Indian Journal of Chemistry Vol. 4 1 A , November 2002, pp. 2244-2250

Synthesis, characterisation and benzylation activity of vanadia impregnated iron pillared montmorillonite

Rahna K Shamsudeen, K Nisha & S Sugunan*

Departmc: nt of App li cd Chemistry. Coch in Uni versit y of Sciencc and Technology, Coehin 686022. Kc:ra i<t. India

E-mail: Ssg@CusaLac. in

l<eceil'l!d I Mav 2002; rel'ised 22 Jill." 2002

Imn pillarc:dmoillmorili onit c: has bc:en synthcsiscd and i t is then wet imprcgnated w ith vanadia w ith difk rent vanadia composition. Th~se ca talysts arc characterised using con vent iona l techn iques such as X RD ana lys is, FTIR analysis. and surface arca and pore vo lumc measurements. Acidity is Illeasured using spectrophotometri c monitori ng of adsorption of pery lcne. thermogravimetr ic desorpti on of 2,6 dimethy lpyr idine and temperatu re programmed desorpt ion el r amilloni a. Acti vity studies arc clone in the li quid phase. It has been concluded that Lew is acid ic sites are responsible for tlie bcn7y lation or tol uene when the benzylating agent is benzy l chloride while Bronsted ac idic sites are responsib le whc:n the reagent is benzy l alcohol.

Supported vanadium ox ide cata lys ts are important in

industrial processesH. Use of supporting vanadium

oxide on another ox ide has several ad vantages over an unsupported oxide, such as higher mechanical strength, better thermal stabil ity and larger surface are</ There are several reports reg;rding the di spersion of vanad ium oxide over basic, amphoteri c and acidi c oxides5

.7

. [ t is poss ible that the di spersion of vanadium ox ide as wel l as the structure can be understood on the bas is of acid-ba e character o f the supports used. The basic and amphoteri c ox ides fa vour the bidimensiona l dispersion, of ten with the formati on of vanadates. T he agg lomeri sa tion o f vanad ia speci es to fo rm crys talline vanad ia IS

fa vou red with the acid charac ter of the support. Except a few repo rt s ~ .') , work on vanadi a impregnati on

over clays tS very rare. Over acid act ivated montmori ll onite, K 10, vanadia load ing by wet i mpregnat ion method was attem pted and compared wi th vanadia-s ili ca by Narayanan and coworkersH

They have inferred that the type of interaction of vanad ia w i th K I 0 montmori 1I0ni te is d i fferent from that w ith silica. Yanadia i s we ll d ispersed as Y~05

crystallites in both the supports. An interacti on of vanad ia w ith the support occurs in the case of sili ca whereas such interaction is absen t in montmorillonite K I O. A n interac ti on between vanad ia and the support leads to the hyper fine splitting o f ES R spectra, and is observed wi th the case of silica but not with K 10 montmori ll onite. [n our present inves ti gati on, we have prepared i ron-pi llared c lay and van ous vanadia

loaded iron-pi ll ared clays by the wet impregnation method. A ll the samples were characteri sed employ ing the tech niques such as EDX, X RD, FT-IR, surface area and pore vol ume measuremcnts and acid ity determinati on studies by pery lene adsorption technique, thermogravi metric desorption of 2.6-dimethylpy ri dine (DMPY) and temperature programmed desorpti on (TPD) of ammoni a. These so lids were tes ted for activity in the Friedel-Crafts benzy lati on of to luene. Friedel-Crafts alky lation reactions are very important from sy ntheti c and industri al po ints of view. These are important means for attach ing alky l chains to aromatic rings. On behalf o f the stri ct env ironmental regul ation s :lI1d drive towards c lean technology , the present tendency is to replace the conve ntional technology using homogeneous cata lys ts by env i ronmentall y fri endl y cata lys is, typica ll y invo lving the use o f so l id aci d

I I i) ca ta ysts .

Mate."ials and Methods Preparation oliron -pillared and vOlladia iouded irnl/ ­pillared I/ lOntlllorilloniles

T he starting material used for the work is KSF montmori llonite obtained from Fluka. Thi s matenal has a chemica l composition (wt %) of Si-59.57%, AI-2 1.05%, Fe- IO.73%, Ca-3.48%, K-1. 86% Mg-2 .32(k and Na-0.97% and a surface area o f 6.6 l11 ~g - l . l ron­

pillared montmorill on ite (FePM) was preparcd by stirring a I % suspension or' the parent montmorill onite in di stilled Welter added to the

SHAMSUDEEN el at.: VANADIA IMPREGNATED IRON PILLARED MONTMORILLONITE 2245

pillaring solution which is previously prepared by partially hydrolysing 0.2 M ferric nitrate solution with 0.3 M sodium carbonate at a base/metal ratio of 2.0 under vigorous stirring at 70°C for two hours followed by continuous stirring for six hours at room temperature. Fe3

+ to clay ratio was maintained as 20 mmol g- I of clay in the final reaction mixture. After reaction, the material was washed until flocculation appeared and it became nitrate free. It was then dried at 110°C, followed by controlled calcination in dry air at 400°C for five hours to get iron-pillared montmori 1I0nite catalyst. Di fferent percentages of vanadia were loaded over FePM according to the wet impregnation method II. Calculated amounts of ammonium metavanadate corresponding to 2, 5, 7, 10, 15 and 20 weight percentages of vanadia were di sso lved in aqueous oxalic acid solution and stirred with iron pi lI ared montmori 1I0nite for 3 hours followed by solvent evaporation on a water bath . Thcse were then oven-dried overnight and calcined at 400°C for fi ve hours. The samples are denoted as X VFe, where X represents the percentage of vanadia loaded over FePM.

The percentage of vanadia was determined by EDX spectroscopy performed over an EDX-JEM-35 instrument (JEOl Co. link system A -1000 Si-li detector, sensiti vity> 0. 1 wt%). The XRD patterns of the prepared sa mples were recorded usi ng Ri gaku Model D/Max C instrument. FTIR spectra of the powdered samples werc measured by the KBr di sk method over the range .. 1-000-400 cm- I. Shimadzu DR 8001 instrument was used for the purpose. Thc surface area of the catal ys ts was determined by BET nitrogen adsorption at liquid nitrogen temperature using a Micromeritics Flow Prep 060 instrument. Prev iously activated samples were preheated and degassed at 473 K for two hours under nitrogen now. The catalyst was then brought to 77 K using liquid nitrogen for adsorbing nitrogen gas at va ri ous pressures. The total pore vo lumes of the sdmples were also measured using the same instrument. This was measured by the uptake of nitrogen at a relative press ure of 0.9. Acidity was determined by three independent methods: (i) Thermogravimetri c desorption of DMPY .. Prev iously activated catalysts were kept in a desiccator saturated with vapours of DMPY at room temperature for 48 hours. Then the catalysts were subj ected to thermogravimetric analysis when the weight loss of the adsorbed sample was monitored for the range 40 .. 500°C at a rate of 20°Clmin. The fraction of weight loss in the range 250-500°C was

found out and taken as a measure of the Bronsted acidity of the samples 12-15. (i i) Perylene adsorption studies-Different concentrations of perylene in benzene were prepared and stirred with about 0.5 g of the activated catalyst at room temperature for four hours. Due to adsorption of perylene on the catalyst from the solution, its concentration in the so lution will be less after adsorption. The di fference 111

concentration before and after adsorption was determined by means of a Shimadzu UV -VIS spectrophotometer. The ahsorbance was measured at a Amax of 439 nm. (iii) Temperature programmed desorption of ammonia-About 0.75 g of the previously activated cata lyst was degassed by heating in a stream of nitrogen at 300°C. It was then cooled to room temperature. The ammonia was allowed to be adsorbed on the catalyst by injecting 15 ml of ammonia gas at room temperature into a stainl ess steel reactor of 30 cm length and I cm diameter in which the catalyst was kept. The catalyst was then heated through a temperature programme in nitrogen fl ow. At each interval of 100°e, the ammonia desorbed was trapped usin g nitrogen as the carrier gas in a stoppered conical flask containing a know n excess of 0.025 N sulphuri c acid for the temperature range IOO-600°C. It is estimated by the back titration of the excess sulphuric acid using standard NaOH . The amount of ammoni ~ desorbed was di stributed in the weak plus medium strong ( I 00-400°C) and strong acid regions (400-600°C) .

Activitystlldies Friedel-Crafts alk ylation of tol uene with two

different benzylating agents such as benzyl chloride and benzy l alcohol was carried out in the liquid phase in a 50 ml R.B. flask equipped with an oil bath , a magnetic stirrer and a water condenser. The reacti on mi xture was analysed at required time intervals using a Chemito GC-8610 equipped with an SE 30 column and flame ionization detector.

Results and Discussion

Methods oIcharacterisatio/l .. XRD a/ld IR studies Table I shows the XRD detail s of FePM and all the

vanadia impregnated iron pillared systems along with those of pure V 20 5 . The characteristic 28 value of the original mon tmorillonite (corresponding to a layer distance of 9.82;\) is shifted from 9 to 5.72 and the layer di stance is correspondingly increased to 15.48;\ for FePM. The other peaks at 19.2 and 25.7 are due to

2246 INDIAN J CHEM . SEC A, NOVEMBER 2002

Tab lt: I - XRD data of iron -pillared and vanad ia loaded iron­pillared Illontillorili onites

Catalyst 20 d spac ing Relati ve Intensity 5.72 15.48 69.2 1

FePM 19.7 4.5 1 100 25.2 3.53 30. 10

5.7 1 15 .50 5 1.50 2V Fe 19. 8 4.49 100

25.1 3.55 27.89

5.72 15.48 38. 10 5V Fc 19.75 4.50 100

25. 1 355 26.24

5.7 1 15.50 26.4 1 7V Fe 19.66 4.52 100

25. 15 3.54 28.0 1

19.66 4.52 100 25 . 15 3.54 17.8 1 26.10 3.4 1 26. 10

10VFe 32.05 2.79 24. 10 35.65 2.5 1 28.10 54 .01 1.69 24.0 1

19<)1 4.46 100 25.48 3.50 16. 10

15VFe 26.1 8 3.40 2809 32. 14 2.78 24.5 1 35 . 15 2.55 3 1.0 I 54.08 1.69 2208

19.76 4.5 1 100 25. 15 3.54 19.62

20V Fe 26.04 3.4 1 3 1.04 32.28 2.77 27.9 I 35.65 2.5 I 44.4 1 550 1 1.66 28.04

20.05 4.43 100 26. 20 3.40 63.87 32.45 2.76 29.3 1 34.45 2.6 45.22 55.75 1.64 15.52

the qu artz impurity. It is well-known that loading of vanadia over bas ic oxides results in compound formation such as orthovanadate or pyrovanadate format ion5

.6 But with acidic ox ides, no such

compound formation is ex pected. Instead, V20 S

aggregates di spersed over the sol id surface are observed 7

. Clays, especially pi lI ared clays are well­known as acidic ox ides. Hence di spersion of V20 S

over the so lid surface is ex pected. From Table I, it is clear that the add iti on of vanadia up to 7 wt% on iron­pi lIared montmori 1I0ni te, did not show any charac teri sti c peak of pure V 20 S' However, with the increased load ing of vanadia such as 10 wt% and above, peaks characteristic of vanadia start to appear. The intensity of the vanadia peak increases with the vanadi a content and the peak corresponding to the

~ L-__ ~ __ ~ __ ~ __ +-__ 4-~ __ ~~ u C ~---------------------------d .0 ~

o

'" .0 <{

1000 2000 3000 4000

Fig I-A. IR spectra of (a) parent mon tnlorill onites and (b) FePM. B. IR spectra o f vanadia loaded iron pill ared mont mori l lonites after subtracting that of Fel'M

ex pansion of layers during pillaring in FePM. becomes weak and it is almost disappeared when the vanadia load ing reaches 10 wt%. These facts suggest the di stribution of vanadia on the pillared montmorillonite as small crystallites. However, these are not detectabl e by XRD technique when the vanadia content is 7% or below. The crystallite size of vanadia seems to increase with vanadia load ing beyond 7% and becomes detectable by XRD. Comparison between the IR spectra of paren t montmorillonite and the pillared samples reveals that the basic structure of clay is not at a ll altered during the pi lIari ng process si nce all the mai n peaks are retained in the pillared samples (Fig. 1 A). IR results are also examined to find out the nature of vanadia species over FePM. Figure I B shows the I R spectra of FePM and the I R spectra of X V Fes after subtracting the TR of FePM . This subtracti on is done in order to nullify the sharp peak due to Si -O bond that may overlap on the peak due to vanadi a and thi s enables to find out the modifi cation caused by

SHAMSUDEEN 1.'1 <II .: VANA DI A IMPREGNATED IRON PILLARED MONTMORILLON ITE 2247

Table 2- Van :ldia pen:..:ntage, <;urface area and pore volumes of vanadia Illlpregnat t:d i ron-pillared montmorillonites

Catalys t Va nadia content BET surface area Port: volume (wt %) (m2g- l) (cm Ig- I)

Ft:PM D.OO 188.7 U. 1435

2VFe 2.0 1 124.9 01040

5V Fc 4.98 94.0 0. 10 17

7VFe 740 72 4 0.0732

10VFt: 10.55 36.6 O.037X

15VFe 15.02 26.0 0.0344

20VFt: 20 .0~ 14.6 O.02lfJ

vanadia impregnati on alone, The characteri stic mode of V:,Os at 1020 cm- t (V=O stretching) is show n by all the vanadi a impreg nated FePMs t6 As the percentage of vanadia increases , the intensity of thi s peak also increases. Thus the distribution of V:,Os ove r FePM is further confirmed by the IR results.

Table 2 presents the vanad ia composition, surface area and pore vo lume of the various systems. The pillaring of montmorillonite causes a dramatic increase of surface area from 6.6 [0 188.67 m:' g- I. But the incorporati on of vanad ia results in the continuous decrease of surface area and also pore vo lume. We have already seen the formation of V20 S on FePM whil e treating it with ammonium metavanadate by analysin g th e XRD and IR results. The decrease in sur face area and pore vo lume suggests the intrusion of vanad ia species on the pores of FePM. As the percentage of vanadi a load ing increases, the surface area and pore vo lume decrease due to the greater intrusion of vanadia inside the pores. The surface area of 20VFe is dramatically red uced to 14.63 m2g- l

,

whi ch suggests almost co mp lete blocking of the pores by the vanadi a species. Similar results were obtained by Narayanan et 0/.

17 The broadening of the peak co rres ponding to the pillaring in FePM in the XRD patt ern also supports the entrance of V:,0 5 in side the pores of the pillared clay. However, the possibility of di spersal of V ~Os over the clay surface cannot also be rul ed out.

Acidit v stlldies Le\vis acidity by pen/ene adsorption lII ethod­

Perylene being an elect ron donor transfers electron to the Lewis acid sites and gets itself ad. orbed as perylene radi ca l cat ion IX, Thu s after adsorpti on, the concen tration of perylene in the solution decreased. Thi s decrease was fo und out using absorbance measurements at A max eq ual to 439 nm, whi ch corresponds to th e amount of pery lene getti ng

adsorbed on the surface. which in turn corresponds to the Lewis ac idity of the sys tem. As the concentrati on of perylene used for adsorption is increased , the amount of perylene getting adsorbed on the so lid surface is also increased. But after a particular concentration of pery lene in benzene, the amount of pery lene adsorbed remains constant. This constant value is referred to as limiting amount. The limit ing amount of perylene indicates the surface electron accepting capac ity i.e .. the Lewis acidity of the samples . Table 3 describes the Lewis acidity of the samples obtained by perylene adsorption method. Vanadia is a very weak ac id and ex hibits main ly Brbnsted ac idit/'i-:' I. From the IR and XRD stu dies we have concluded that vanadi a is dispersed as V~Os

in side the pores and surface of the clay latt ice . Now. during the perylene adsorp tion experiments, \-ve observed a signi ficant loss of Lewis sites. It can be confirmed that loadi ng of vanadia and thereby dispersion of van ad ia as V20 S did not enh ance the Lewis acidity of iron pill ared systems, Not only that, addition of vanadia signifi cant ly reduces perylene adsorpti on, due to the blocking of surface acidic sites by vanadi a spec ies .

At low vanadium loaciings of 2 and 5%. Lewis ac idic sites are not very much red uced. But at higher vanadium load ings, Lewis acidity is dimini shed due to the coverage of ca ta lys t surface with parti cles of V:,Os. At low vanad ium load ings, Lewis acidity due to exposed Fe and AI ions is present. Lewis ac idity can be assigned to un saturated vanad ium ions also, due to the adsorbed V20 S on the clay surface to so me ex tent. At high vanadium loadings, crystalline vanadia covers almost the major portions of the clay, and reduced the Lewis acidity .

Deterlllination ql Bronsted acidify by thenJ/ogrm 'i­lIIetric desorption of 2.6-dilllethy /pyridine- DMPY is a useful probe molecule for the selective determination of Brbnsted ac id sites. Severa l researchers applied DMPY as a probe for Brbnsted ac id sites l

:'- 15. DMPY desorbed above 250°C is solely attributed to the Brbnsted acid sites. The se lec ti ve adsorption of DMPY on Brbnsted acid sites is attributed to the steri c hindrance of the meth yl groupS22 . The thermogravimetric desorption res ults or DMPY on the prepared samples arc reported in Table 3. Vanadia addi tion improved the Brbnsted ac idity for the loading of 5VFe, 7VFe and IOVFe, Th is is in agreement with the Brbnsted acidic nature

"

V 0 I() · ~ I A I d' I d" f' o ' :' 5 -. s t 'le vana Ia oa lIlg t ncreases rom 2%, the amount of DMPY clesorhed from the catalys t

2248 IND IAN J C HEM, SEC A. NOVEM BER 2002

Table 3- Acid ity va lues o r different vanadi a impreg nated iron pillared sys tems

Catal yst Limi ting aillount Fract ion of Aci d site di stri butio n by NH, TPD (Ill lllo ig - I)

of pery lene DMPY Wea k Mediulll Strong Total (IO-~ Illlllo ig- I ) desorbed ( 100-200°C) (200-400°C) ( 400-600°C)

FePM 1.2 104 0.024 0.2726 0. 1992 0. 1163 0.588 1

2VFe 1.1 01 0.0 19 0.2985 0. 16 10 0.13 14 0.5909

5V Fe 0.9088 0.039 0.426] 0.4225 0.3726 1.22 14

7V Fe 0.73 16 0.059 0.5095 0.8006 0.6 186 1.92RO

10VFe 0.4 14 0.04 0.3887 0.2682 0.]754 1.0324

15VFc 0.]4:21 0028 0.3207 0. 178 1 0.0534 0.5522

20VFe 0. 2461 0.004 0. 18 10 0. 101 1 0.0190 0. 30 11

also increases . up to 7% and then it decreases. We have already seen thal va nadia is well dispersed over the clay surface up to the weight percentage of 7. Afte r th al, onl y large crys tallites of V20 S are formed over the surface. These crys tallites enter in to the pores and thu s reduce the surface area and also pore volume. It is reported that va nadia load ing creates Bri.insted acidity. assigned to the V-O H spec ies of crystalline V20 /,1. The dispersion of V20 S increases the Bri:insted <K idity cons iderably in the case of 7VFe. However, the fo rmati on of large crys tallites of V20 S

red uces th e Bronsted acid ity or the ca tal ys t. Telllpera/llre progral/l1l/ed desorption of'

al// I/ /O I/ia- The distribution or acid site.' and total acidi ty va lues of the vari ous vanad ia loaded iron pill ared montmorill onites are also shown in Table 3. We ca ll see that the tOlal ac idity va lues inc reases as the vanad ia percentage increases from '2 to 7 and then thcse acidil Y va lues start to decline . Thi ~ trend may be due to the high degree of di spersion of vanadi a over FePM up to 7% and the form ali on of la rge crys tallites at load ings greater than 77r . The large crysta lli tes block the acid sites of the catalyst and hence the acidi ty is red uced. We have already seen the increase in Bronsted acidity and decrea~ e in Lewis acidity upon vanadia addition. Thus we can ass ign the increase in total acidity to the effect or increasing Bri'instcd ac idity. For 7VFe. the total acid ity measured is three times greater than thaI. of the parent iron pillared montmorillonite. Lewi s acidity is dec reased LI S evident from the vei/ues of limiting Llmoun ts of pc rylene adsorbed . Howeve r, w,~ can not ru le out the possibility of NH, molecul e bCl ng ~ id,(ll hell on tho 'ie Lewi s acid s it es whi ch perykl1c mu l 'cuk cou ld no' .. It may be suspected thal the total acid ity value:, are increased (when deten nlllcd by tem pnature programmed desorpt ion of I Hi .-, du e tt' this factor :t1 so. along wi th enhancement or Bronsted ac idity. The cata lys ts up to 10 wt% vanadi a loadin g. possess

Table 4- Data o f benzy lation of to luene with benzy l chl o ri de over va nadia loaded FePM

I Reacti on conditi ons: 0. 1 g catal yst activated at 500°e. benzyl chl oride: to luene molar rati o I: 10, reaction te lilperature 60°C and time of reac ti on - 30 min .]

Catal yst Conversio n (%) Seiec ti vi ty (o/r )

(J-MDPM p-MDPM

Fe PM 86.82 23.50 76.50

2VFe 87.8 1 22. 10 77.90

5VFe 90. 2 1 20.4 t 79.59

7VFc 92. tl 2 1.68 n.32

IOVFe 64.80 19. 15 flO.85

15VFe 50. 10 14. 10 85.90

20VFe 3 1.01 t2.21 87.79

more acidity than the parent iron-pi lI ared montmori lloni te, whereas ISVFe and 20VFe possess lesser num ber of ac idic sites, when determined by N H,-TPD method.

Catalv tic activity stl/dies Bell zyl eli/oride as the reagenr- The products

identifi ed were o-methy ld iphenylrnethane and p-methy ldipheny lmethLlne. The percentage conversion and se lectivities of o- MDPM and p-MDPM over vanadia i mpregnated ~ystems are show n in Table 4. The presence of iro n in the sys tems makes these ca ta l ys ts efficient towards benzy lat ion of tol uene. In addi tion , vanadia 10Llding increased the number of aci dic sites considerably up to a va nad ia loading of 7C!1c (Table 3) . Thi s increase in acidity is reflected in the ir activi ty ,L wt'll. Cata lysts ha \'ing a vanadi a IOLlrli ng: of 7o/c or below are found to be more active I_ h ~ n FePM . These c, ta lysts enjoy hi gh Briinstcd acid ity and the [(\ta! aCidi ty IS al w h i ~h T than paren t ITI'M Others ';how lowcl Jct i·;l ty. Al il lf hcr \'<111 adia load ings, th e pores are almost blocked and thi s res uit s in the large reduction in su rface area and catalyti c

y

SHAMSUDEE N er al.: VANAD IA IMPR EGNATE D IRO N PILLARED MONTMORILLON ITE 2249

Table 5- Benzylat ion of to luene wi th benzy l alcohol over vanadia loaded FePM

Reac tion conditions: 0. 1 g catalyst act ivated at 500°C, benzy l alcohol: to luene molar rati o I: 10 , reac tion temperatu re- I 10°C and time of reac tion - 60 min .

Catalyst Conversion Selec ti vit;t (%) ('Yo ) (J-MDPM p -MDPM Dibenzy l

e ther

FePM 47. 19 5 1.72 33.32 14.95

2VFe 27 .H4 SO. I I 37.58 12.3 1

5VFe 3 1.92 50.65 36.23 13. 12

7VFe 50.2 1 54.62 34. 17 11.2 1

10VFe 4 1.62 39.2 1 42. 15 18.64

15VFe 2 1.11 22.62 37. 10 40.28

20VFe 9.62 18. 19 22 .37 59.44

act ivity. Pore blockage leads to higher se lecti vity fo r p-MDPM than o-MDPM . However, the conversion is very low for sys tems with hi gh vanadi a loading such as 15 and 20%. compared to systems hav ing low vanadi a loadings.

Benzyl alcohol as the reagellt- We have a lso carri ed out the Friedel-Crafts reac tio n using a different benzy lating agent such as benzy l a lcoho l. It is well establi shed that the Brbnsted acid sites present in a catalys t are taking part in the reac tion when the benzy lating agent is benzy l a lcohol2425

. Apart from 0-

MDPM and p-MDPM, dibenzy l ether is also detec ted as a side product in thi s reac tio n which is formed by the se lf condensation of benzy l a lcoho l. Table 5 depicts the conversio n of benzyl a lcohol and the product se lecti vity pattern for the reac tio n between toluene and benzy l a lcoho l. Yanadi a additi on enhanced the Brbnsted acidity of the sys tem with 7% vanadi a loading and the cata lytic acti vity is found to be more than the other loaded systems. For hi gh vanadi a loaded sys tems such as 15 and 20YFe, ve ry low conversion is observed ow ing to their low surface areas.

The acidity determined by TPD method and thermogravimetric desorpti o n of DMPY revealed the regular increase in both , as the loading increases. To gel an insight into the decisive part pl ay ing in both the set of reactions (to luene + benzy l chloride and to luene + benzy l alcohol) we have carried out the moi sture influence for the two reac tio ns. 0.1 g each of fresh catalyst and moisture adsorbed catalyst were put in to the reaction mi xture containing benzy l chloride and to luene with 1: I 0 molar rati o in separate R.B. fl asks and the reaction was carri ed out at 60°e. At regular intervals a small quantity o f reaction mixture was

100

so

If' 60 c: .2 ~ 40 ~

" ;> c: C u 20

0

0

100

~ 80

~ c 60 .'2 V>

<; ;> c 0 40 u

20

A

/~-I . I o~o~o-o

20 40 6() 80 T ime (min )

B

30 40 50 60 70

T ime (min)

- --Fresh ca talyst

100 120

80 90

- 0-- Catalyst adsorbed with moisture

Fig. 2 -A. Conversio ns o f benzy l chlo ride over fresh cata lyst and cata lys t ad sorbed with moisture (0. 1 g FePM, benzy l chlo ri de : to luene mo lar ratio- I: 10, reaction temperature-60°C) and B. Conversions of benzyl alcoho l over fresh catalys t and catalyst adsorbed with mo isture (0. 1 g FePM. benzy l alcoho l: to luene mo lar ra ti o- I : 10 , reac ti on te mperat ure- 11 0°C)

withdrawn from each of the R. B. fl asks and analysed by gas chromatography. The results are illustrated in Fig. 2 A. The percentage conversion gradually increased as a function of time and became 100% at 50 minutes of the reac tion, when fresh cata lys t was used fo r reactio n. But when the moisture-adsorbed catalys t was used fo r reactio n, we have observed only negli gible conversio n even after forty minutes. (At thi s time the conversion is 94 .22% for the fresh cata lys t). The percentage conversion is suddenly raised fro m 5 .29 tc5 69.99% in the time range 40-50 minutes. After thi s time, the convers io n is increased steadily and became 100% after 60 minutes. Thus there is a time period fo r which the cata lys t is inacti ve towards reaction , when it is adsorbed with moisture. The moisture gets adsorbed o n the active sites, (Fe3

+

and AI 3+) present on the interl ayer and surface of the pillared clay and thi s prevents the benzy l chloride molecule to react over these sites. However, after the

2250 INDIAN J CHEM, SEC A, NOVEMBER 2002

inducti on peri od, the reaction proceeds as in the case of fresh catalyst itse lf.

Adsorpti on of moisture on the catalys t do not cause any decrease in cataly tic activity when the reagent is benzy l alcohol (Fi g. 2B). Instead, we noticed the catal y ti c act i vity [ 0 be higher when the moisture­adsorbed catalyst was used. When we u.·ed the catal ys t after adsorbi ng with moisture for carry i ng out th e reac tion between benzy l chl orid e and to lu ene, we have clearly noticed an " induction peri od" to start the catalytic reaction. In the present case, such an induction period is not at all observed. From the results obtained, we can record that moisture has a beneficial rate enhancement effect on the react ion. The enhancement effec t of catal y tic act i vity by adsorbing the catalyst with moisture can be ex plained as fo llows. Moisture is getting adsorbed on the metal ions (Lewis acid sites) present in the catalys t and in effect, these ac t as Brbnstecl acid sites2

(, . By adsorbing on the elec tron defi cient metal cati ons, the water molecules get polari sed and thi s po larisat ion effect makes them effi cient to carry out the reaction26

.n . Thu the effect ive number of

Brbnsted acid sites taking part in the reacti on when the cata lys t is adsorbed w ith mois ture will be greater than those taking part in th e absence of moisture. Thi s leads [ 0 [he higher percentage conversion when the reaction is carried out with a cata lys t, which is adsorbed with moisture. These results indicate the invo lvement of Brbnsted sites in the reaction.

Conclusion Vanadia impregnat ion over iron pillared clay

resulted in the well dispersion of V 20 S crystal s at low vanadia load ings and agg lomeri sation of the spec ies at high vanadi a load ings. Vanadia di spersion increased the ac idity in al l regions up to a parti cular vanadi a load ing, namely, 7 vvt%. However, the Lewi s acidity is reduced cl ue to the block ing of active sites by vanadia species as ev ident from the pery lene adsorption method. M oisture influence for the benzy lati on react ion reveal ed that Brbnsted ac idic sites are responsible for the reaction between toluene

and benzy l alcohol whereas Fe3+ and A l '+ are the

active sites for the reaction between toluene and benzyl chloride. It is also concluded that vanadia addition can enhance the Friedel-Crans reaction.

References I Hucknal D J, SeleCl il'e oxidmioll oj hvdrocorbol/S (Academic

Press, London). 1974. 2 Villadsen J & Li vberg H. Caral ReF Sci Ellg, 17 ( 1978) 203. 3 l3 ielanski A & Haber J. Cawl Rev Sc i El!g. 19 ( 1979) I. 4 Villa P L in Catalysl deaClivOIioll , edited by B Delmon & G

F Fromenl (E lsev ier, Amsterclam). 1980, p- I 03. 5 Blasco T. Lopcz-Nieto J, Dejoz A & Vasquet M I, J Catal,

159 ( 1995) 27 1. 6 Shon J R, Cha S G, Pae Y I & Hayash i S, J Cawl. 159

( 1996) 170. 7 Lopez-Ne ito J M. Kremnik G & Fiero J L G. Appl Cawl. 6 1

( 1990) 235. ~ Narayanan S & Des hapande K, Micropor Mm('/'. I I ( 1997)

77. 9 Narayanan S & Deshapande K. Apl ,1 CUlal A Cel/. 135

( 1996) 125. 10 Clark J H ,Cullcn S R, Barlow S J. Bastock T W. J ('/Ii'1II Soc

Perkill TrollS II. ( 1994) 111 7. II l30nd G C & Tah ir SF. i\ppl COWl, 7 1 (199 1) I. 12 Jacobs P A & Hey len C F. J Ca/(/ /. 34 {1974) 267. 13 Knozi ngcr H. Kri wtcnbrin k H & Ratnas31llY P, J Cawl, 48

( 1977) 436. 14 COrln a A, Rodell as C & Fornes V, J Cilo l. 88 ( 1984) 374. 15 Mi ya ta H & Moffat J B. J Cowl. 62 ( 1980) 357 . 16 Reddy 13 M, Vijayak ull1nr M & Jeeva Ratnam K. i\ppl Cawl

A Cel/, I ~ ( 1(99) 77. 17 Na rayanan S. Deshapande K & Prasad 13 P, J I/ wlec Cawl,

88( 1 9~4)L27 1.

18 Kij enski J & Baiker A, Cawl Torlav, I ( 1939). 19 Dines T J. Rochester C H & Ward A W, .I chelll Soc Farad

Trail S. 87 ( 199 1) 16 11. 20 Takalji M, Qwai T K, Soma M. Onoshi & Tamaru K. J phrs

Chelll, 80 ( 1976) 430. 2 1 Takalji M. Qwai T K. Soma M. Onm,hi & Tamaru K . .I

Calal. 50 (1977) 44 1. 22 Satsuma A. KamiYJ Y, Wcsti Y & Ha ttori T, Appl Callil A

Cell . 194 (2000) 253. 23 Khader M M, J 1II0iec COlal , A Ch 1'111 , 10.1 ( 1995) ~7 .

24 Cseri T. Bckassy S. FigucrJs F & Ri Fner S . .I lIIolec Cawl A Chell/, 98 ( 1995) 10 1.

25 Yiu H H P & Bl"Own D R. Cawl Lell, 56 ( 1998) 57. 26 Arata K. Hi no M &Yamagata N. /Jil/l chelll Suc .Iapal/ , 63

( 1990) 2 1. 27 Choudhary B M, Bhaskar V, Ka ntam M L. Raa K K &

Ragha van K V. C reell Chelllislry. 67 (2000).