.

~ ~

APPLIED

E LS EV I E R

Applied Catalysis A: General 129 (1995) 57-67

CATALYSIS

A: EEA GNRL

Cubic and hexagonal faujasites with varying Si/A1 ratios II. Catalysis of n-heptane cracking
G. Lischke a, E. Schreier b, B. Parlitz c, I. Pitsch c, U. Lohse c,., M. Woettke"
" Technische Fachhochsehule Wildau, Friedrich-Engels-StraJ3e 63, 15742 Wildau, Germany b Humboldt-Unversitiit Berlin, Hessische Strafle 1, 10115 Berlin, Germany blstitutfiir Angewandte Chemie Berlin-Adlershofe. V., Rudower Chaussee 5, 12484 Berlin, Germany

Received 22 December 1994: revised I 1 April 1995; accepted 12 April 1995

Abstract Faujasite (FAU) and its hexagonal analog (EMT) was investigated in the catalytic cracking of nheptane. Measurements were conducted with Aerosil-200 bound catalysts in an integral fixed bed reactor at 723 K. Catalytic activity is related to the concentration of Br0nsted acid sites of the zeolite framework. Catalytic behaviour of dealuminated and subsequently acid leached varieties evidences that extra-framework aluminium species formed during dealumination contribute to catalysis. The distribution within the spectrum of the products of reaction clearly demonstrates that pathways producing alkenes and free hydrogen are involved in the scission of the parent hydrocarbon.
Keywords: Heptane cracking; Zeolites: Acid sites; Reaction pathways

1. Introduction The nature of active sites that catalyze reactions of saturated hydrocarbons has been the subject of profound discussion in the fields of heterogeneous catalysis and surface science for about half a century. Apart from electron deficient centres (Lewis sites), zeolites contain proton donating hydroxyl groups attached to their frameworks (BrOnsted sites) which can exert catalytic effects [ 1-3]. Thus, it could be shown that the activity of zeolite catalysts in the cracking of nhexane [4], n-heptane [5], and n-octane [6] as well as in the isomerization of various hydrocarbons [7] is clearly related to strong protonic acidity.
* Corresponding author. 0926-860X/95/$09.50 1995 Elsevier Science B.V. All rights reserved SSD109 2 6 - 8 6 0 X ( 95 )00093-3

58

G. Lischke et al./Applied Catalysis A: General 129 (1995) 5 7 4 7

Additionally, it became evident that Lewis sites can also be involved in the catalytic process. Their role, however, is still far from being well understood [810]. With Y zeolites, extra-framework aluminium (EFAL) generated by dealumination has turned out to have its own catalytic activity in the cracking of nheptane and gas oil [ 11,12]. Catalytic studies of ethylbenzene disproportionation [13] gave evidence for a synergistic interaction between BrCnsted sites attached to the framework and Lewis acid sites located on EFAL species. With H-ZSM-20, extra-framework aluminium species of a certain state were presumed to be able to exert inductive effects on structural hydroxyl groups, thus enhancing their acidity [ 14]. On the other hand, cationic EFAL species are able to poison strong acid sites [15]. Recent studies [5,16] demonstrate that products resulting from the cracking of n-heptane on Y zeolite catalysts consist essentially of C3 and C4 compounds. In the cracking of n-octane on H-mordenite [ 17], the scission of the alkane molecule into two parts to produce C3-, C4-and Cs-aliphatics is the dominating primary step. Its relative probability amounts to about 60%. Simultaneously, direct cleavage into three parts occurs. The scission of the parent molecule at the end of the carbon chain ( a and/3 position) appears to be a rather rare event. Like with various other types of hydrocarbon reactions [18-20], hydrogen transfer, which strongly influences the distribution between olefinic, paraffinic and aromatic hydrocarbons, participates in the overall cracking process of saturated hydrocarbons [ 21-23 ]. The intensity of its effect correlates with the acidity of the catalyst and depends on the degree of conversion. The aim of this work is to demonstrate whether and how intensively n-heptane cracking on faujasites is influenced by their physico-chemical properties. Experimental conditions are varied to specify the effect of structure, acidity, and the presence of extra-framework aluminium (EFAL) within the zeolites on their catalytic behaviour (activity, selectivity, stability). To gain information on the initial stages of the reaction, measurements at extremely low degrees of conversions of n-heptane ( 1% and less) were included in the experiments.

2. Experimental
Investigated catalysts are based on two different structures (cubic and hexagonal) of Y zeolites and their hydrothermally dealuminated and acid leached varieties. Acid leaching is a proper means to remove parts of the extra-framework aluminium. Detailed results of physico-chemical characterization of the samples are described previously (part I). Table 1 represents the acidity and pore data of the samples. In the catalytic experiments, cylindrically shaped pellets (diameter 1.5 mm) containing 35 mass-% of zeolite bound with Aerosil-200 were used as catalysts. Samples were dried at 393 K prior to calcination at 823 K. Experiments were performed in the gas phase by using an integral fixed bed reactor made of stainless

G. Lischke et al. /Applied Catalysis A: General 129 (1995) 57-67

Table I Physico-chemical characterization of zeolite samples Sample H EMT D EMT 500 D EMT 500ex D EMT 600 D EMT 600ex D EMT 800 D EMT 800ex H FAU D FAU 500 D FAU 500ex D FAU 600 D FAU 600ex H +/u.c. TPDA 28 5.6 4.3 3 7 1 4,1 27.5 5.6 3.1 3 4.5 Pore volume of micropores (cm 3/g ) 0,30 0.28 0.15 0.28 0.28 0.09 0.15 0.32 0.31 0,11 0.27 0.11

59

Pore volume of mesopores (cm3/g) 0 0 0.11 0.15 0.20 0 0 0,06 0.04

steel with an independently controlled two-zone heater. Standard conditions of the reaction were the following: catalyst volume: 6 ml, temperature: 723 K, concentration of n-heptane in nitrogen: 10 vol.-%, rate of the feed gas: 1.11 ml/s. The mixture of the reactant feed was obtained by passing the nitrogen flow through a n-heptane saturator at 308 K. Conversion of n-heptane was varied by changing the duration, as well as ---deviating from the above standard conditions - - b y the variation of the volume rate of the feed gas. To determine the concentrations of the reactant and of the products of reaction, samples of the reactor outlet gas were injected into an on-line gas chromatograph (GC). The GC device was equipped with two parallel operated columns (packed with Fractonitril on Porapack and Gel II, respectively) and a flame ionization detector coupled to a computing integrator. Off-line GC devices were used to detect and to determine free hydrogen (2 m molecular sieve column at 328 K with helium as carrier gas with a flow rate of 0.4 ml/s) and to analyze the butene fraction (combined column system of di-npropyl-sulfone and kieselgel at room temperature and nitrogen carrier gas with a flow rate of 0.42 ml/s) of the exit gas.

3. Results and discussion

Catalytic cracking of n-heptane on Y zeolites at 723 K was studied by constructing conversion / time on stream (TOS) and molar portion / conversion plots.

3.1. Activity
Conversion/TOS plots (Fig. 1 and Fig. 2) demonstrate that all catalysts undergo rapid deactivation. Losses in the total balance of the gas-chromatographically analyzed substance of the reacted mixture (Table 2) indicate that consid-

60
8O

G. Lischke et al. /Apph'ed Catalysis A: General 129 (19951 57~57

conversion %

a-!

8O

i conversion %

b ~

7O

70i 60: O 50 ',

4o

50 i 40 30
20 10~ i 0 t 0 ,.~.. time-on-stream h 1 2 3 4 5 6 7 8

20 ~

- w'zl)4~t

j.

10 ~ [
0

&IA
....

'. . . . . . . . .

u & A A~lc~k~lc& A ~k~icA A time-on-stream h

~. . . . . . . . . . . . . . .

Fig. 1. Catalytic activity of faujasite (FAU) based catalysts in the cracking of n-heptane in the gas phase at 723 K. Influence of dealumination and extraction of extraframework aluminium (EFAL). (a) (@) H FAU; I + ) D FAU 500; ( ) D FAU 600. (b) ( 0 ) H FAU: ( ) D FAU 500ex; () D FAU 600ex.
80

o conversion Yo 7O iO 60 ',

80

conversion % 70 i l 60 ,~

so

'.

~o
40

: "x 40: 4. . ~ ' , ' ]11, 4-- '

30 "" " , , " .

30

70
0 1 2 3 4 5 6 7 8 . 0~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 1 2 3 4 5 6 7 8

Fig. 2. Catalytic activity of EMT based catalysts in the cracking of n-heptane in the gas phase at 723 K. Influence of dealumination and extraction of extraframework aluminium (EFAL). (a) (O) H EMT: ( + ) D EMT 500; ( ) D EMT 600; (C]) D EMT 800, (b) (O) H EMT; ( ) D EMT 500ex; (C)) D EMT 600ex: ( ) D EMT 800ex.

Table 2 Catalytic cracking of n-heptane on H EMT at 723 K catalyst volume: 7.0 ml (weight: 2,1 g) Selectivities (%) Relative conversion of n-heptane (%) 0.83 Propene Butenes Propane Butanes Sum Methane Cz-aliphatics C~-aliphatics Sum Benzene Toluene Xylenes C9 aromatics Sum Total 36.0 26.4 6.8 22.2 91.4
1.5

1.8 35.6 24.0 10.0 22.4 92.0

1.2

3.5 33.7 20.7 11.7 27.9 94.0

0.9

9.9 26.5 17.7 16.4 31.9 92.5

1.1

13,2 25,4 15.7 19,0 33.5 93,6 0.7 1,9 2.3 4,9 0 0.7 0.1 0 0.8 99,3

17,1 24.5 13.3 21.6 35.2 94.6 0.6 1.8 1.5 3.9 0 0.7 0.2 0 0.9 99.4

19.8 23.4 12.4 22.6 36.3 94.7 0.5 1.7 1.6 3.8 0 0.7 0.2 0.1 1.0 99.5

27.0

32.3

39.0

44.6

48.2 11.4 8,3 31.3 41.1 92. I

1.1

68.2 9.0 7.8 32.9 41.3 91.0

(1.4

2 0 . 1 18.0 10.9 9,9 24.9 26.6 38.7 40.0 94.6 94.5 0.5 1.6 1.8 3.9 0 0.6 0.2 0.1 0.9 99.4 0.5 1.6 1.8 3.9 0 0.6 0.3 0.1 1.0 99.4

1 6 . 1 13.4 9.2 8.5 28.0 30.3 40.9 42.0 94.2 94.2 0.5 1.6 2.0 4.1 0 0.6 0.3 0.1 1,0 99.3 0,5 1.7 2.0 4.2 0 0.5 0.4 0.1 1.0 99.4

2.6 3.2 7.3 0 0.4 0 0 0.4 99.1

2.4 3.2 6.8 0 0.6 0 0 0.6 99.4

2.0 1.8 4.7 0 0.8 0 0 0.8 99.5

2.0 2.9 6.0 0 0.8 0.1 0 0.9 99.4

2.3 2.0 5.4 0.1 1.2 0.6 0.1 2.0 99.5

1.7 1.8 3.9 0 0.5 0.4 0.1 1.0 95.9

G. Lischke et al. / Applied Catalysis A: General 129 (1995) 57-67

61

erable amounts of coke or coke precursors are formed at the high conversion level immediately after the start of the reaction. During this period (the first 15 to 20 min), about 3% of the reacted n-heptane is converted into material deposited on the catalysts. As TOS grows, the portion of non-detected substance sharply decreases. Losses in the GC balance then lie within the margins of error. Concerning the balance of substance and the sums of selectivities of the three groups of chemical compounds presented in Table 2, all investigated Y zeolites show nearly identical behaviour. Spent catalysts can be regenerated by a 2-h treatment in a stream of air (3.6 1/ h) at 723 K. Then, initial activity is re-achieved. For both types of Y zeolites, initial activities of the corresponding catalysts decrease as the temperature of hydrothermal treatment during dealumination is raised. With the varieties dealuminated at 873 or 1073 K, partial extraction of extraframework aluminium (EFAL) results in an increase of the activity. With D EMT 500 and D FAU 500, acid leaching, however, exerts a negative influence: activity values are further decreased. The comparison of reaction data with the results of physico-chemical characterization (part I) reveals a qualitative correlation between initial catalytic activities and the densities of BrCnsted acid sites. The latter are derived from the temperatureprogrammed desorption of ammonia (TPDA) (Table 1 ) and are further evidenced by the changes of the intensities of the IR bands of OH vibration (part I). In any case, a stepwise decrease of the number of protonic sites per unit cell causes a decrease of the initial activity. As was expected, both effects are most pronounced after treatment at 1073 K (D EMT 800). Similarly, changes of the number of the protonic sites observed after the partial extraction of EFAL originating from dealumination give rise to a corresponding change of catalytic activity. In all cases when the number of Br0nsted acid sites after acid leaching is re-increased (D EMT 600ex, D EMT 800ex, D FAU 600ex), the initial catalytic activity increases, too. It is worth mentioning that acid leaching results also in a partial exchange of Na ions present within the zeolite framework. This effect additionally contributes to an increase of BrCnsted acidity and hence of activity. With D EMT 500 and D FAU 500, acid leaching causes a further decrease of catalytic activity. Parallel, BrCnsted acidity is decreased. This is indicated by losses of the volumes of the micropores (Table 1 ) and might be due to partial damages of the zeolite structure. Differences in the porous structure between the two types of Y zeolites are expressed by a higher portion of mesopores found after hydrothermal treatment with the EMT samples (Table 1 ). They are reflected by the generally slightly higher level of catalytic activity. A quantitative comparison of the activity data and the results of physico-chemical characterization of the zeolite samples, however, evidences that the protonic acid sites of the zeolite framework are not the only source of catalytic activity. Apparently, EFAL species are able to contribute considerably to the catalytic activity.

62

G. Lischke et al. /Applied Catalysis A: General 129 (1995) 57~57

This is indicated by the comparatively high catalytic activity of most of the dealuminated and extracted varieties, which generally exhibit substantially lower H + / u.c. values than the parent zeolites (Table 1). The existence of EFAL active sites could explain the remaining catalytic effect of D FAU 500ex, which --according to IR vibrational spectra--has no free bridging hydroxyl groups within the zeolite framework at all.

3.2. Selectivity of products

The conversion of n-heptane is dominated by cracking reactions which give rise to the formation of C3-and C4-aliphatic fragments. Within a broad range of nheptane conversion (Table 2), their total selectivity amounts to about 94%. The formation of other products (CI-, C2-and Cs-aliphatics; C6- and Cg-aromatics) plays a minor role and depends only weakly on the degree of conversion. Somewhat higher selectivities for C5 aliphatics at low conversions may indicate that these compounds are primary products and can undergo further chemical changes as reaction proceeds. C6 aliphatics, heptenes, and isomer heptanes are detected in traces only. Among the aromatic compounds, toluene and --at higher degrees of conversion --the xylenes are dominating. The selectivity values of the C3-and C4-aliphatics (Table 2) clearly show that the ratios of the amounts of saturated and non-saturated compounds characteristically depend on the level of conversion of n-heptane. In the initial stages of the reaction, selectivities of the olefinic hydrocarbons are substantially higher than those of the corresponding saturated compounds. With increasing degree of conversion, the portions of propane and butanes within the product grow. Hydride transfer gains influence: primarily formed C3 and C4 alkenes are transferred into alkanes. At n-heptane conversions of about 10%, alkenes and alkanes are contained in approximately identical amounts in the mixture of the product. At higher levels of conversion, saturated hydrocarbons clearly dominate. These results indicate that primary cracking is not completely described by the scission of the intermediately formed branched CTH~ ions [5] following the conventional stoichiometries:
C7HI6 ~ C3H s q- C4H 8

(1)

and
C7HI6 ~ C4Hloq-C3H6

(2)

producing equimolecular amounts of olefinic and saturated hydrocarbons. Apparently, only pathways comprising the formation of alkenes are involved in the reaction network. The presence of free hydrogen in the gaseous product of the reaction could be proved by off-line gas-chromatographic analysis. Its amount markedly depends on the degree of conversion of n-heptane (Table 3). At extremely low conversions

G. Lischke et al. / Applied Catalysis A: General 129 (1995) 57~5 7 Table 3 Content of free hydrogen in the gaseous product on H EMT, reaction temperature: 723 K Degree of conversion ( % ) 1.8 19.8 48.2 Concentration of hydrogen ( vol.-% ) 0.053 0.173 O. 160

63

(1.8%), the concentration of hydrogen amounts to 0.053 vol.-%. This value is nearly equivalent to the concentrations of propene and butenes (averaged value 0.0474 vol.-%, Table 4) formed concurrently to conventional cracking according to Eqn. 1 and Eqn. 2). This fact is indicative of a 'dehydrogenating cracking', by which scission of the chemisorbed species results in free hydrogen and equivalent molecular amounts of the olefinic fragments, e.g. following the stoichiometry:
C 7 H 1 6 --~ C 4 H 8 + C 3 H r + H 2

(3)

It is known from literature that this kind of fragmentation of the parent molecule can prevail at low conversion levels in cracking processes. On SAPO and HY materials, cracking of n-octane at a conversion of about 2% resulted in high portions of unsaturated aliphatics [24]. More recently published results of heptane cracking in the presence of feed hydrogen [ 25 ] confirm that free hydrogen is able to hydrogenate unsaturated species which are chemisorbed at the surface of the catalyst. Thus, at higher degrees of conversion, hydrogenation of primarily formed C3 and C4 alkenes by free hydrogen can be assumed to be a source of propane and butanes. This reaction is a parallel path to the usual hydrogen transfer accompanied by aromatization and/or coke deposition. As could be shown by off-line GC analysis, the distribution among the butenes is near to thermodynamic equilibrium. At high conversions of n-heptane, observed with the fresh highly active catalyst, however, isobutene markedly prevails. Its concentration reaches about 65% within the butene fraction. Numerical ratios of the two parallel pathways of initial cracking can be gained by extrapolating molar portion/conversion plots for the C3/C 4 aliphatics (Fig. 3) to zero conversion of n-heptane (MPo). Extrapolated MPo values (Table 5) show that stepwise decreasing of the Br~nsted acidity by dealumination at 773 or 873 K exerts weak influences on the distribution among the C3 and Ca aliphatics. Only
Table 4 Concentration of C3-and C4-aliphatics in the gaseous product at the conversion of n-heptane of 1.8% Compound Propene Butenes Propane Butanes Concentration (vol.-%) 0.1225 0.0826 0.0341 0.0762 Alkenes formed excessively to conventional cracking (vol.-%) 0.0463 0.0485

64

G. Lischke et a l . / Applied Catah'sis A." General 129 (1995) 57-67

50 t molar portion %

<,

40 I

~f~tt.~-~

--

40

50 iI molar portion % ~

E]FJ

FJ

[J

:.,

{ ~ ~

*i

3o~

i~, ~

**
H EMT
'Q

~--ml- + + : #
0 0 10 20 30 ...... 40 conversion % i '. . . . . 50 60 70 conversion % 0 . . . . . . . . . . . . . . . 0 20 40 60 80

50

molar portion %

--

, !

50

molar portion %

D
30
20

c~
,,. ,~_ ;5 E] E~

dz

~.

9e
D EMT 500 5~J~r: , ~ e.% D EMT 500ex

"
10

.
conversion %

ii
0 60 __ ~__t . . . . . 10 20

+
conversion % 40 50

10

20

30

40

50

t . 30

50

molar portion %

50 i molar portion %

40 0,

401
D ~[~[t-

~]

E~

[]

'

30 ; ' % 1 20

:{Ye"e:o -Et
i I ~ E ] ~

~ _
D EMT 800

[]

30 20 10 "

]~-:'~F

W,

: ~

s.

D EMT 800ex

10 ~ . 0 . . . . . 10 . . . 20 . . conversion % . . 30 40

-b
10 20 30 40

+
50

+
60 70

i 0

conversion %

Fig. 3. Molar portions of aliphatics within the C~/C4 fraction of products of the cracking of n-heptane obtained with various FAU and EMT catalysts. ( 0 ) Propene; ( + ) butenes; ( ) propane; ([3) butanes. Table 5 Molar portions (%) of C~-and C4-aliphatics formed at zero conversion of n-beptane cracking (Y zeolites at 723 K) MPo

H
FAU

D
FAU 500 39.5 28.0 7.5 25.0 0.89

D
FAU 600 41.0 30.5 7.0 21.5 0.92

D
FAU 500ex 36.0 31.5 13.0 19.5 0.96

D
FAU 600ex 38.0 30.5 8.5 23.0 0.96

H
EMT

D
EMT 500 39.0 31.0 7.5 22.5 0.87

D
EMT 600 40.5 29.0 7.0 23.5 0.90

D
EMT 800 44.0 37.0 6.0 13.0 1.00

D
EMT 500ex 38.0 33.0 9.0 20.0 0.89

D
EMT 600ex 38.0 25.0 8.0 29.0 0.85

D
EMT 800ex 36.0 28.5 10.0 26.0 0.85

C3H~ C4H8 C3H8 C4Hlo Molar ratio C3/C4

42.0 30.0 8.0 20.0 1.00

41.0 31.0 6.0 22.0 0.89

G. Lischke et al. /Applied Catalysis A: General 129 (1995) 57-67

65

Table 6 Contribution ( % ) of normal and dehydrogenating cracking to the primary C3/C4 scission of n-heptane ( Y zeolites at 723 K) H FAU D FAU 500 D FAU 600 D FAU 500ex D FAU 600ex H EMT D EMT 500 D EMT 600 D EMT 800 D EMT 500ex D EMT 600ex DEMT 800ex

Cracking
Normal 56 Dehydrogenating 44 65 35 57 43 65 35 63 37 56 44 60 40 61 39 38 62 58 42 74 26 72 28

after hydrothermal treatment at 1073 K (D EMT 800) when EFAL is present in higher amounts, the MPo values of the alkenes clearly increase. With this sample, partial removal of EFAL by acid leaching (D EMT 800ex) results in a decrease of the initial molar portions of the olefinic compounds. Then, product distribution achieves values similar to those obtained with the other samples. Quantified contributions of normal [Eqn. 1 and Eqn. 2] and dehydrogenating [ Eqn. 3 ] cracking to the primary scission of the heptane molecule can be estimated from the MPo values using the following relations: contribution of normal cracking: CNC (%) = 2 (MPo(C3H8) + MPo (C4Hlo) ) contribution of dehydrogenating cracking: CDC (%) = 100 - CNC. The data (Table 6) reveal that in the primary steps of C3/C4 scission, usually about 35 to 45% of the cracked products are formed via dehydrogenating cracking. Only with D EMT 800, which owing to dealumination contains larger amounts of EFAL, cracking of heptane following this pathway is markedly enhanced (62%). After partial removing of EFAL by acid leaching, pathways of normal cracking clearly predominate again. It should be taken into consideration that the mechanism of cracking is not necessarily based on a monomolecular scission of chemisorbed C7 species. Meusinger et al. [25 ] suggested that the formation of parts of the product compounds should be attributed to a bimolecular cracking mechanism, which assumes Cv dimer (C~4) adsorbates at the zeolite surface. Cracking then obeys a more complex stoichiometry; C3 and C4 hydrocarbons are formed no longer in equivalent molar amounts.

4. Conclusions

Y zeolite based catalysts are active catalysts in the cracking process of n-heptane. With increasing time on stream, the catalysts undergo rather rapid reversible deactivation. Initial catalytic activity is re-established by regeneration of the spent

66

G. Lischke et al. /Applied Catalysis A: General 129 (1995) 57-67

catalysts in air at 723 K. Dominating products of the reaction are the C3 and C4 aliphatics. Pathways producing alkenes and free hydrogen are involved in the primary steps of reaction (scission of the parent hydrocarbon). During the initial stages of the reaction (low degree of conversion), propene and butenes clearly prevail within the spectrum of the cracking products. Owing to hydrogenation and hydrogen transfer, the portion of the saturated compounds (propane and butanes) rapidly increases, as the degree of conversion is raised. Catalytic properties of Y zeolites are influenced by dealumination. The effect is caused by the decrease of the Br0nsted acid sites of the framework as well as by the formation of extra-framework aluminium species. As was expected, the removal of framework aluminium ions, to which the Br0nsted acid hydroxyl groups are attached, results in a decrease of the catalytic activity. Extra-framework species exhibit a double function. On the one hand, they can give rise to a decrease of catalytic activity probably by an electro-chemical neutralization of the strong acid hydroxyl groups of the framework. On the other hand, extra-framework aluminium can be the carrier of catalytically active sites. The first function of EFAL --probably intensified by the simultaneous exchange of Na ions - - i s demonstrated by the increase of the activity following the acid leaching of dealuminated samples (D FAU 600ex, D EMT 600ex, and D EMT 800ex). This effect is strongest if the degree of dealumination is high, that is if EFAL is present in rather high amounts. TPDA and IR data show that partial extraction of EFAL is accompanied by a re-increase of BrOnsted acidity. The second function is indicated by the remaining low activity even in the total absence of spectroscopically detectable bridging hydroxyl groups ( D FAU 500ex). Further evidence of a specific contribution of extra-framework aluminium to catalysis is given by the fact that a low level activity favouring dehydrogenating cracking can be observed even after strong dealumination (D EMT 800).

Acknowledgements
The authors would like to thank Chem.-Ing. R. Eckelt and Chem.-Ing. E. Lieske for running the hydrogen determination and the analysis of the individual compounds of the C4 fraction of the product gas, respectively.

References
[ I ] C.D. Chang, Catal. Rev.-Sci. Eng., 25, l (1983) 1. [2] P. Jarlagadda, C.F.-R. Lund and E. Ruckenstein, Appl. Catal., 54 (1989) 135. [3] G. Ohlmann, H.-G. Jerschkewitz, G. Lischke, B. Parlitz, R. Eckelt and M. Richter, Z. Chem., 28, 5 (1988) 161. [4] J.T. Miller, P.D. Hopkins, B.L. Meyers, G.J. Ray, R.T. Roginski, G.W. Zajac and N.H. Rosenbaum, J. Catal., 138 (1992) 115.

G. Lischke et al. /Applied Catalysis A: General 129 (1995157-67 [5] [6] [7] [8] ]9] [ 10] [ 11 ] [ 12] [ 13] [ 14] [ 15] [ 16] [ 17] 118] [ 19] [20] [21] [22] [23] [24] [25]

67

F. Lemos, F. Ramoa, M. Kern, G. Gianetto and M. Guisnet, Appl. Catal., 29 ( 19871 43. N.L. Spiridonova, S.E. Spiridonov and S.N. Khadshiev, Kinet. Katal., 30 ( 1) ( 19891 243. G. Bourdillon, G. Gueguen and M. Guisnet, Appl. Catal., 61 (1990) 123. G. Ohlmann, H.-G. Jerschkewitz, G. Lischke, R. Eckelt, T. Gross, B. Parlitz, I. Schulz, K. Wehner and D. Timm, Stud. Surf. Sci. Catal., 65 ( 1991 ) 415. M. Richter, Angew. Chem. In[. Ed. Engl., 30 ( 1991 ) 607. J. Cejka, A. Vondrowa, B. Wichterlova, G. Lischke, H.-J. Jerschkewitz and E. Schreier, CCCC (in press). A, Corma, V. Fomes, A. Martinez and A.V. Orchilles, Am. Chem. Soc. Syrup. Set., 368 ( 19881 542. J.M. Cruz, A. Corma and V. Fornes, Appl. Catal., 50 (1989) 287. N,P. Rhodes and R. Rudham, J. Chem. Soc. Faraday Trans., 89 (14) ( 19931 2551. Yao Sun, Po-Jen Chu and H. Lunsford, Langmuir, 7 ( 1991 ) 3027. A. Macedo, A. Auroux, F. Raatz~ E. Jacquinot and R. Boulet, Am. Chem. Soc. Syrup. Ser., 368 ( 19981 98. F. Dougnier, J. Patarin, J.L. Guth and D. Anglerot, Zeolites, 12 (1992) 160. I. Abbot and F.N. Guerzoni, Appl. Catal. A, 85 (1992) 173. B. Mercier des Rochettes, C. Marcilly, C. Gueguen and J. Bousquet, Appl. Catal., 58 ( 19901 35. E. Jacquinot, A. Mendes, F. Raatz and C. Marcilly, Appl. Catal., 60 (1990) 101. A.T. Townsend and J. Abbot, Appl. Catal. A, 95 ( 19931 221. A. Corma, M. Faralcos and A. Mifsud, Appl. Catal., 47 (1989) 125. D.B. Lukyanov, J. Catal., 145 (1994) 54. D.B. Lukyanov, V.I. Shtral and S.N. Khadzhiev, J. Catal., 146 ( 19941 87. M. Briend, M. Derewinski, A. Lamy and D. Barthomeuf, Stud. Surf. Sci. Catal., 75 { 19931 409. J. Meusinger, J. Liers, A. M6sch and W. Reschetilowski, J. Catal., 148 (1994) 30.