interestingly, the impact of H2SO4 on KM–I precursor is greater than SO2 on BASF). The decrease in catalysts’ activity upon poisoning is accompanied by suppression of the emission of highly excited potassium states. This is in-line with earlier


149

studies [10] that such states can serve as indicators of the actual potassium promoting efficiency in iron catalyst.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

P. Stolze: Ammonia Catalysis and Manufacture, Chapter 2, Springer–Verlag, Berlin 1995. M. Bowker: The Chemical Physics of Solid Surfaces, Vol. 6, chapter 7. Elsevier, Amsterdam 1993. R. Schlögl: Catalytic Ammonia Synthesis, Ch. 2. Plenum Press, New York 1991. A. Kotarba, K. Engvall, M. Hagström, J.B.C. Pettersson: React. Kinet. Catal. Lett., 63, 219 (1998). K. Engvall, L. Holmlid, A. Kotarba, J.B.C. Pettersson, P.G. Menon, P. Skaugset: Appl. Catal. A, 134, 239 (1996). A. %DUDVNL 5 Dziembaj, A. Kotarba, A. *Rá ELRZVNL = Janecki, J.B.C. Pettersson: Studies in Surface Science and Catalysis, 126, 229 (1999). P.E. Højlund Nielsen: Catalytic Ammonia Synthesis, p. 288. Plenum Press, New York 1991. V.I. Sharypov, B.N. Kuznetsov, N.G. Beregovtsova, O.L. Reshetnikov, S.V. Baryshnikov: Fuel, vol. 75, 1, 39 (1996) J. Bernard, J. Oudar, N. Barbouth, E. Margot, Y. Berthier: Surf. Sci., 88, L35 (1979). L.D. Kuznetsov, L.M. Dmitrenko, P.D. Rabina, Ju.A. Sokolinskii: Sintiez ammiaka, Khimiya, Moskva 1982. K. Engvall, A. Kotarba, L. Holmlid: Catal. Lett., 26, 101 (1994). G.A. Somorjai: Surface Chemistry and Catalysis, Ch. 5. John Wiley & Sons, Inc., New York 1994. K. Engvall, A. Kotarba, L. Holmlid: J. Catal., 181, 256 (1999). J. Dmytrzyk, A. %DUDVNL : Arabczyk, A. Kotarba: Proc. of International Symposium Catalysis in XXI Century, p. 51, Krakow, Poland, 4-7.05.2000. J. Wang, K. Engvall, L. Holmlid: J. Chem. Phys., 110, 1212 (1999).