Indian Journal of Chemist ry Vol. 44A, September 2005. pp. 1772-1 78 1

Selective benzylation of benzene over alumina pillared clays

Manjll Kurian & S SlIgllnan*

Department of Applied Chemistry. Cochi n Un ivers ity of Science and Technology. Kochi 680022. India Emai l: ssg@cusal. Jc. in

Neceived 13 Decelllber 2004; revised 16 Jllile 2005

Al umin ium pi ll ared clays prepared by part ial hydrolysis method has been subjected to roo m temperature exchange with u'ans ition metals of the first se ries. The resulti ng materials ex hibit a fine porous network of exceptional stability as characteri zed by XRD. FT IR and MR spectroscopy. EDX and surface area measu rements indicate the presence of exchanged mewls allached to the pillars. Acidi ty of the pillared clay benefits from the in se rt ion of transi tion metal s. Benzylat ion of benzene occu rs effic ient ly over these catalys ts with 100% monoalkylated product selectivi ty. An extensi ve study or the reaction variables points towards a carbocati on mechani sm for the reaction. The reaction depends to a large exten t on the number and avail ability of the Lewis acid sites.

IPC Code: int. Cl 7 BOIJ2 1/ 16

Pillar in terlayered clays constitute one of the widely studied groups among the new mic roporous materi als developed by molecular engineering. These solids, al so known as cross linked clays are obtained by exchanging the interl ayered cations of layered clays with bulky inorgani c pol yoxocations, followed by ca lcination. The interca lated polycati ons increase the basa l spaci ng of clays and upon heating, are converted to metal ox ide clu sters by dehydration and dehydroxy lation . As a res ult, a two dimensiona l porous network is generated . Maintenance of a well­defined porous network up to re lati ve ly high temperatures along with the presence of acid si tes immediately suggest potential catalytic applications for these solidsl.2

.

Fr iedel-Crafts' alkylation is an important means of attaching alkyl chai ns to aromatic rings . o i phen y lmethane and substi tuted di pheny lmethanes are industri ally significant compou nds used as heat tran fer fI uids, aromat ic sol vents, fragrances and monomers for po lyca rbonate res ins. They are also precursors to benzophenones, synthesised by an air ox idation step in ,:c ~tic acid medium in presence of manganese based catalysts3. Traditionally, homogeneous ac id cata lysts like AICl3, BF3 and H2SO.. are used fo r Friedel-Crafts alkylations". However. use of Lewis ac id catal ys ts is laden with seve ral problems like difficulty in separation and recovery of prod ucts, di sposa l of spent catal yst, corrosion , high toxicity etc5

.6

. Most of the catalysts

have to be added in stoichiometric amounts, thereby addi ng to the cost of the des ired product. Hence efforts are being made to replace the present environmentally ma li gnant catalys ts with so lid acid catalys ts.

In thi s paper, we report the relevance of metal exc hanged aluminium pi ll ared clays for the benzylation of benzene usi ng benzy l ch loride . Se lecti ve formati on of the monoa ikylated product occ urs. The observed act ivities coul d be corre lated wi th the strong/Lewis acidit ies of the catalys t. A thorough study on the structural stabi lity of the catalysts as well as reaction va riab les has been ca rri ed out.

Materials and Methods Aluminium pi ll ared clay was syn thesised by the

partial hyd rolys is of 0. 1 !VI AI(N0 3)3 solutton by dropwise addit ion of 0.3!V1 a2C03 solu ti on under vigorous stirring. N2 gas was bubbled through the solution to remove excess CO2 anu aged for 24 hou rs at room temperature. Intercalation of pillaring species into the cl ay layers was done by trea ti ng the pillaring solution with a prev iously swo llen clay sLl spension at 70°C (OHImetal ratio: 2 and metal/clay rati o: 20 mmol/g clay) . The clay, afte r exc hange, was washed several times with distilled water and filtered . This was dri ed in an air oven at l lOoC overn ight. fo llowed by calcinat ion for 6 hOllrs at 450°C in a muffle furnace. Exchange with transition metals was done using 0.1 molar aqueous sol utions of the

KURIAN & SUGUNAN: SELECTIVE BENZYLATION OF BENZENE OVER ALUMINA PILLARED CLA YS 1773

corresponding metal nitrate. For exchange with vanadium, reqUisite amount of ammonium metavanadate was dissolved in oxalic acid. Pillared cl ays were stirred mechanically with salt solutions for 24 hours at room temperature. The clay after exc hange was washed 5-6 times with distilled water. This was filtered and dri ed in an air oven at llOoC overnight and calcined for 5 hours at 500°e. The pill ared clays synthesised for the present study are notated as XI AI PM where X is the transition metal exchanged.

Cation Exchange Capacity measurements were done by ammonium acetate extract method. The evo lved ammonia was quantified by microKjeldahl method. EDX analysi s of the prepared samples was done in a Jeo l JSM-840 A instrument with a resol uti on of 1.3 e V. Samples were prepared by dusting the clay powder onto double sided carbon tape, mounted on a metal stub. The diffractometer traces of the catalyst sa mples were taken in Ri gak u D/MAX-C inst rume nt llsing Cu-Ka radiation (I, = 1.5405A) . The simultaneou'i determination of surface areas and pore volumes of the catalyst sa mples was done on a Micromeritics Gemini analyser. Previously activated samples were degassed at 200°C under nitrogen atmosphere for 2 hours and then brought to nitrogen boiling poi nt. FTlR spectra of the sa mpl es were recorded using a Perkin-Elmer RX-I spec trometer by the KBr disc method in the range 400- 4000 cm·l. NMR specU'a of sa mples were recorded on a 300 DSX Brucker spectrometer. Temperature programmed desorption of ammonia was done using a conventional flow type apparatus. Cumene cracking test reaction was done in vapour phase in a fixed bed, downflow vertical glass reactor. The catalyst (0.5 g) activated at 500°C was immobilised inside the reactor usi ng glass wool, sandwi ched between inert silica beads. Cumene was fed into the reactor with the help of a sy ringe pump at controlled fl ow rate. The products were collected downstream ina recei ver connected through a water condenser and analysed using a Chemito 8610 gas chromatograph equipped with flame ioni sation detector and FFAP column.

The liquid phase benzylation of benzene was carried out in a closed 50 ml round bottomed glass flask equipped with a reflux condenser, magnetic stirrer and provIsion for withdrawing product samples . The temperature of the r'::action vessel was maintained using a thermostated oil bath. In a typical

run, appropriate amounts of benzene, benzyl chloride and catalyst were allowed to react at speci fied temperatures under magnetic stlrnng. Reaction mixture was withdrawn at specific interval s and analysed using gas chromatography using SE-30 column. The percentage conversion (wt%) of benzy l chloride is the total percentage of benzyl chloride transformed into the products.

Results and Discussion Residual cation exchange capacity measurements

Cation exchange capacity of pillared clays is only partially compensated by the charge of oligomcrs. Even if the oligomer to clay ratio is very hi gh, a part of exchangeable cations remains unchanged. The res idual CEC provides an estimation of the fraction of laye r charge, which is not compensated by cationic pi lIaring species. Duri ng exchange process, sQme monomeric forms of the pillaring spec ies cO~lld as well be exchanged. The protons that are formed in final calcination step of pillaring process . when intercalated oligomeric cation is deco mposed into metal oxide pillars, can also restore CEC of clay . Thus, residual CEC gives a measure of effectiveness of pillaring process . Montmorillonite shows a cation exchange capacity of 0.91 mmol g.1 which red uces to 0.56 mmol g.1 on pillaring with aluminium. 61.5 ~ of the original CEC is retained on the clay. Aluminiull1 polymeric solution contains AI monomeri c and dimeric species and they also are considered to be exchangeable. Hence, res idual CEC of AI PM is tile CEC that has not been exchanged with Al l:; pol ymer or it is an indication of effectiveness of pillaring \Vith AI 13 polymer.

Energy dispersive X-ray analysis Energy dispersive X-ray analysis yie lds the

chemical composition of the prepared samples. The e lemental compositions of individual syste ll1s are presented in Table 1. It can be seen from the tabl e that Sil AI rati o of parent clay decreases from 3.18 to 2.0. Decrease in SiiAI rati o is a direct consequence of the Increase in aluminium content as a result of pill ari ng rather than structural in stability of the clay laye rs. Aluminium content increases by 9.14% \Vith corresponding decrease in amount of exc hangeable cations. From the data, it can also be concluded that pillaring occurs at interlamellar sites in the place of exc hange cations. This fact is supported by CEC measurements . Exchange with transition metal s incorporates about 1-2% of the meta! at the expense

1774 INDIAN J CHEM, S EC A, S EPTEMBER 2005

Table I - Ele mental composit ion o r aluminium pill ared syste ms

Catalyst

VIAl PM Mn/AI PM ColAI PM Nil AI PM Cul AI PM Zn/AI PM Al PM Mh

a

0.45 0.49 0 .63 0 .88 0.49 0 .69 1.44 2. 18

Mg

1.54 2. 19 1.67 2.03 1.25 1.24 2 .96 2.8 1

" Exchanged trzlIlsiti on metal: h MOlllmori:l onite KSF

El ement(%)

AI Si

24.85 57 .71 24.51 58.65 23 .19 59 .55 24 .08 58.32 23.28 59.42 22 .20 58.4 26.87 54.42 17 .74 56.47

K Ca Fe T M"

2.7 0.34 10.0 2. 36 2 .42 0 .33 10.38 1.03 2 .65 0 .64 10 . .19 1.08 2.7 0 .46 10 05 1. 15

2.46 0.74 10.24 1.06 2. 63 0.82 10.78 1.62 3. 16 0 .94 10.91 4 . 11 5. 01 1 1.67

Tabl e 2 - Surrace ,l'ld pore volullle or aluminiu lll pill ared seri es

Surface area (nlg· l) Po re vol (ecg' l

)"

Cata lys t BET Langmuir

VIAl PM 105.2 150 Mn/ AI PM 120.7 163.9 Col A I PM 125 .2 171.8 Nil AI PM 12 1.3 165.2 Cu/AI PM 126.9 179.1 Zn/AI PM 11 68 168.9 AI PM 132.9 192.3

" Pore volume measured at 0 .9976 P/ Po

of interlame ll ar cations. Vanadium exchanges best. Thus, exchange with the transition metals of the first se ri es further replaces these cations.

In order to show the effects of exchange process on the original constituents of the clay, the elemental we ights can be reca lculated to 100% after excluding the fixed transition metal oxide. This indicates that transition metal oxide was introduced without a significant deleteriou s effect on the clay structure and the pillars. Thus, on exchange with transition metal , metal oxides are incorporated into the pores or in between the pillars rather than exchanged in the location of pillars.

Surface area and pore volume measurements The determination of surface area and pore volume

in clays is a subject of controvers/. The most convenient approach for textural characterisation of these material s is to obtain the adsorption isotherm at a te mperature lower than or equal to critical adsorbate temperature. N2 (at 77.2 K) is traditionally used as adsorbate. As gas pressure :ncreases, adsorption proceeds by pore filling, stalting from smaller pores. The potential energy fields from neighbouring surfaces overlap and the total interaction energy with ad sorbate molecules becomes substantially enhanced g iving ri se to high gas volume adsorption at very low relative pressures. The surface areas of pillared clays

M icroporous External

7 1.2 34.0 01432 74.6 46.1 0. 1493 73 .8 5 1.4 0. 1539 720 49.3 0 . 1502 74.9 52.0 0. 1592 70 .2 46.6 0 . 1498 86.9 46 .0 0.1623

are typically obtained by applying BET equation. The range of va lidity of BET equation fo r these materi als is usually between PIPo = 0.01 and 0.1. However, in microporous solids like pillared clays where the interlamellar distance is of the order of a few molecular diameters , monolayer formati on on clay silicate layers occurs. Thus, surface areas approximated by Langmuir equati on are reasonable representations of pillared clay surface areas8

. Hence in the present study, BET and Langm ir surface areas of various systems obtained directl y are tabul ated. The external and microporous surface areas were calculated from l-plot9

.

Table 2 furni shes the surface area and pore volumes of various aluminium pillared syste ms. Surface area and pore volume of montmorillonite increases as a result of pillaring. Aluminium pillaring increase BET surface area to 132.9 m2il and

. .., -I LangmUIr surface area to 192.3 m-g . The pore volume increases to 0.1623 ccg·l. About one third of the area can be attributed to external surface. As a result of transition metal exchange, surface area and pore volume decreases. Zinc and vanadium exchange brings about this effect to the maximum. Decrease in surface area and pore volume can be correlated we ll with the amount of metal oxide incorporated into the pi I!ared system.

KURI AN & SUGUNAN: SELECTIVE BENZYLATION OF BENZENE OVER ALUMINA PILLARED CLA YS 1775

Increase in surface area and pore volume can be ascribed to the pillaring process. CEC measurements showed that 38.5% of original CEC was replaced by Al l) polymer. while a lower surface area was obtained for the final pillared solid. This can be due to breaking of pi liars on calci nation. Si mi lar observations were reported by Bakas et al lo

. Breaking of pillars di srupts the porous network and hence decreases the percentage of microporous surface area compared to iron pillared systems. Transition metal exchange deposits the metal oxides nea r the pillars resulting in lower percentage of external surface.

X-ray diffractiun

Apart from surface area and pore volume measurements, the easiest way to determine whether pillar intercalation is successful is to record the X-ray diffraction pattern of an oriented film of the product. Pillared clays are semi-crystalline in nature. The broad bands obtained in the XRD spectrum, instead of sharp peaks. can be attributed to the semi-crystalline nature of clays. Hence, indexing of the spectrum is not possible for thi s type of solid acids. The only data that can be obtained is the d spacing of (001) plane, which indicates the extent of propping apart of clay layers. X-ray diffraction peaks show that long range face-to-face layer aggregation is present in the pillared sa mple . Thus , it can be safely assumed that the sample is not an edge-to-face delaminated clay.

The characteristic dO~ I spacing of montmorillonite is seen at 28 value of 9.8A. For AI PM, this value increased to17 .3 A. Shifting of 28 values clearly suggests ex pansion of clay layer during pillaring process. The major interca lated species giving rise to stabie basal spacing in AI PM is the so called AI I3 polyhydroxy polymer or Keggin cation which has been characterised by small angle X-ray scattering ll

and 27AI NMRI 2. This polymer with structural formula, lAI04Al d OHb(H20 )lz]7+ is a tri-decamer composed of one aluminium tetrahedron surrounded by 12 aluminium octahedra. It contains four layers of superimposed oxygen atoms needed for expanding clay basal spacings to 18 A. The expansion of (00l) plane of clay layer near this value suggests that the intercalated species is the Al l) polymer in AI PM.

The effect of exchange with transition metals on the XRD patterns of the pillared system was studied for representative samples. The XRD patterns were exactly identical to that of the parent pillared sample. Thus, it can be concluded that insertion of the second metal a fter the formation of stable pi liars does not

destabilise the porous network. Additional peaks corresponding to the exchanged metal oxides we re not noticed. This may be due to the diminutive amounts (1-3 %) of the exchanged metal in these samples.

FTiR spectroscopy The structural evolution of the aluminosili cate

layer was characterised by FTIR spectroscopy. The parent montmorillonite shows a large band at 3620 cm· l, typical of smectites with large amount of AI in the octahedral }. Intensity of this peak decreases upon pillaring. Formation of a new band in the range of 3740-3770 cm·1 is an important observa ti on. The identification of hydroxyl spec ies on pillared clays is extremely difficult, because of the complexity of the system and the opaqueness of the sample in the corresponding IR region. It has been reported that isotopic exchange with mild deuterating agent Cc, D6

has allowed identification of two acidic hydrox yls. with OH stretching modes at 3660 and 3740 cm' l, the former referring to the unstructured band of the parent montmorillonite and the latter which seem to ari se from the sealing of montmorillonite layer and the pillar l4 . These can be represented as Si-O-AI-OH or AI-O-Si-OH. Thus, FTIR spectra of the three pillared samples also are indicative of effective pillaring.

The IR spectra in the fingerprint region are characterised by absorptions at 1200-1000 CIll,I due to asymmetric stretching vibrations of apical oxygens of Si02 tetrahedra and the 18rge band due to combined stretching and bending vibrations of the Si-O bonds related to basal oxygens l5. The band around 900 cm' l often provides information on the composition of the octahedral sheets. In mcntmorillonite, it reflects partial substitution of octahedral AI by Mg. Absorptions at 526-471 cm' l echo bending Si-O vibrations. Thus, the framework vibrations contain information about the structural characteri stics of the material and their preservation after thermal treatments may be considered as a proof of the structural stability on pillaring. Absence of additional peaks suggests that no bond formation occurs between the montmorillonite and the pillars unlike other clays I'k ,16 t 'e sapolllte .

27 AI NMR spectroscopy Clay minerals and pillared clays have been the b· f I I'd NMR . .. 1718 su ~ect 0 severa so t state mvesttgattons . .

27 AI NMR can detect the co-ordination of AI atoms in clays containing as little as 0.26% of A1 20 3.

Therefore, this technique is of particular use in

1776 INDIAN J CHEM, SEC A. SEPTEMBER 2005

stud yi ng octahedral and tetrahedral sites in the extra framework AI spec ies contained in aluminium pillared samples. Montmorillonite shows two resonances: one at +1.38 ppm that can be ascribed to octahed ral AI atoms and the other at +66.0 ppm attributable to tetrahedrall y coordinated AI atoms. For aluilliniulll pillared samples, 8 ppm values for octahed ral AI atoms shifts to + 1.98 ppm. This is due to overlapping of signals of AI atoms of intercalating polymer wi th that of orig inal signal. Two resonances can be di stingui shed at tetrahed ral reg ion: resonance at +66.4 ppm belonging to clay tetrahedral sites and resonance at +60.3 ppm corresponding to tetrahedral sites in the pillar. An important conclusion is that the intercalated spec ies is not transformed into a spinel li ke structure. When aluminium hydroxides like gibbs ite and boehmite are the pillars , corresponding resona nces are between +8 and +9 ppm and above +68 ppm, signi fy ing spinel-l ike structure. FLII1her, no resonance is observed at around +30 ppm demonstrating the absence of pentacoordinated AI atoms. Thus, it can be unambi guously concluded that pillared spec ies is the Keggin AIJ3 polymer for Al pillared systems. The Kegg in cation contains a central AI tetrahedron linked to 12 AI octahedra and has the formu la IA IO.jA ldOHb(H20)d7

+.

The effect of transition metal exc hange on structu ra l stabi lity of pillared clays was exa mined by taking 27 AI NMR spectra of copper and cobalt doped aluminium pillared systems as representative systems. As a result of exchange with transition meta ls, peak width as well as peak positior:s do not vary. This shows that incorpora tion of transit ion metals does not affec t the structural stability of layers and pillars. Hence 27 AI NMR data supports the inference drawn from surface area and pore vo lume meas urements, that metal ox ides are incorporated into porous network rat her than aitached to pill ars.

2~S i NM R sperl roseopy

For Si nucleus (spin I = 112) , chemica l shift is affected mainly by the electron density on oxygen atoms of Si tetrahedron. Therefore, nature of nei ghbouring atoms coordinated to these oxygen atoms can influence the shi ft. The spread of these shifts is not on ly a function of the st ructural di sorder but depends also on nature of the second neighbours. The connectivity of Si nuclei in si li cate spec ies is described by the usu al Q" notati on. Q represents a silicon nucleus connected to four oxygen formlll g a tetrahedron. The su perscript n is the number of other

Q units attached to the central SiO.j tetrahedron. Thus QO denotes monomeric orthosi li cate anion SiO.j-. QI denotes end groups of chains, Q" denotes middle groups in chains or cycle, Q3 denotes cha in branching sites and Q.J, three dimensionally cross linked groups. Montmorill onite shows a single resonance centred at -93 ppm. The peak shows a main peak at -93.98 ppm and two shoulder peaks; a large one at - 104.5 ppm and a diminutive one at -90.3 ppm. The peak at -93 .98 ppm can be attributed to Q 1 (Si J AI) units representing Si(IV) atoms linked through oxygen atoms to three other Si(IV) and to one AI(VI) (or Mg) in the clay octahed ral layer. The shoulder peak at - 104.5 ppm can be ascribed to Q' (S iOA!) whr:- re Si is linked to Si onl y through oxygens, wh il e the sma ll peak at -90.3 ppm is due to Q' (Si2A I). Thus , majority of the silicon tetrahedra is linked to 3 Si atoms and one AI atom. A very small porti on is linked to 2 AI atoms and 2 Si atoms, while a part of the Si tetrahedra IS linked to Si atoms alone. Th is di stribution of silicon tetrahedra into va rI OUS environments is not affected by pill :lring. A sli ght shift in ppm values was noticed for pi llared samples. Data show that Lowenstein rul e, wh ich states that two tetrahedral AI can not be next neighbours, is obeyed.

In version of silicon tetrahedra can occ ur with interca lation of polymeric species in c lays like beidellite and saponite where the layer charge is loca lised in tetrahedral layer l9. Such an inversion is not anticipated in montmorillonites since layer charge is not loca lised. 29Si NMR spectra confi rm thi s poin t since the shift in 8 ppm va lues is less tha n 1- 2 pplll ill a ll pillared samples. The contributi ons of different Si environ men ts also remain the sa me. Th is conl~rms the nonex istence of chemical bond s between exchanged polymeric species and clay layers. Thus. as anti cipated, pilbring which is on ly a cation exchange process does not affec t the short rJnge order within clay layers. If any bond that fornr AI-O-Si lin kage between the pillars and silicate layers been formed. the spectra would ha ve conta ined Q3 (Si3AI ) resonance.

The effect of exchange with tranSItI on metal s on the structural stability of clay as \Ne ll as pillaring process was exposed by tak ing the 29Si NMR spectra of two of the exc hanged systems of the alumin ium pill ared series. viz., ColAI PM and ZnlA I PM. It can be inferred that 8 ppm values do not alter mu ch as () result of exchange with transition metal s. Thus incorporation of transiti on metal s on the pillared

KURI AN & SUG UNAN: SELECTI VE BENZYLATION OF BENZENE OVER ALU MIN A PILLARED C LA YS 1777

Tab le 3 - Acid site di stribution/ mass of aluminiu m pill ared series

Weak Medi um Strong Cumul ati ve

Catalyst (35-200°C) (20 I-400°C) (40 I-600°C) (mmolg-I)

VIAl PM 0 .52 1 Mn/AI PM 0 .530 Col AI PM 0.655 Ni/AI PM 0.669 Cu/A I PM 0.5 14 Zn/AI PM 0 .53 AI PM 0.652

syste ms does not a lter the local e nvironme nt o f Si atoms. Hence, it can be infe rred that metal ox ides a re not in the immedi ate environment o f Si layer. They may be present in the porous network of the pillared system. Thi s has been e videnced by surface area data and 27 AI NMR spectra.

Temperature progI'ammed desorption of ammonia It is generally recogni sed that ammonia is an

excell ent probe molec ule for testing acidic propelties of solid cata lysts as it s strong bas icity and small molec ular s ize all ow detecti on o f acidic s ites located in very narrow pores al so. Although, there is widespread use of TPD in the studies of surface ac idity, NHr TPD spec tra a re often poorl y reso lved . However, the procedure is a standardi sation method since ammoni a a llows the de te rminati on of both protonic and cati onic acidities by titrating acid strengths o f any strength .

Ammoni a adsorptio n on pillared cl ays can be ph ys ica l (!1 H ;::; 13 Kea lmo r l) or che mical (!1H ;::; 33 Kcalmo rl ) type. Acid site di stribution profiles show the presence of weak (ammoni a desorbed between 35-200°C), medium (201-400°C) and strong (40 I-600°C) ac id sites. The ac id site di stribution at various temperatures has been calcul ated as functions of mass of the sample .

The ac id ic structure of aluminium pillared cata lysts as obtained from NH}-TPD is presented in Table 3, Exchange with transiti on meta ls does not improve the ac idi ty and weak and medium ac id sites predominate the strong s ites. T hus, a good di stribution of weak and strong ac id sites is present. Acidity in the medium strength region decreases as a result of incorporati on of metals. Acidi ty expressed as a func ti on o f surface, has va lues comparabl e to AI PM . It can be conc luded that an e ffec ti ve di stributi on o f acid sites of varying strength is present in the prepared sys te ms. Consequently , they ca n be used as suitable cata lysts in ac id cata lysed reactions. Exchange of transitio n

0 . 133 0 .223 0 .894 0 .248 0 . 100 0.868 0 .394 0 .079 1.1 28 0.334 0. 192 1.195 0.253 0.058 0 .825 0.427 0. 169 1.1 26 0.441 0. 105 1.1 98

meta ls blocks the hydroxyl groups in the structural framework, reduc ing the Bronsted ac idity .

NHr TPD me thod does not di sc rimin ate the type o f acid s ites (Bronsted and Lew is) . However, it is generally accepted that e vacuati on of a mmo ni a adsorbed at 400°C remo ves most o f the Bronsted ac id sites20

. For pillared cl ays, it has been doc umented that a mmoni a adsorbs in Bro nsted s ites at temperatures a round 250 °C (ref.ll ). Aga in , it is implied tha t coordina ti ve ly bound a mmoni a o n strong L e wi s s ite can be desorbed o nl y at hi g h te mpe ratures and he nce acidit y in strong reg ion can be corre lated to the amo unt of Le wi s s ites. In pillared clays, Le wi s acidity is co ns idered to ori g inate fro m pill a rs whe reas Bro nsted ac id ity ari ses fro m struc tura l f ra mework o f aluminos ilicates. Pore vo lume va lues indi cate presence of exchanged cation s ins ide the poro us ne two rk . He nce, inc rease in numbe r o f strong sites can be attributed to contributi on of these cati o ns in pillars. Thi s is substantiated by the fac t that a mo un t of medium acid s ites (corre lated to Bronsted ac idity) , an s ll1g fro m struc tura l frame work dec reases as a res ult of trans iti on meta l exc hange . Thi s may be due to shi e lding o f these s ites by the depos ition of meta ls in the po res, nea r the structura l framework .

Cumene cracking C umene is a conventional mode l compound for

testing catalytic activity since it undergoes di ve rse reactions over different types of ac id sites. Maj or reac tions taking pl ace during cracking of cumene are dealkylatio n or cracking to benzene and propene and dehydrogenation to a -methyl styrene, Sma ll amo un ts of ethylbenzene and to luene can be formed by cracking of side cha in , whic h on dehydrogenati on gives styrene. C rac king o f cumene is genera ll y attributed to Bronsted ac id sites by a carbonium ion mechani sm whereas a-methy l styrene is fo rmed on Lew is acid si tes2

l.22 .

1778 IND IAN J CHEM, SEC A. SEPTEMBER 2005

Tahlc 4 -Catal ytic act ivity of'various systcms towards cumcnc cracking. [Temp: 400°C; WHSV : 7 h· l; time on stream: 2hl

Selectivity (%) Catalys t Conversion (%) a-Methyl styrcnc Dcalkyl ation products LClVis/Bronstcd

VIA l PM 28.8 5S.S MnlA I PM 26.0 55.4 ColAI PM 3 1.4 50.S NilAI PM 33.6 56.S Cul AI PM 19.2 44 .9 Znl AI PM 26.5 W I AI PM 28.5 5S.6

The catalyti c performa nce of AI PM and the effect of transiti on metal exc hange on pillared clays are given in Tab le 4. Al l systems show considerable acti vi ty towards cumene cracking under the spec ified cond itions . Appreciable selectiv ity towards the dehydrogenated prod uct is obtai ned for all cata l Ethy lbenzene and styrene were formed in small quantiti es and in some cases toluene al so was detected. All the dea lkylated products are grouped toge ther and Lew is to Bronsted ratio gives the ratio bet ween the dehydrogenated and cracked products. Aluminiu m pill ared c lay ex hibits hi gh ac tivity towa rds the reaction with Lewi s/Bronsted ac id rati o of 1.42. Exchange with transit ion metal s results in reduction in acidi ty except for nickel and cobalt exchanged systems as indicated by convers ion of cumcne. A critical scruti ny of Table 4 shows that the dec li ne in acid sites is mainl y due to decrease in the number of Lewis acid sites.

i3cnzylalion (If benzene with benzyl chloride

Benzy lat ion of benzene is an important probe react ion to chec k the acidity of the catalysts since the rlllg itsel f is not ac ti va ted or deacti vated by subst itut ion. Moreover, the reaction gives diphcnylmet hane, an important synthetic intermed iate as the product. The reacti on always gave monoalkylated product regard less of temperature, ime and reactant mo lar rati o. The catal ytic

perfo rmance of the prepared systems towards benzene benzylation is given in Table 5. Al PM shows a cOllversion of 12% in 1 hour under refluxin g condi ti ons. However, catal yti c ac ti vi ty increases as a re LIlt of transiti on metal exc hange. VIAl PM exhibits 100% conversion of benzyl chloride in I hour. A 100% selectivity to monoalkylated product is not changed with any catalyst. The three dimensional porous Iletwork or pil la red clay systems brings about the react ion between the aromatic species and benzy l chloride in a shape selective manner, leading to cent percent d i phenyl methane selecti vity. Presence of L, Li):~ ngcd cation increases the cataly i,: efri ciency.

41.2 1.42 44.6 1.24 41.7 1.22 43.2 1. 22 55. 1 O.S I 41.1 1.44 4 1.4 i-+2

Tablc 5 - Activi ty or alum in ium pillared ,ystcm, l'or hcnzy lati on o r bcnzene

[Temp: soac; henzcne/ BC: 5: limc: I hour: catal yst/BC: 0.15 151

Catal ys t Convcrsion ('If) Select ivit y ( r.-i )

VIA l PM 100 100 MnlAI PM 72.3 100 NilA I PM 48.5 100 ColA I PM 69.S 100 CulA I PM 24.S 100 ZnlA I PM 44.5 100 AI PM 12.0 100

100

\ 70

80

~r ~

~ ~

6O~ ~ 60 a;

VI C

~ 0 '0; OJ CD ~ > c 40 VI

8 >. 50 :=

~ 20

b

0 40 ~ ~ ~ ~ ~ ~ ~ 0.. 0.. 0.. 0.. 0.. 0.. 0..

~ ~ ~ ~ ~ ~ ~

> :::,

~ ~ z ()

Fi g. I - Correl ation of convcrsion (%) lVith (I-Illet hyl styrene se lccti vity.

1n the present study, an attempt has been made to correlate cata lytic activity with the amount of Lew is acidity obtained by the two independent acidity es timation tec hniques elaborated earli er. Lewis acidity as indicated by a-methyl styrene se lecti vity in cumene cracking reaction is in line with catalyt ic performance of aluminium pillared systems (Fig. I).

Figure 2 shows that catal yt ic act ivity along the series fo ll ows the order of the amount of strong acid sites (ammoni a desorbed in the tempe rature range of

KURIAN & SUGUNAN: SELECTIVE BENZYLATION OF BENZENE OVER ALUMINA PILLARED CLA YS 1779

100 \ 0.3

80 IJ)

~ IJ)

~ 0.2 ~ e.... 60 0

'" c: Cl 0 c: '0; e Qj en > c: 40 '0 0 0 0.1 C

:J 0 E

20 <{

0 0 ~ ~ ~ ~ ~ ~ ~ Cl. Cl. Cl. Cl. Cl. Cl. Cl. ~ ~

~ ~ ~ « ~

3; c :::, ;G Z :l

~ 0

Fig. 2 - Dependence or activi ty on amount of strong acid sites from NHr TPD.

40 I-600°C). Greater effici ency of V/AI PM is due to increase in amount of strong acid sites . Thus, conclusive confirmati on of the relat ionship between the amount of strong/Lewis acid sites with catalytic acti vi ty has been ev idenced. Thus, strong acid sites may be considered to be in volved in the benzy lation of benzene with benzy l chloride. Though the pillared clay surface prov ides both Lewis and Bronsted acid sites. the above observations clearly indicate the dominating impact of Lewis/s trong ac id sites for the benzy lati on of benzene with benzy l ch loride. From va ri ous spectroscopic ev idences it has been concluded that the exchanged metals are depos ited in the pill ars. Literature suggests that the Lewis acidity in pillared clays ori gin ates mainl y from the pillars and hence the mod ifica ti on of the pill ars with meta l ox ides improves the catalytic activity.

St ructural stability of the catalys ts

An essential requi site of a heterogeneous catal yst is the stability of act ive sites under the reaction conditi ons. Hence, stability of the catal ysts was checked by metal leac hing and moisture adsorption ex periments.

Leac hing of metals from the cata lyst surface can occu r without much transfo rmati on in the reaction profi Ie, gradual I y changing the nature of reaction from trul y heterogeneous to partl y homogeneous23

.

The probability of aluminium leaching is remote and has not been reported so far. In the present study, metal leac hing was studied at 35 min by continuing the reaction for furthe r 25 min , after filtering off the

catalyst. No noticeable change was observed even after 25 min of reaction , in the absence of catalyst. Hence a trul y heterogeneous reaction can be envisaged with aluminium pillared systems.

In order to test the effect of moisture on cata lys t performance, the catalyst and the substrate were saturated with water vapour. by keepi ng them over deioni sed water in a dess icator for 72 hours at room temperature. The induction peri od for the reacti on increased in presence of moisture. After thi s short span of time (l hour), the reacti on proceeded as in moisture-free conditions. Presence of inducti on peri od is suggestive of water molecules occ upying the acti ve sites prior to reaction . from where cr ions di spl aced them. The hi gh induction period ca n be attributed to the hi gh activation energy over these cata lysts.

Effect of reaction varia bles Influence of va ri ous reaction variab les like time.

benzene to benzy l chl oride ratio and substrate and cata lyst concentrati on on the extent of reacti on was studied ex tensive ly using Mn/AI PM as reference in order to arri ve at a plausibl e mechani sti c path way for the reaction . Study shows the reaction rate to be nominal for a particular period of time and then continually increas in g with time. Absence of pol yalkylated products is a significant res ult. The cont inual increase in percentage conversion with time is indicative of heterogeneous nature of the reaction. Selective formation of the monoalkylated prod uct can be attributed to the porous two dimensional structure of the pillared clays res tricting the attack of the benzyl cati on on the bulky product spec ies.

Benzene to benzy lati ng agent molar rati o was studied by tak ing appropriate amounts of benzene. keeping the amount of catalyst and benzy l chl oride fixed. An inverse relationship was observed between benzene to benzyl chloride molar rati o and conversion % at any given time. Higher amou nts of benzene in the reac ti on mixture prolonged the time required for complete conversion. It is worth menti oning thal polyalkylation does not occur even at hi gh concentrations of benzy l chl oride. Since benzene is taken in excess, the reaction is supposed to follow pseudo unimolec ular mechani sm. The refore, the reaction should have an equivalent relat ion with the amount of benzy l chloride. Thus, the cata lyst is very effic ient for the creation of benzy l carbocations even at hi gh concentrations of the reagent.

The effect of catalyst concentration on the reac ti on was studied by varyi ng the amou nt of catalyst added.

1780 INDIAN J CHEM, SEC A, SEPTEMBER 2005

keeping the amount of reactants constant. Results show that the presence of catalyst even in trace amounts has a marked difference in the product yield. Increase in cata lyst concentration , enhanced the pe rcentage conversion and at a stage, rate of the reaction levelled off. After thi s, increase in ca talyst concentration had no pronounced effect on the reaction rate. Thus, onl y a sma ll amount of cata lyst is needed for easy completion of the reac ti on. Cent pe rcentage selectivity to monoalkylated product even at high concentrations of the cata lyst is commendable. Increase in reaction rate with cata lyst concentrat ion is sugges tive of heterogeneous mec hani sm for the reaction. Higher amounts of catalyst result in higher amounts of active sites and hence the inc rease in reaction rate. Since the sys tems under study are V".)

efficient Friedel-Crafts catalysts, only a sma ll amount of catalyst is needed.

The effect of substrate on percentage conversion was studied with different substrates. The observed order of reactivity of substrates is a-xylene> toluene > benzene. As expected , the reaction rate for hyd rocarbons was lowered over aluminium pillared sys tems and halobenzenes did not react at a ll over

(interlamellar)

dipheny! methane.

these catalysts under the specified condi tions . From the resu lts it can be inferred that the act ivating and deactivating effect of the substi tuting groups in the benzene ring has profound influence on reaction rate . The observed order of reacti vity is exac tly the same as that of e lectron releasing effect of the substituting group in the benzene ring. The induct ive effect of methyl group makes the reaction more facile in the case of toluene and sti ll higher for xy lene due to the cumulative effect of two methyl groups. Similar res ults have been reported by Jun et (/ 1.2~. The hi gher activat ion energy due to e lectron withd rawing effec t of the ha lo group is respon sible for the non occurrence of the reaction over these cata lysts.

Mechanism of the reaction Severa l mechanisms have been put forward by

various authors for benzylation with benzy l chl oride. Predominant among them are the class ical carbocation mechanism and the redox mec hani sm involving free radicals . Yadav et ai. investi ga ted the a lky lation of toluene by benzyl chloride on sulphated zirconia, a very strong Lewis acid and reported a surpris ingly low activit/5

. Hence, a redox mec hani sm was suggested for the reaction .

o 6 6

pi complex

pi-sigma 11 rearr.,gement ~

~:..-o:

-M t sigma complex

Scheme 1

KURIAN & SUGUNAN: SELECTIVE BENZYLATION OF BENZENE OVER ALUMINA PILLARED CLA YS 1781

Based on the observations of present study , a plausible reaction mechanism is suggested for the reaction (Scheme 1). The mechanism is similar to the mechanism of alkylation reaction catalysed by an M+ -H species present on synthetic transition metal oxide Si02-AI 20 J systems26

.27

. The alkylating agent, benzyl chloride interacts with the active species and forms the alkylating moiety M+-CH2-C6HS. which in turn attacks the benzene molecule forming a TC

complex. This is possible due to the partially filled p­orbital s of AI (III) and d-orbitals of Fe(lII) present in the pillars. The framework AI and Fe does not contribute to the formation of complex or pure montmorillonite would have shown much better activity. The TC-complex rearranges to give the alkylated product diphenylmethane. The platelets of clay intercalated with transition metal oxide pillars are capable of stabilising highly polar transition states converting it into products whereas it is formed in lesser amounts or even not obtained in homogeneous conditions28

. Thus, the intermediate of the alkylating moiety, the TC complex and the sigma complex could be stabilised which in tum enhances the conversion to diphenylmethane. When the reactants are constrained to diffuse in a porous solid, which have layered structure like clays, their encounter frequencies increase. Also, organic molecules congregate in the compartment like structures of the clay matrix . Thus, pores locally IIlcrease the interaction bet ween reactants29

.

Acknowledgement Financial assistance from CSIR, New Delhi to MK

is gratefully acknowledged.

References I Kloprogge J T, J Porolls MareI', 5 (1998) 5. 2 Gil A, Gandia L M & Vicente M A, Caral Rev Sci Ellg, 42

(2000) 145 .

3 DeWitt L A, Su H & Mathew C T. US Parellf 5.723.676 ( 1998).

4 Olah G A, Friedel-Crajr:; Chellli:;rrv. (Wiley . Ne w York) 1973 .

5 Ratnaswamy P, Singh A P & Sharma S. Appl Caral II : (iell. 135 ( 1996) 25.

6 Chaudhary V R. Jana S K & Kiran B P. J Caral. 192 (2000) 257.

7 Moreno S, SunKou R. Molina R & POllce lot G . .I Coral. 182 ( 1999) 174.

8 Occelli M L, Bertrand J A. Gould SAC & Dominguez J M. Microporolls Mesoporolls MareI'. 34 (2000) 195 .

9 Rutherrord D W, Chiou C T & Eberl D D. Clays Cia." /II/iller. 45 ( 1997) 534.

10 Bakas T, Moukarika A, Papae fth ymiou V & Ladavo> A. Clays Clay Mill er. 42 (1994) 634.

I I Rausch W I & Bale H D . .I Chelll Ph),. 40 (1964) 3391 . 12 Helmy A K. Ferreiro E A & deBussetti S G. Clays Cia\'

Miller, 42 (1994) 444. 13 Madejova J, Vibrariollal Specrrosc . 3 1 (2003) I. 14 Bodoardo S, Chiappetta R. Onida B. Figueras F & Garrone

E, Microporolls Mesoporolls Morel', 20 ( 1998) 187. 15 Tichit D, Fajula F. Figueras F. Ducourant B. Mascerpa G.

Guegen C & Bousquet J, Clays Clay Mill er. 36 ( 1988) 369. 16 Li L, Liu X, Ge Y, Xu R. Rocha J & Klinowski J. J Pin'

Chelll, 97 ( 1993) 10389. 17 Occelli M L, Aurox A & Ray G J. MicroporollS Mesopol"OlIs

Maler, 39 (2000) 43 18 Plee D, Borg F. Gatineau L & Fripiat J J. J 11111 ChelJl Soc.

107 ( 1985) 2362 19 de Lucas M C M, Rodriguez F. Prie to C. Verdag uer M &

Gudel H U. J Phy Chelll Solids, 56 ( 1995) 995 . 20 Misra T & Parida K M, IIppl Cu ret! A : Cell . 174 ( 1998) 9 1. 21 Mokaya R & Jones W, J Caral, 153 (1995 ) 76. 22 Ghorpade S P. Dharsane V S & Di xit S G. Appl CW(i/. A:

Cell , 166 (1998) 135. 23 Jun S & Ryoo R . .I Catal. 185 (2000) 237. 24 Yadav G D, Thorat T S & Khumbar P S. Terrahedroll Lefl.

34 (1993) 529. 25 Sabu K R. Rao K V C & Nair C G R. /3111/ CI,elll Soc .lapall.

64 ( 1991 ) 1920. 26 Sabu K R, Rao K V C & Nair C G R, /3111/ Chelll Soc Japall .

64 (1991 ) 1926. 27 Laszlo P, Pllre IIppl Chelll, 62 (1990) 2027. 28 Laszlo P,Acc Chelll Res, 19 ( 1986) 12 1.