J . Chem. SOC., Furaduy Trans. 1, 1989, 85(8), 1945-1962 Zeolites treated with Silicon Tetrachloride Vapour Part 5.-Catalytic Cracking of n-Hexane Michael W. Anderson, Jacek Klinowski and John M. Thomas? Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW Michael T. Barlow B. P . Research Centre, Chertsey Road, Sunbury-on- Thames, Middlesex T W16 7LN The catalytic activity of a series of samples of zeolite Y dealuminated with silicon tetrachloride vapour was tested by cracking of n-hexane at various temperatures. A reaction mechanism, involving high-molecular-weight intermediates, is proposed to account for the observed product distribution. Preliminary experiments with the cracking on dealuminated ferrierite and zeolite omega (synthetic mazzite) are also reported.Earlier papers' have described the preparation and characterization of a series of samples of zeolite Y dealuminated with silicon tetrachloride vapour at elevated temperatures. We now consider the catalytic performance of these zeolites with reference to their acidic properties and the catalytic behaviour of hydrothermally dealuminated samples. Several investigations have been carried out of the catalytic properties of aluminium-deficient zeolites. In a study of the rate of iso-octane cracking over ultrastabilised zeolite Y Barthomeuf and Beaumont2 found that all samples with more than 37 aluminium atoms per unit cell exhibited similar cracking activities. Below this aluminium content the activity dropped, suggesting that the strong acid sites, of which there are ca.40 per unit cell, were the seat of the cracking activity. The distribution of products did not change a great deal over a wide range of aluminium contents. Using aluminium-deficient zeolites Y prepared by various techniques Jacobs et aL3 found that the nature of the acid site involved varies from one catalytic reaction to another. For example, the site for toluene disproportionation appeared to be stronger than that required for cumene cracking. Abbas et aL4 investigated the isomerisation of cyclopropane over acid-leached and over steam-treated zeolite Y. They found a variation in activation energy with aluminium content, and related this effect to a change in acid strength of active sites upon dealumination. We have used the cracking of n- hexane as the test reaction for the measurement of catalytic activity.This approach was initiated by Miale et aL5 who found that cracking of n-hexane over a variety of zeolites and amorphous silica-aluminas revealed the existence of a wide range of catalytic activities. However, despite this variation, the apparent activation energy, E,, was in most cases close to 120 kJ mol-'. This fact was used to derive the activity parameter, a, defined as the activity of the zeolite relative to that of amorphous silica-alumina at an extrapolated temperature of 538 "C. a ranged from unity to over lo4, and the samples with the highest values are referred to as 'superactive'. t Present address : Davy-Faraday Research Laboratory, The Royal Institution of Great Britain, 21 Albemarle Street. London WlX.19451946 Catalytic Cracking of n- Hexane Table 1. Heats of adsorption of n-hexane on zeolite Y no. of aluminium Si/AI atoms per unit cell AH/kJ mol-' 2.5 54.8 -42.7 6.2 26.7 -41.4 7.0 23.4 -48.1" 9.0 19.2 - 42.7 100 1.9 -48.3 "Sample which may have been partially dea- luminated (ultrastabilised) by deep-bed treat- men t. Experiment a1 The cracking of n-hexane was performed in a Pyrex microreactor holding I .5 cm3 of catalyst of 10-25 mesh crystal size. Hydrogen forms of the zeolites were prepared by calcining the ammonium-exchanged forms. Nitrogen was passed through the reactor at a rate of 10 cm3 min-l, which corresponds to a contact time of 9 s. The nitrogen could be passed either directly through the reactor or diverted via a bubbler containing 99% pure n-hexane maintained at 25 0.1 "C.Prior to testing, catalysts were activated by heating in air to 538 "C for 15 min. All glassware was packed with fused alumina in order to reduce dead space. Conversion was monitored after 5 min on stream by sampling with a syringe through a septum placed close to the catalyst. All products were analysed by gas chromatography and conversions were kept between 5 and 40% to avoid analytical inaccuracies and mass- and heat-transport effects. With some of the less siliceous material extensive coking of the catalyst meant that catalytic runs at different temperatures had to be performed on fresh samples. Initial heats of adsorption, H , of n-hexane over the dealuminated zeolites, were determined by the pulse-flow technique described later.The activation energy for n-hexane cracking, E, is related to the apparent activation energy E,, determined from kinetic studies, by E = E,+( - A H ) (1) where AH is the heat of adsorption. In order to determine AH a 6 cm chromatographic column was packed with zeolite of 60-80 mesh screen size. Hexane in a nitrogen carrier gas was passed over the catalyst at adsorption temperatures between 100 and 200 "C, and the retention times measured. The preparation of the SiC1,-treated catalysts was described in detail elsewhere. In Table 1 shows the Si/Al ratios of the zeolites tested. In the parent material 92% of the sodium cations where replaced by protons. In the dealuminated materials some of the cation-exchange capacity is satisfied by residual extra-framework cationic aluminium species.At Si/Al = 7.2 this accounts for half the exchange capacity. At higher Si/AI ratios of ca. 50 the extra-framework A1 content is very 10w.l~ Results and Discussion Heats of Adsorption of n-Hexane Determination of the heat of adsorption by gas-solid chromatography is described by Greene and Pust' and Hamerski' and the application of the method they used to the study of zeolite X was examined by Eberly et aZ.**' The relationship between the chromatographic data and the adsorption equilibrium constant, B', is Llt, ue = 1/B'M. W. Anderson, J . Klinowski, J . M. Thomas and M . T. Barlow 1947 2.4 4 a' I . 2.2 2.3 2.4 2.5 2.6 lo3 K/T Fig. 1. Logarithm of the corrected retention time us. 1/T for the absorption of n-hexane.Si/AI: 0 , 2.5; W, 6.2; 0, 7.0; A, 9.0; +, > 100. where L is the length of packed column, t, is the retention time of pulse maximum, Ue is the superficial linear gas velocity, i.e. velocity in a completely empty column and B is the adsorption equilibrium constant, defined as the ratio of the number of molecules adsorbed per unit volume of column at equilibrium to the number of molecules in the same volume of gas. Since B' is a thermodynamic equilibrium constant we may write B = Bexp(-AH/RT) (3) Int, = C - A H / R T (4) where B is a constant. Substituting eqn (3) in eqn (2) and rearranging we have where C = In (BL/ Uc). The observed retention time must be corrected to 25 "C and atmospheric pressure where V z is the limiting retention volume, pi is the column inlet pressure, p, is the column outlet pressure, is the column temperature, & is the temperature of flow measuring device and F is the flow rate of carrier gas.As the column is very short, the ratio pi/po is close to unity and therefore t,(corr) = t , </T. (6) In(?, TJT) = C-AH/RT. (7) Inserting t,(corr) into eqn (4) instead o f t , gives A plot of In(?, C / T ) t's. I/Tshould therefore be a straight line with slope - A H / R . The advantage of this method for determining AH over more conventional methods is that AH is calculated at high temperature and low sorbate loading under conditions which closely simulate those used for cracking tests. Plots of t,(corr) us. I / T for a series of dealuminated zeolites are given in fig. 1, and the values of AH are listed in table 1.AH is fairly constant, ca. 42 kJ mo1-', for samples with different Si/AI ratios. The sample with Si/AI = 7.0 which may have undergone deep-bed treatment simultaneously with treatment with SiCl, vapourfd has a slightly larger value of AH, as does the highly dealuminated sample (Si/AI = 100).1948 Catalytic Cracking of n- Hexane I I I I I I 1 1 1.9 1.8 1 7 1.6 1 5 1.4 1.3 1.2 103 K I T Fig. 2. Arrhenius plots for the cracking of n-hexane over zeolite Y. Cracking of n-Hexane The degree of conversion of n-hexane to different products over the various zeolite samples and different temperatures is listed in the Appendix, while the corresponding Arrhenius plots are given in fig. 2. The latter were calculated on the assumption that the reaction kinetics are first-order and obey the equation (8) k = A exp (- EJRT).The first-order rate constant is obtained from the degree of conversion, E , by 1 k = Aln [l/(l - E ) ] z (9) where t is contact time. This relationship was verified by varying t and plotting it us. In [1/(1 --&)I. As is shown in fig. 3 the plots are linear, confirming the assumption of first-order kinetics. Cracking on zeolites with high aluminium contents causes extensive coking of the sample and consequent deactivation. Upon dealumination the apparent activation energy for the cracking of n-hexane decreases, as does the activation energy. The profile shown in fig. 4 is very similar to that found by Abbas et al.4 for the isomerisation of cyclopropane over hydrothermally treated zeolite Y , although the absolute values are somewhat different.The activation energy, E, drops from 160 kJ mol-' for the parent material to ca. 112 kJ mol-' at an aluminium content of ca. 20 atoms per unit cell. Between 20 and 4 aluminium atoms E remains constant, but below 4 aluminium atoms it rises steeply. A plot of the framework A1 content, nA,, us. conversion, E , at 315 "C over the range of constant activation energy is linear (see fig. 5). For the parent material Ea is 117 kJ mol-l, which agrees with the values reported by Miale et al.5 for a variety of aluminosilicates. Abbas et aL4 suggested that the initial activation energy could be explained in terms of the Sanderson electronegativity model as described by Mortier. lo According to this, dealumination of the framework increases its electronegativity, which in turn increases the polarisation of the framework 0-H bonds, making acid sites stronger and reducing the activation energy for protonation of hydrocarbons.This implies further that the nature of the Brarnsted-acid sites plays a fundamental role in cracking.M . W. Anderson, J . Klinowski, J. M . Thomas and M . T. Barlow 1949 2 6 10 14 18 TI s Fig. 3. Contact times, t, us. In[l/(l -&)I. 0, Si/AI = 20.7; +, Si/AI > 100. 2 00 160 120 80 I =' & I I I I I 1 10 20 30 40 50 60 no. of Al atoms per unit cell Fig. 4. E (+), E, (m) and - A H (0) us. aluminium content of the samples for the cracking of n-hexane over zeolite Y. An increase in the acid strength with decreasing aluminium content for the same samples reported in this work was observed previously.le This was based on a shift of the infrared band at 3640 cm-', corresponding to an 0-H stretch, to lower frequency.Interestingly, at ca. 15 aluminium atoms per unit cell the 3640 cm-' band no longer moves to lower frequency. This correlates closely with the point at which the activation1950 Catalytic Cracking of n- Hexane 1.0 0-8 0.6 & 0-4 0.2 0 0 4 8 12 16 20 24 nAl Fig. 5. Conversion of n-hexane, extrapolated to 315 "C us. the number of framework aluminium atoms per unit cell, nA,, for a series of SiC1,-treated samples of zeolite Y. energy for n-hexane cracking levels off. This indicates a direct correlation between acid strength and n-hexane cracking activity. It is also possible that activation energy is reduced because the reaction takes place in the newly formed Lewis acid sites thought to be associated with extra-framework aluminium.It is well known that in the presence of strong Brsnsted acids Lewis acids are ideal for forming a carbonium ion from an alkene. This may provide a mechanism for the isomerisation of unsaturated reaction products but is less likely to be the driving force behind the initial carbocation formation. Reaction Mechanisms It has been proposed1' that the formation of a carbonium ion is initiated by thermal production of an olefinic species: R,-CH,-CH,-R, --+ R,-CH=CH-R, + H,. This is a gas-phase reaction and its activation energy is likely to be high. Protonation of the olefinic species leads to the formation of a carbocation:12 R -CH = CH-R*+ HZ -+ R - C H ~ H - R +z- I HM .W. Anderson, J. Hinowski, J. M. Thomas and M. T. Barlow 1951 where HZ denotes the Brernsted-acid site. From studies of the rates of cracking of n-hexane13 it is known that the second reaction is over 200 times faster than the first. It would therefore appear that olefin formation is the rate-determining step. If this is the case, modification of the acid sites (for example by dealumination of the zeolitic framework) should have little effect on the activation energy, unless changed electrostatic potentials in the interstitial space alter the activation energy for the production of an olefinic species. However, as is evident from fig. 4, the activation energy changes considerably, and it is therefore necessary to postulate another mechanism for carbocation formation in which framework hydroxyls play a fundamental role in altering the rate-determining step.There are two possible routes to carbocation formation which would intimately involve the catalyst. First, the catalytic dehydrogenation of n-hexane may be initiated by extra-framework aluminium. This would then be followed by protonation by the zeolite as, indicated in the above reaction, to form a carbenium ion. The other possibility is the formation of an Olah-type non-classical carbonium ion1* by direct protonation of the n-hexane by the zeolite. This is shown below: This reaction would then be followed by a cleavage of the three-centre bond to yield a conventional carbenium ion : The above mechanism offers an explanation for the drop in activation energy in the cracking of n-hexane upon dealumination.Indeed C i species have only been shown to exist in superacidic environments (such as fluorosulphuric acid, H , = - 15.6) which are similar to the intracrystalline zeolitic environment. After the formation of the carbonium ion the reaction path becomes rather complicated. Cleavage of carbonxarbon bonds may take place via a a-scission one carbon away from the electric charge,". l5 producing an olefin and a new carbonium ion : H H H H H I I I l l + I I I I H 3 C : C : C : C : C : C : R H H H H H H H 1 1 1 7 7 : C : C : R I I H H This primary carbonium ion is very unstable and must isornerise to carbonium The olefin is then very likely to form a carbonium ion a secondary itself, which1952 Catalyric Cracking of n- Hexane would then undergo isomerisation.The possible reaction paths for this mechanism are given in scheme 1. nC4 +C2 + iC4 +C2 C3 +C2 +C, + iC4 +C2 iC5 +C1 P H -H2 + nC6 +/\/\I+ /\/\/+'ic6 ;c6 nC5 +C1+ iC5 +C1 L c3 +c2 +c* J L iC4 +C2 iC5 + C t Scheme 1. Assuming that secondary reactions are insignificant, the underlined species will be the major products. The absence of secondary reactions was verified by varying contact times which did not alter the distribution of reaction products to any great extent. According to the reaction scheme the product distribution should contain all species from C, to C,. Inspection of the product distributions given in the Appendix reveals that for all samples except those with very high Si/Al ratios there are virtually no C, or C, species.This discrepancy between the expected and observed product distributions is well do~umented.'~~ 16-19 The distribution found for the parent material closely resembles that reported by Tung and McIninch." The main product for cracking over the low Si/A1 ratio materials is propane followed by isobutane and isopentane. The straight- chain hydrocarbons are in lower abundance, while skeletal isomerisation amounts to ca. 5% of the product. If the reaction scheme shown above were to be operating then the ratio of the rates of formation of each kind of secondary carbonium ion would be approximately : + + with C , w C, and C, w C4, w products for cracking over a samples except that with the The distribution of reaction ,here the Cs are concentrations.variety of dealuminated samples is given in table 2. For all very high Si/Al ratio the amounts of C, and C, species are very low. The need to postulate a reaction mechanism other than a simple p-scission of the carbocation was first realised by Bolton and Lanewalal' when interpreting theM. W. Anderson, J. Klinowski, J. M . Thomas and M. T. Barlow Table 2. Distribution of products for the cracking of n-hexane on zeolites Y 1953 2.5 2.5 2.5 2.5 2.5 4.8 7 .O 7.0 7.6 7.6 20.7 33.7 33.7 48.8 48.8 75.0 75.0 75.0 > 100 > 100 > 100 ~~ 54 33 1 54 340 54 354 54 363 54 372 33 23 1 24 230 24 300 22 276 22 282 8.8 300 5.5 312 5.5 353 3.8 316 3.8 349 2.5 403 2.5 445 2.5 494 I .9 499 1.9 538 I .9 548 > 100 > 50 > 100 > 100 > 100 > 100 > 100 > 100 > 100 > 100 > 100 > 50 > 100 > 100 > 100 > 30 23 2.3 3.4 2.3 2.1 21 15 18 18 18 > 50 > 50 > 100 > 50 > 100 > 50 > 30 18 21 17 7 7 3 .O 1.9 1.8 I .7 3.6 3.1 3.1 3.1 3.1 7.0 4.9 5.3 7.5 8.1 5.8 4.9 4.6 4.2 4.2 2.9 2.3 2.2 0.8 1 .o 0.9 I 5.5 4.6 4.7 6.2 5.0 2.9 6.0 6.3 6.3 7.2 10.8 6.0 6.4 4.7 5.2 2.5 3.4 1.6 1 .o 0.9 1.1 ~ ~~ N , , denotes the number of aluminium atoms per unit cell.I11 Scheme 2. IV I1 c c c \ I \ / \ c c c I I c c c /c\ /c\ 7 0.40 0.57 0.42 0.45 0.46 0.12 0.10 0.13 0.16 0.12 0.22 0.24 0.33 0.23 0.35 0.47 0.74 1.07 I .25 1.46 1.15 distribution of products of isomerisation of hexane over Pt-loaded zeolite Y. They considered disproportionation with the production of four intermediates based on cyclohexane, as shown in scheme 2.Bolton and Lanewala were able to show that the main products of catalytic cracking of these species were 2- and 3-methylpentane, with some 2,3-dimethylbutane but no 2,2-dimethylbutane. Bolton and Bujalski suggested1954 Catalytic Cracking of n-Hexane rr 0 100 200 300 400 TI "C Fig. 6. The number of supercages per molecule of adsorbed n-hexane on dealuminated zeolite Y us. temperature. later18 that the formation of C,, C, and C, species could be understood in terms of the reactions : 2C6 + K 1 2 l + 3c4 + 4c, +c,+c4+c5 with no C, or C, species being formed. Closer examination of the products formed by scission of the reaction intermediates in scheme 2 suggests that the main product would always be C,.Haag et al.', demonstrated that the structure of zeolite ZSM-5 imposes shape selectivity on the activity of n-hexane and interpreted their findings in terms of a large reaction complex H c c which is anchored in the zeolite as a charge-balancing cation. This seems quite likely if we consider the reaction as follows. First a carbocation is formed, probably by direct protonation of hexane by a framework hydroxyl. The carbocation cleaves according to the rules of P-scission : + /\\I + c; The reactions (i) are less likely than reaction (ii) since they result in a primary carbocation which may only be stabilised by hydride transfer.12 Reaction (ii) will now be considered in some detail. The primary carbocation forn. -d by reaction (ii) will isomerise via a hydride shift to give the secondary isopropyl carbocation.Because of its positive charge the carbocation may become pinned to the site at which it was formed by the negative charge on theM . W. Anderson, J . Klinowski, J . M. Thomas and M. T. Barlow 1955 zeolitic framework. If it were not adsorbed in this manner it would be unlikely to come into contact with another molecule of n-hexane, as the concentration of the reactant in the zeolite at reaction temperatures is very low. Also, the kinetics of the reaction would probably be second-order. It can be seen from fig. 6, which shows the uptake of n-hexane on zeolite Y at elevated temperatures, that above 150 "C there is less than one molecule per supercage. At 300 O C , a typical reaction temperature, there is only one molecule of n-hexane per 10 supercages.The probability of two molecules coming together under these conditions is very low. By being in close contact with the zeolite framework, the secondary carbocation, which is inherently unstable, may stabilise by spreading its charge on to the framework. Such hyperconjugation could occur by overlap of the vacant p orbital of the carbocation with the lone pair of electrons in the sp*-hybridised orbitals of the framework oxygen. Reaction of the secondary carbocation with another n-hexane molecule to form a larger tertiary carbocation would further increase its stability according to scheme 3, which also gives the four possible tertiary carbocations and their products. C c c c C+-H+C C C / \ / \ I \ / C C Ci +iC4 Ct + iC5 C; +23DMB c; +2MP 23DMB = 2,3-dhethylbutane 2MP = 2-methylpentane Scheme 3.1956 Catalytic Cracking of n-Hexane Reactions which result in the primary CT and C l carbocations have been neglected for reasons given previously.In this way a product distribution consisting of mainly C,, C,, C, and iso-C, compounds will result as observed experimentally. We suggest a reaction mechanism involving the following: ( I ) the lifetime of carbocations C l and C;t is too short for them to react with n-hexane molecules to a significant extent. They are neutralised by hydride transfer. (2) Once formed, a carbocation is either neutralised by hydride transfer or becomes pinned to the framework to await an n-hexane molecule. (3) Carbocations rearrange by hydride shift to tertiary carbocations. (4) Large reaction intermediates are formed which crack via p-scission. ( 5 ) Reaction intermediates leading to products larger than C, are negligible. (6) Secondary reactions other than simple isomerisation to more stable species are negligible.For very low aluminium contents (Si/AI 2 75) much more C, and C, is produced, the ratio C,/2(C4 + C,) approaches unity, there is little isomerisation, not all the olefins become saturated, and the activation energy increases rapidly. All these factors are characteristic of thermal cracking, which is consistent with the fact that higher temperatures are required for cracking over low-aluminium-content samples and that they contain very few catalytically active centres. The role of Lewis acidity in all these reactions is not clear.We know that the Brarnsted sites are playing an important role in the catalytic cracking as the 3640 cm-l infrared band, corresponding to Brarnsted-acid sites in the supercage, is eliminated during the cracking process.lf However, the possibility that Lewis acidity also plays a role in carbocation formation by hydride extraction cannot be ruled out: where [L] denotes a Lewis-acid site. Isomerisation of olefins by Lewis acids is well known, but their interactions with the much more stable paraffins is less well understood. We have shownld that Lewis acidity is present not only in the dealuminated samples but in the parent zeolite H-Y as well. The formation of Lewis acidity upon dealumination cannot therefore be the only cause of the changes in activation energy and product distribution.a-values could not be calculated for these catalysts as the activation energy for cracking varies between different zeolite samples. Such a comparison, as reported by Miale et al.,, is only valid if all the catalysts exhibit the same activation energy. Cracking by Other Dealuminated Zeolites Cracking of n-hexane over zeolite omega (synthetic mazzite) and zeolite omega treated with silicon tetrachloride leads to a similar distribution of reaction products to that found for zeolite Y of similar Si/AI ratio (see Appendix). The fact that the parent zeolite omega shows a similar product distribution to dealuminated Y with about the same Si/AI ratio suggests that Lewis acidity, which is only present to any significant extent in the latter, does not play an important role in the process.The activation energy for zeolite omega is ca. 65 kJ mol-', rather lower than for zeolite Y. The correspond- ing Arrhenius plots are given in fig. 7. 1.r. spectra of zeolite omega in the hydroxyl stretch region indicate the presence of only one kind of acid site, and the frequency (ca. 3580 cm-l) suggests that it is stronger than in zeolite Y. This is likely to be respon- sible for the low activation energy. In the case of ferrierite the product distribution is rather different from that for zeolites Y and omega (see table 3). Over the parent sample the products are mostly saturated hydrocarbons, which is indicative of catalytic, rather than thermal, cracking. The C, species are the most abundant, and the ratios C,/C, and CJC, are much closer to unity than for zeolite Y.There is also little isomerisation. The most likely reason for thisM. W. Anderson, J . Klinowski, J . M . Thomas and M . T. Barlow 1957 0 9 - 2 - -Y s - 4 - 2.2 2.0 1.8 lo3 KIT 1.6 Fig. 7. Arrhenius plots for the cracking of n-hexane over parent zeolite omega (m, Si/AI = 4.24) and partially dealuminated omega (0, Si/Al = 6.00). Table 3. Distribution of products for the cracking of n-hexane on zeolites on zeolites omega and ferrierite Si/AI N* I T/"C CJC, C,/C, iC,/nC, iC5/nC5 C3/2(C, + C,) __ __ - - . __ _ ~~ __ 4.24 4.24 4.24 4.24 6.0 6.0 6.0 4.6 4.6 4.6 9.3 9.3 9.3 9.3 6.9 6.9 6.9 6.9 5.1 5.1 5.1 6.4 6.4 6.4 3.5 3.5 3.5 3.5 omega 200 > 100 > 100 232 > 100 > 100 252 > 100 > 50 268 > 100 > 100 209 > 100 > 100 235 > 100 > 100 252 > 100 > 50 ferrierite 280 9.4 2.6 300 8.I 2.8 320 5.4 2.5 376 1.9 0.9 389 2.2 1 .O 410 2.2 1.2 430 2.1 1.2 8.8 6.7 5.0 4.6 8.5 5.3 4.4 0.5 0.5 0.5 0.6 0.6 0.7 0.8 > 100 7.9 6.0 5 .O 6.4 5.1 > 100 0.4 0.6 0.3 0.0 0.0 0.0 0.0 0.05 0.08 0.10 0.12 0.07 0.09 0.12 0.57 0.56 0.58 1.07 0.97 0.87 0.84 product distribution is the steric restrictions imposed on the prqducts and reaction intermediates by the small pore size of ferrierite (4.3 x 5.5 A). Large reaction intermediates of the kind shown in scheme 3 can no longer form, and cracking must proceed along the more conventional routes shown in scheme 1. This gives further support to our postulate that large reaction intermediates are formed in wide-pore zeolites.Cracking over ferrierite treated with SiCl, produces olefinic C, and C , species, which suggests that insufficient hydrogen is available for the products to be saturated hydrocarbons. Thermal cracking is unlikely as the activation energy is low. Restricted1958 Catalytic Cracking of n-Hexane 0 - 2 * - 4 Ei - 6 - 8 / 0@ a /@ /@' 1.74 1.58 1.4 2 103 K/T Fig. 8. Arrhenius plots for the cracking of n-hexane over parent ferrierite (m, Si/AI = 4.6) and partially dealuminated ferrierite (@, Si/AI = 9.3). deposition of coke in the narrow channels of ferrierite might reduce the amount of available hydrogen. The activation energy is ca. 85 kJ mol-' and the corresponding Arrhenius plot is given in fig. 8.Conclusions It is clear that zeolite Y treated with silicon tetrachloride vapour exhibits interesting catalytic properties. The activation energy for the cracking of n-hexane is reduced by nearly 30% on dealumination. This is probably caused by an increase in acid strength resulting from an increased electronegativity of the silicon-rich zeolitic framework. The postulated reaction mechanism requires the formation of a carbonium ion via a non- classical CRS species followed by a sequence of reactions during which the carbonium ion is pinned to the zeolitic framework as a charge-balancing cation and large reaction intermediates are formed. In the medium-pore zeolite ferrierite no such large reaction intermediaries can be formed and this is reflected in the product distribution. We are grateful to B.P.Research Centre, Sunbury-on-Thames, for support and to Professor J. H. Lunsford, Texas A & M University, for his comments on the manuscript. References I ( a ) M. W. Anderson and J. Klinowski, J. Chem. SOC., Faraday Trans. 1, 1986, 82, 1449; (6) M. W. Anderson and J. Klinowski, J . Chem. SOC., Faraday Trans. 1, 1986, 82, 3569; (c) M. W. Anderson, J . Klinowski and J. M. Thomas, J . Chem. SOC., Faraday Trans. I , 1986,82,2851; ( d ) M. W. Anderson and J. Klinowski, Zeolites, 1986, 6, 150; (e) M. W . Anderson and J . Klinowski, Zeolites, 1986, 6, 455; (.f) M. W. Anderson, Ph.D. Thesis (Cambridge University, 1984). 2 D. Barthomeuf and R. Beaumont, J . Catal., 1973, 30, 288. 3 P. A. Jacobs, H. E. Leeman and J. B. Uytterhoeven, J .Catal., 1974, 33, 31. 4 S. H. Abbas, T. K. Al-Dawood, J. Dwyer, F. R. Fitch, A. Georgopoulos, F. J. Machado and S. M. Smyth, in Catalysis by Zeolites, ed. B. Imelik, C. Naccache, Y. Ben Taarit, J. C . Vedrine, G. Coudurier and H. Praliaud (Elsevier, Amsterdam, 1980), p. 127.M . W. Anderson, J . Klinowski, J . M. Thomas and M. T. Barlow 1959 5 J . N. Miale. N. Y. Chen and P. B. Weisz, J. Cural., 1966, 6, 278. 6 S. A. Greene and H. Pust, J. Phys. Chcm., 1958, 62, 55. 7 J. J. Hamerski, Ph.D. Thesis (University of the Pacific, 1963). 8 P. E. Eberly Jr, J . Phys. Chem., 1961, 65, 68. 9 P. E. Eberly Jr and E. H. Spencer, Trans. Furuday Soc., 1961, 57, 289. 10 W. J. Mortier, J. Cutul., 1978, 55, 138. 1 1 P. A. Jacobs, in Carhoniogenic Activity qf Zeolifes (Elsevier, Amsterdam, 1977).12 V. Haensel, Adz?. Cutul., 1951, 3, 179. 13 W. 0. Haag, R. M. Lago and P. B. Weisz, Furuduy Discuss. Chem. Soc., 1981, 72, 317. 14 G . A. Olah, G . K. Surya Prakash and J. Sommer, Science, 1979, 206, 13. 15 F. Whitmore, Chem. Eng. News, 1948, 26, 673. 16 S. E. Tung and E. McIninch, J. Catal., 1968, 10, 175. 17 A. P. Bolton and M. A. Lanewala, J. Cutuf.. 1970, 18, 1. 18 A. P. Bolton and R. L. Bujalski, J. Cutul.. 1971, 23, 331. 19 J. A. Rabo, Cutul. Rev. Sci. Eng., I98 I , 23, 293. Paper 8/01890A; Received 13th May, 1988 Appendix Distribution of products and the total conversion (YO) for the cracking of n-hexane by various zeolitic catalysts at temperatures indicated. (a) Catalyst: H-Y, Si/Al = 2.5 T/"C 33 1 340 354 363 372 0.0 1 0.02 0.02 0.03 0.12 0.22 0.32 0.47 C2H6 0.03 0.05 0.06 0.09 3.78 6.14 8.46 12.37 iC, 2.5 1 2.98 5.29 7.48 nC4 0.69 0.96 1.70 2.42 nc, 0.24 0.26 0.52 0.54 iC, 0.70 I .2 2.5 0.6 conversion 9.43 13.04 21.38 27.36 CH4 C2H4 c3 iC5 1.31 1.20 2.46 3.34 0.04 0.60 0.12 14.70 9.68 3.14 4.5 1 0.90 2.0 35.71 (6) Catalyst: Na-Y treated with SiCI,, Si/Al = 7.0 T/"C 240 260 0.0 I 0.0 1 0.09 0.07 0.04 0.04 c3 2.91 4.65 iC, 9.61 14.00 iC5 5.10 8.44 iC, 3.0 8.9 conversion 22.9 39.3 iC, are mostly 2- and 3-methylpentane with some 2.3-dimethylbutane and very little 2,24imethylbutane.CH4 CJ-4 C*H, nC4 1.32 1.75 nc, 0.59 0.77 .~1960 Catalyric Cracking of n- Hexane (c) Catalyst: Na-Y treated with SiCl,, Si/AI = 20.7 T/"C 276 284 306 CH4 0.0 1 C2H4 0.0 I C2H6 0.0 1 iC, 3.45 iCS 1.99 c3 I .68 nC4 0.5 I nC5 0.26 iC, 3.96 conversion 2.25 0.00 0.03 0.02 2.34 4.9 I 0.69 2.97 0.35 10.34 21.78 0.0 I 0.09 0.04 4.42 7.73 1.19 4.85 0.55 13.94 32.85 (d) CataIyst: Na-Y treated with SiCI,, Si/AI = 33 T/"C 292 312 3 34 353 CH4 0.06 C2H4 0.04 c3 0.56 iC, 0.62 iC, 0.36 nCS 0.16 conversion 3.47 'ZH6 0.02 nC4 0.20 iC, I .43 ~~ 0.0 1 0.0 1 0.02 0.04 0.08 0.19 0.02 0.03 0.07 1.37 2.09 4.91 1.59 2.15 3.77 0.32 0.44 0.82 0.76 1.43 2.48 0.20 0.26 0.39 3.80 5.25 6.59 8.11 11.74 19.24 (e) Catalyst: Na-Y treated with SiCI,, Si/AI = 49 T/"C 306 316 349 369 0.00 0.04 c3 0.75 iC, 0.99 iC, 0.67 nC5 0.13 iC, 3.29 conversion 6.12 CH4 C2-4 0.01 nC4 0.22 C2H4 0.0 1 0.01 0.02 0.05 0.17 0.29 0.02 0.05 0.08 1.14 4.31 7.16 1.12 3.05 4.81 0.29 0.73 1.12 0.80 1.98 3.01 0.17 0.38 0.56 3.7 1 4.57 6.1 7.38 15.3 23.24M .W. Anderson, J . Klinowski, J . M . Thomas and M . T. Barlow 1961 (f) Catalyst: Na-Y treated with SiCl,, Si/AI = 75 T/"C 303 403 445 494 524 0.00 0.19 c3 0.23 iC, 0.39 nC4 0.23 iC5 0.26 nC5 0.03 iC6 negligible conversion 1.43 CH4 '2*6 0.00 C2H4 0.04 0.3 1 0.05 3.57 1.88 0.65 0.89 0.35 7.39 0.05 0.28 0.09 5.67 1.87 0.80 0.88 0.26 9.61 0.54 1.60 0.52 16.41 4.37 2.02 0.77 0.47 27.76 0.66 2.05 0.80 20.93 4.45 2.63 1.45 0.56 33.57 (g) Catalyst: Na-Y treated with SiCI,, Si/Al > 100 T/"C 403 499 532 538 548 CH4 0.0 1 C2H4 0.02 'ZH6 0.0 1 c3 0.39 iC, 0.06 nC4 0.04 iC5 0.03 nC5 0.03 iC6 negligible conversion 0.6 0.10 0.22 0.29 3.33 0.45 0.54 0.17 0.17 0.25 0.62 0.68 7.25 1.10 1.28 0.34 0.3 I 0.26 0.85 0.60 9.32 1.29 1.30 0.29 0.32 5.3 11.87 14.26 0.44 1.1 I 1.08 10.80 1.76 2.0 1 0.49 0.43 18.17 (h) Catalyst : NHi-exchanged zeolite omega T/OC 200 232 252 268 0.00 0.00 0.04 0.00 0.00 0.02 'ZH6 0.00 0.02 0.06 c3 0.12 0.66 2.37 n c , 0.06 0.29 1.08 iC5 0.69 1.74 4.21 nC5 0.00 0.22 0.70 iC6 0.87 8.72 9.69 conversion 2.26 13.58 23.63 CH4 C2H4 iC, 0.53 1.94 5.45 0.03 0.02 0.06 3.92 7.84 1.72 6.00 1.19 12.74 33.501962 Catalytic Cracking of n-Hexane (i) Catalyst: NHf-exchanged dealuminated omega, Si/AI = 5.4 T/"C 209 0.00 0.00 c3 0.05 iC, 0.17 iC5 0.17 iC, 4.12 conversion 4.55 CH, C2H, C2H6 0.00 nC, 0.02 nc, 0.00 235 252 0.00 0.02 0.00 0.02 0.00 0.03 0.30 1.20 0.74 2.28 0.14 0.52 0.64 1.77 0.10 0.45 4.47 6.44 6.40 12.65 . .. ~ . __ - - ( j ) Catalyst: NHf-exchanged synthetic ferrierite, Si/AI = 4.6 T/ "C 280 300 320 0.12 0.09 C2H6 0.93 iC, 0.95 nC4 1.74 iC5 0.30 nC5 0.83 iC6 negligible conversion 9.3 I CH, c3 4.34 C2H4 0.25 0.15 1.69 8.13 1.66 3.51 0.73 1.29 17.42 0.5 I 0.27 2.93 12.44 2.68 5.25 0.71 2.05 26.85 (k) Catalyst : NHi-exchanged SiCl,-treated ferrierite, Si/AI = 9.3 ~ _ ~ ~ ~ ~ _ _ _ ~ _______. - ~ T/OC 280 376 389 410 430 0.00 0.19 0.25 0.03 0.46 0.66 C2H6 0.07 1.21 I .57 iC, 0.04 0.56 0.91 nC, 0.00 0.90 1.39 nC5 0.00 0.37 0.56 iC, negligible conversion 0.48 7.62 10.88 CH4 C2H4 c3 0.26 3.92 5.54 iC5 0.07 0.00 0.00 0.4 1 1.16 2.3 1 8.92 1.18 2.40 0.00 0.9 1 17.92 0.64 1.84 3.23 12.30 2.72 3.32 0.03 1.29 25.38