Catalytically active centres in H2S +02 reaction on faujasites Maria Zidtek and Zdzistaw Dudzik Laboratory of Zeolite Catalysis, Institute of Chemistry, A. Michiewicz University, 60- 780 Poznarl, Poland (Received 9 December 1980; revised 10 March 1981)

The influence of the basicity and acidity of faujasite-type zeolites on their catalytic activity in the H2S + 02 reaction was investigated. The catalytic activity increases with the increase in the zeolite basicity. Br6nsted-type acidity retarded the catalyst activity. Oxygen anions bound to aluminium cations in the zeolite framework play the role of catalytically active basic sites. In the sodium form series (Na-X, Na-Y, Na-Y deal.) the direct correlation between the number of alumina tetrahedra per unit celt and the catalytic activity was observed. The catalytic activity of faujasites modified with K +, Li +, and Ca 2+ cations was found to be inversely proportional to the electrostatic potential of the exchanging cation.

INTRODUCTION

In recent publications concerning the catalytical oxidation of hydrogen sulfide with molecular oxygen, the question of the active centres involved in this reaction was not clarified 1-9. Therefore we studied in detail the influence of the basic sites, acidic sites and the cations on the catalytic activity of the synthetic faujasite-type zeolites.

EXPERIMENTAL

Catalysts Zeolites Na-X and Na-Y were used as the parent material. Na-X was Linde 13X Lot No. 2098760 with initial composit ion Na87(A187Si10s)-FAU, and Na-Y was Linde SK-40 Lot No. 3606411 with initial composit ion NasT(AIsTSilas)-FAU. Before use, the samples were treated twice with 1 M solution of NaC1 to remove cations other than Na +, and to bring the A1/Na ratio to equilibrium. Modified forms were prepared using 0.25 M aqueous solutions of appropriate chlorides at room tempeature. The following catalysts were

e x e x prepared: K40Na41-X, Li~Na4s-X, Ca12Nas2-X, ex .~× y (NH~ x)s2Naas-X, K3sNa22-Y, L13sNa22 - ,

Ca~gNa26-Y, (NH]X)29Na28-Y.

The chemical composit ion of the parent sodium forms, exchanged zeolites and their extent of exchange were determined by chemical analysis: SiO 2 gravimetrically, aluminium by EDTA titration, Li, Na and K by atomic absorption spectroscopy (Unicam SP-90 instrument), and Ca content by a volumetric technique.

Aluminium removed from the zeolite during the ion-exchange process was estimated by EDTA titration of the evaporated solutions.

Hydrogen forms of Na-X and Na-Y zeolites were obtained by calcination of the ammonium forms in the shallow bed at 673 K for 4 h, in a flow of pure, dry nitrogen. Dealuminated Na-Y zeolite was prepared by slowly adding EDTA to the zeolite

suspension at the temperature of a boiling water bath 1°. The sample obtained contained 48 A1 atoms per unit cell. In the catalytic experiments the zeolite crystallites were tableted under pressure, ground, and sieved to the 0.5-1 mm diameter range.

For X-ray structural investigations, an HZG-3 diffractometer was employed. The percentage of the zeolite structure preserved was estimated by comparison of peak intensities.

Catalytic experiments The continuous flow technique was used for measurements of the catalytic activity. The apparatus has been described previously 6. 1.1 g portions of zeolite in the dehydrated form were used. The catalyst was activated before reaction in flowing purified nitrogen at 623 K for 4 h.

A 0.012 m 3 h -1 mixture containing Fluka research grade H2S 0.5% (v/v), 02 3.75% (v/v), and nitrogen as carrier gas was passed through the catalyst bed to a gas chromatographic assembly. A 2 m column containing Porapak Q at 253 K served to separate the products.

The catalytic experiments were conducted mainly at 343 K, at the reagents contact time 2 × 10-4h. Our former experiments 11 indicated that over Na-X zeolite, the H2S + 02 reaction proceeds more efficiently at such short contact times at constant catalyst loading. At the H20 + 0 2 reaction temperatures higher than 363 K the dealumination of the Na-X zeolite occurs s.

For the i.r. investigations in the region 300-1200 cm -1 all samples were hydrated by equilibrating them over saturated NH4C1 solution. The pellets contained 1 mg of hydrated zeolite and 200 mg of KBr.

Only the measurements in the 3000-4000 cm -1 region were conducted in the cell connected to a vacuum system. Pressed zeolite wafers had an

0144-2449/81/020117-05502.00 © 1981 IPC Business Press ZEOLITES, 1981, Vol 1, July 117

Active centres in the H2S + 0 2 reaction: M. Zi&lek and Z. Dudzik

average weight of 0.04 mg mm-L These wafers were dehydrated under vacuum (~ 10 .3 Pa), at 623 K for 4 h.

A Perkin-Elmer 180 i.r. spectrometer was employed.

R E S U L T S

Influence of Br6nsted acidity o n the catalytic act iv i ty The curves showing the H2S conversion at 343 K over Na-X, Na-Y faujasites, and their hydrogen forms vs. reaction time are presented in Figure 1. Zeolite Na-X shows the highest catalytic activity (curve a). At the stationary state the catalytic activity of the zeolites Na-Y and (H, Na)-X is practically the same (curves c and b).

Considerable differences appeared in the catalytic activity of these zeolites at the initial stage of the H2S + C2 reaction. Na-Y zeolite also has a relatively long period of 100% H2S conversion (1 h). On hydrogen forms of X and Y zeolites the 100% conversion of H2S at 343 K does not occur. The X-ray structural investigations indicated that the crystallinity of (H, Na)-Y zeolite was practically unchanged. The degree of crystallinity of (H, Na)-X zeolite, measured after catalytic reaction, was still substantial (>60%).

At 458 K the difference between the catalytic activity of Na-X and Na-Y zeolites is much smaller. At the initial stage of the reaction the conversion of HzS is slightly higher over Na-X, and at the stationary state is higher over Na-Y (38% for Na-Y and 30% for Na-X). We found previously that in the course of H2S + 02 reaction zeolite

9ci\ / ,,

/ o ~%5C "1- ~6

o ,-o- A

\ \

-- e _~ " "~.r- z~-~ z~--~C

I00 200 300 t (min)

Fi~lure 1 Catalyt ic act iv i ty o f sodium and hydrogen forms of X and Y zeolites in H2S + O~ reaction at 343 K. (A) Na-X; (B), (H, Na) -X ; (C); Na-Y, (D), (H. Na) -Y

4 0 0

5 0 0

4 0 0

,~ 3 0 0 x

2OO

IOO

\ \

\

N\ \

\ \

\

N

, I, [, K I Ne Li 2 Ce

_0 ¥

Figure 2 The inf luence o f cations electrostatic potent ial on the catalyt ic act ivi ty.(~s~D--) zeol i te X; ( - -x - - - -x - - ) zeol i te Y

Na-X, serving as a catalyst is substantially de- aluminated, and Na-Y remains unchanged 8.

Catalytic activity o f faujas i te- type zeolites m o d i f i e d wi th Li ÷, K ÷, and Ca 2+ The aim of these experiments was to investigate the influence of the zeolite cations polarizing power, which is a function of their electrostatic potential, on the catalytic activity of the modified zeolite. This proper ty increases in the sequence: K < N a < L i < C a .

In Figure 2 the amount of H2S converted during 6 h of reaction was plot ted as a function of the electrostatic potential (e/r) of the cations present in the faujasite structure. All modified type X zeolites demonstrated higher catalytic activity than the respective Y forms.

In the sequence of type Y zeolites the introduction of a cation with higher electrostatic potential causes the decrease in catalytic activity. In the case of type X zeolites the trend is the same with one exception: Na-X zeolite possesses the highest catalytic activity.

Influence o f the alumina tetrahedra The catalytic properties of faujasite-type zeolites with different numbers of alumina tetrahedra in the unit cell (87, 57, and 48) were investigated. The results are presented in Figure 3. At 343 K the catalytic activity of faujasite-type zeolites in the H2S + 02 reaction depends markedly on the Al/Si ratio in the framework.

On the graph (Figure 4), presenting the % of H2S converted at the stationary state vs. number of

118 ZEOLITE& 1981, Vol 1, July

90

80

70

| § 1 o ~ 4 0

30

100

-I \ \

\

1 \ I \ I \ I \ I \ I

c - o A

zk.

I ~ x " x , .

tO - ~ . . x / . . "x-- - x - - x C

I I I I I00 200 300 400

t (rnin)

Figure 3 Catalytic activity of sodium forms of faujasite-type zaolites with dif ferent Si/AI ratio, at 343 K. (A), Na-X; (B), Na-Y; (C), Na-Ydeal.)

913

30

8C

70

6c

5c

% -r

3O ~6

20

Number of (At04)- per unit cell

Correlation between the number of alumina tetrahedra in F i g u r e 4

100

the sodium faujasites unit cell and catalytic activity of the zeolites

alumina tetrahedra per unit cell, the points fall on a straight line, indicating a direct correlation between the catalytic activity of sodium faujasites in the H2S + 02 reaction and the number of alumina tetrahedra in the unit cell.

I.r. s t u d i e s The spectroscopic investigations in the region of 3000-4000 cm -1 indicated that the initial acidity of the catalyst used (hydroxyl band ~3580 and 3660 cm -1) was negligible. The spectrum of dealuminated Na-Y demonstrated the absence of

Active centres in the H2S + 02 reaction: M. Zi~tek and Z. Oudzik

acidic hydroxyl bands. Only on the spectrum of (Ca ex, Na)-X zeolite was the 3660 cm -1 band quite distinct.

The ion-exchange, conducted under the con- ditions used, caused partial dealumination of the faujasite structure. Using the method described previously 7 we estimated the number of alumina tetrahedra in the framework unit cell. Aluminium cations, removed from the zeolite framework during the ion-exchange process, may occupy cationic sites in the form of AP +, AI(OH) 2+, AI(OH)~, A10 +, or may pass to the solution. Both phenomena were observed. Chemical analysis proved that aluminium was present in the ion- exchange solution. The presence of aluminium cations in the cationic sites was confirmed by the i.r. spectroscopy method. A new band (~600 cm -1) appeared due to the interaction between alumina tetrahedra of the zeolite structure and A1 a+ (ref 12). Using chemical analysis alone it is possible to estimate only the total amount of aluminium present in the zeolite, i.e. in the zeolite framework and in cationic sites. The degree of framework dealumination of modified X and Y zeolites estimated from the i.r. spectra are presented in Table 1.

The X-ray structural investigations of ion- exchanged samples demonstrated that only in the case of Ca ex, Na-X partical collapse (~20%) of the faujasite structure occurred during the ion- exchange process. The framework of the remaining cation-modified zeolites had not changed signifi- cantly, the differences in the diffraction spectra being within experimental error.

DISCUSSION

The experimental results indiate that Br6nsted- type acidity retarded the catalytic activity of faujasite-type zeolites in the H2S + 02 reaction. The hydrogen forms of type X and Y zeolites were less active than their sodium forms, especially in the first period of the reaction (Figure 1). On the basis of these results it is possible to conclude that the basic sites in the faujasite-type zeolites play the dominant role as the catalytically active centres in the H2S + 02 reaction. The basicity is caused by the presence of the negatively charged alumina- oxygen tetrahedra in the zeolite framework. Taking into account the size of the aluminium cation and the oxygen anion it seems probable that

Table 1 Degree of f ramework dealumination after ion-exch.ange

Zeolite Kex, Liex, CaeX, H e x Na-X Na-X Na-X Na-X

Amount of AI in the frame- 83 77 77 68 work (per unit cell)

Zeolite K ex, Li ex, Ca ex, H ex, Na-Y Na-Y Na-Y Na-Y Na-Y deal.

Amount of AI in the 57 54 51 48 48 framework (per unit cell)

ZEOLITE& 1981, Vol 1, July 119

Active centres in the H2S + 0 2 reaction: M. Zi~gek and Z. Dudzik

oxygen anions bound to aluminium cations play the role of catalytically active basic sites. The role of alumina tetrahedra was discussed by Yashima et al. la and recently by Ono in his review paper 14. The catalytic activity in the H2S + 02 reaction of faujasite-type zeolites having the same cations depends predominantly on the number of alumina tetrahedra in the unit cell. Such direct correlation was observed in the case of sodium forms (Figure 4). It should be stressed that such distinct correla- tion exists only in the region of low reaction temperatures (273-343 K). At the stationary state of reaction conducted at 458 K the H2S con- version over Na-X zeolite is slightly higher than the conversion over Na-Y zeolite. At least two factors caused the deactivation of the Na-X zeolite during the reaction time:

(1) the dealumination of the Na-X zeolite frame- work

(2) the generation of acidic hydroxyls on the Na-X surface

The quantitative data obtained from the i.r. spectra are subject to experimental error. Therefore, the accuracy of the estimation of framework A1 content may be challenged. The faujasite-type framework dealumination problem is not the main topic of this study and will be discussed in detail in a separate paper zl.

Zhdanov et al. is observed that i.r. vibrations of the dehydrated zeolite framework depend very much on the exchanged cation in it. Contrary to these observations Flanigen et al. 16 found no consider- able changes of the characteristic spectra after cation exchange in a limited number of fully hydrated exchanged forms of type A, X and Y zeolites, and only slight spectral changes after dehydration. Maxwell and Baks iv have shown that in the shifts of i.r. framework vibrations caused by cations, a similar series of changes exists for A and X zeolites. Radak et al. 18 have found similar correlations for monovalent cation-exchanged Y-type zeolites.

In the last two papers 17,18 the biggest shifts were observed for the band in 460 and 570 cm -1 regions. Radak and coworkers investigated the i.r. spectra of cation modified zeolites only in the 200-850 cm -1 region and stated that this spectral range was chosen because any influence of the cations on framework vibrations only appears within these limits.

To estimate the degree of framework dealumina- tion in zeolites X and Y we investigated mainly the changes of the band position in the 1000 cm -1 region, and to some extent in the 700 cm -1 region. The number of A1 cations occupying cationic sites was estimated in some cases, with limited accuracy, by evaluating the intensity of the ~ 6 0 0 cm -x band, which appeared due to the interaction between alumina tetrahedra of the zeolite structure and AP +. The amount of aluminium removed from the framework and transferred to the ion-exchange solution was estimated by chemical analysis.

Recently we performed an experiment which supports our findings concerning the lattice de- alumination during ion-exchange 21. In the Na-X zeolite ~ 2 5 % Na ÷ were exchanged for Li +. On the spectra of the parent and modified zeolite a shift of the band in the 1000 cm -1 region was observed, from 970 cm -1 (Na-X) to 978 cm -1 [(Li ex, Na)-X]. Subsequently Li + was exchanged for Na ÷. The exchanged sample contained only traces of Li, but the 978 cm -1 band remain unchanged. These results indicate that the shift was caused by frame- work dealumination and was not influenced by the change of cations.

In an earlier paper a we reported the dealumination of the Na-X zeolite framework which occurred when the H2S + 02 reaction was conducted over this zeolite. Dealumination of its lattice increased with increasing reaction temperature, and at 458 K ~ 10 alumina tetrahedra per unit cell were removed from the lattice. In the case of Na-Y no noticeable changes in the zeolite lattice were observed after a reaction time of 8 h. Karge and Rasco 9 used infrared spectroscopy for investiga- tions of H2S chemisorption on faujasite-type zeolites. They found that on alumina-rich faujasites ( type X) heterolytic dissociative chemisorption occurs:

H 2 S a d s -+ HS- + H +

H + forming a surface hydroxyl with the lattice oxygen. On Na-Y the bands due to formation of HS- and surface hydroxyls were not observed.

Recently Karge et al. 19 reported that strong bands due to acidic OH groups appeared during the HzS + 0 2 reaction in the spectrum of Na-X, bu t were missing for Na-Y. The retarding effect of the Br/Snsted-type acidity is demonstrated also by Figures 1 and 4. Both Na-Y dealuminated and (H, Na)-Y contain the same number of alumina tetrahedra per unit cell, but the former is more active.

Studies on the influence of the cations on catalytic activity in the HzS + 0 2 reaction are strongly impeded because usually the modified zeolites differ from the parent form. The dealumination occurring during the ion- exchange process was discussed in an earlier paper 7, and the degree of dealumination of the modified samples investigated in this series of experiments is presented in the Table 1.

A second problem is the deviation of the Na+/A104 ratio from unity. Comparisons of analytical data on exchanged samples (see Experimental) with dealumination data (Table 1) shows that (K ex, Na)-X is cation deficient, and (Li ex, Na)-X has excess cations. Such deviations were recently observed by Hardin et al. z0, however these authors did not take into account the framework de- alumination occurring during the ion-exchange process.

Finally the picture is complicated by the

120 ZEOLITE& 1981, Vol 1, July

~OC

90

8O ~0£

~OC

~.ec

3G IOC

7 0

= o"

so i

o"

i E

40

O

\ 1 \ \

\ \

i I i L I h K I I~I I 2¢m

÷ Figure 5 The influence of Si/AI ratio in the faujasites framework and the electrostatic potential of their cations on the catalytic act ivi ty of these zeolites in H=S + 02 reaction. (--o--o--) zealite X; (--x----x--) zeolite Y

preference of the cations for different types of cationic sites. In spite of these complications, the influence of the cations electrostatic potential on the catalytic activity of modified zeolites in the H2S + 02 reaction was demonstrated quite clearly.

In Figure 5 the amount of H2S converted during 6 h of reaction and the number of alumina tetra- hedra in the unit cell were plotted versus the electrostatic potential of the cations present in the zeolite.

Experimental results presented on this figure suggest that the catalytic activity of synthetic faujasites in the hydrogen sulfide oxidation reaction with molecular oxygen depends on at least two factors. The catalytic activity is directly proportional to the number of alumina-oxygen tetrahedra present in the unit cell, and inversely proportional to the electrostatic potential of the cations present on the faujasite framework.

The influence of cations on the catalytic activity may be estimated more precisely for the series of

Active centres in the H2S + 02 reaction: M. ZiSlek and Z. Dudzik

Y-type faujasites because the dealumination of these zeolites, occurring during the ion-exchange process, was negligible. The results indicate that the basic zeolites (potassium and sodium forms) are much more active catalysts in the H2S + 02 reaction than the acidic zeolites (hydrogen and calcium forms). Based on these results it is possible to state that the catalytic activity of faujasite-type zeolites in the hydrogen sulfide oxidation reaction increases with the increase of the zeolite basicity.

ACKNOWLEDGEMENTS

The authors wish to thank Mrs. H. Nowicka for excellent technical assistance.

REFERENCES 1 Dudzik, Z., Bilska, M. and Czeremuzi~ska, J. Bull. Acad.

Polon. ScL, Ser. ScL Chim. 1974, 22, 307 2 Dudzik, Z. and Bilska-Zi6fek, M. Bull. Acad. Polon. ScL, Ser.

ScL Chim. 1975, 23 ,699 3 Steijns, M. andMars, P.J. Catal. 1974,35,11 4 Steijns, M., Derks, F., Verloop, A. and Mars, P. J. Catal. 1976,

42, 87 5 Steijns, M., Koopman, P., Nieuwenhuijse, B. and Mars, P.

J. Catal. 1976, 42, 96 6 Dudzik, Z. and Zi6fek, M.J. CataL 1978, 51 ,345 7 Zi6~ek, M. and Dudzik, Z. React. Kin. CataL Lett. 1979, 12,

213 8 Zi6I-ek, M. and Dudzik, Z. React. Kin. CataL Lett. 1980, 14,

213 9 Karge, H. G. and Rasco, J. J. Colloid Interface ScL 1978, 64,

522 10 Kerr, G. T.J. Phys. Chem. 1968,72,2594 11 Zi6tek, M. and Dudzik, Z. Przem. Chem. in press 12 Chukin, G.D. andSmirnov, B.V. Zh. Fiz. Khim. 1976,50,

141 13 Yashima, T., Suzuki, H. and Hara, N. J. Catal. 174, 33,486 14 Ono, Y. Studies in Surface Science and Catalysis, Vol. 5,

(Ed. B. Imelik) Elsevier, 1980, p. 19 16 Zhdanov, S. P., Kiselev, A. V., Lygin, V. J. and Totova, T. J.

Russ. J. Phys. Chem. 1964, 38, ! 299 16 Flanigen, E. M., Khatami, K. and Szimansky, H. A. Adv.

Chem. Ser. 1971, 101,201 17 Maxwell, l .E. andBaks, A. Adv. Chem. Ser. 1973,121,87 18 Radak, V. M., Cerani~, T. S. and 2~ivadinovid, I. M. Z.

Naturforsch. 1978, 33B, 1116 19 Karge, H. G., Ladebeck, J. and Nag, N. K. Preprint, 7th

Canadian Symposium on Catalysis, Edmonton 1980 20 Hardin, A. H., Klemens, M. and Morrow, B. A. J. CataL 1980,

62,316 21 Zi6~ek, M. and Dudzik, Z. Unpublished results

Z E O L I T E S , 1981, Vo l 1, Ju l y 121