11_chapter 3.pdf

34

CHAPTER 3 CATALYTIC DECOMPOSITION OF CYCLOHEXANOL

3.1 INTRODUCTION Catalytic dehydrogenation of cyclohexanol to cyclohexanone is an industrially important reaction. Cyclohexanone is the raw material for the manufacture of adipic acid (starting material for the manufacture of nylon-6,6) and for the manufacture of caprolactum (raw material for nylon-6). Cyclohexanone is also an important industrial solvent for PVC polymers, lacquers and also used as a building block in the synthesis of many organic compounds such as pharmaceuticals, insecticides etc. Cyclohexanone is used in the manufacture of magnetic video tapes (Kirk Othmer, 1979). Cyclohexanone is manufactured by dehydrogenation of cyclohexanol which is produced either by hydrogenation of phenol or by air oxidation of cyclohexane. A number of catalysts are reported in literature, most of which are based on copper, nickel and some metallic oxides either alone or as combination of two or mote with or without support. The dehydrogenation of cyclohexanol has been earned out over oxides of copper and zinc (Yu-Ming Lin et al 1988) and oxides of copper and magnesium, Ni-Sn-Sn02, Cu-Zn0-Al203, Cd-Cu-Al203 (Sivaraj et al 1988, Sideltseva and Erofeev 1986). However, these oxides are susceptible to sintering and limit the temperature of the reaction well below 300°C. Hence, the dehydrogenation of cyclohexanone has been a subject of intensive research by major caprolactum and adipic acid producing industries with a view to achieve higher conversion, selectivity, improved catalyst life time and lower operating temperature. Dehydrogenation of cyclohexanol is an endothermic reversible reaction. (Dake et al 1990). C6HnOH

C6H10O + H2

AH = 14.5 k.cal/mole

35

The decomposition of cyclohexanol leading exclusively either to cyclohexene or cyclohexanone has often been used as a model reaction in studying the mechanism of dehydration or dehydrogenation and in the evaluation of the catalysts (Morita et al 1970). It has also been reported that the quantity of cyclohexene formed depends on the density of Bronsted acidic sites (Bezouhanova and Al Zihari, 1992). Scheme 3.1 gives the different decomposition reactions of cyclohexanol (Richardson and Lu 1976) yielding a variety of products in single or multistages. 1.

Dehydrogenation of cyclohexanol to cyclohexanone.

2.

Dehydration of cyclohexanol to cyclohexene.

3.

Dehydrogenation (nuclear) to phenol.

4.

Self condensation of cyclohexanone to different oxidation products (BePskaya 1982).

5.

Carbon skeletal isomerisation of cyclohexene to methyl cyclo pentene isomers.

According to the above scheme, cyclohexanol may undergo dehydrogenation to phenol over platinum catalyst. The formation of cyclohexanone from phenol might occur through the formation of cyclohexene-l-ol as the intermediate. Hydrogenoiysis of phenol may give benzene (Manninger et al 1977). Cyclohexanol dehydrates on platinum-alumina catalyst to give cyclohexene over the acidic sites of alumina and cyclohexene disproportionates to benzene and cyclohexane. Cyclohexene also undergoes carbon skeletal isomerisation to methyl cyclopentene isomers over the acidic sites. The hydroxyl group of cyclohexanol interacts with platinum metal via oxygen atom leading to cleavage of O-H bond and formation of O-M bond simultaneously giving rise to adsorbed C = O-M species. This surface carbonyl species is held by the metal either by its x-bond or one of the lone pairs of electrons of oxygen. The alcoholate then transforms to a n bonded ketone and is desorbed as cyclohexanone. The cyclohexanone thus formed may be converted to benzene through cyclohexene.

36

0

OH

0

Benzene Pt

o

Cyclohexane

Cyclohexene

Scheme 3.1

Methyl Cyclopentene isomers

Different decomposition reactions of cyclohexanol

37

In addition to the above products, high boiling products like 2-{l-cyclohexenyl)-cyclohexanone (A), 2-cycIohexyliden cyclohexanone (B) and l-{l-cyclohexenyl)-cyclohexen (C) were also reported to have formed (Gut and Jaeger 1982) as shown in Scheme 3.2.

(c)

Scheme 3.2

High boiling products obtained from cyclohexanone

Metals catalysing the decomposition of cyclohexanol have been classified into two types (i) dehydrogenating metals (Os, Co, Fe, Ru, Re) and (ii) aromatizing metals (Pd, Pt, Ni) (Dobrovlszky et al 1981). Chen and Lee (1992) investigated the transformation of cyclohexanol over Cu-Zn0-AJ203 and Cu-Fe and reported that the dehydrogenation of cyclohexanol proceeded on metallic copper and dehydration on acidic sites. However, Fridman etal (1988 and 1992) reported that monovalent copper ions are the active sites for the dehydrogenation of cyclohexanol over copper-chromium catalysts. Sivaraj et al 1990 reported that Cu-Al203 catalysts prepared by coprecipitation using the urea hydrolysis procedure found to be highly active for the reaction of cyclohexanol to cyclohexanone and cyclohexene. They have also adopted ammonia chemisorption to assess Cu-A1203 catalysts for their relative selectivity to cyclohexanone and cyclohexene. Sideltseva and Erofeev (1986) have studied the dependence of the metal deposition procedure on the specific surface

3B

and catalytic properties of Cu/MgO catalysts. Wen-Shing Chen et al (1992) observed that dehydrogenation of cyclohexanol to cyclohexanone proceeded mainly on reduced metallic copper in Cu-Fe304 system. They found that Cu-Fe304 gave much higher yields to 79% at 663K than Cu-ZnO (ca.56%). Hsiu-Fuchang et al (1994) correlated the catalytic activity with acidity and surface area of copper of the Cu/A1203 catalyst in the dehydrogenation of cyclohexanol. Stoklosa et a/ (1996) studied the influence of zinc ions on the rate of cyclohexanol dehydrogenation by a copper catalyst. Raja and Santhanalakshmi (1996) obtained higher percentage conversion and yield of 2-methyl cyclohexanone in the dehydrogenation of 2-methyl cyclohexanol with calcined 3:1 Ni-Fe layered double hydroxide. Mg3(P04)2. yields cyclohexene preferentially and some cyclohexanone when modified by the addition of sodium carbonate in gas phase conversion of cyclohexanol (Aramendia et al 1996). Kuriacose et al (1968) illustrated that selectivity is determined not by the catalyst alone, but by both, the adsorbate and the catalyst. Athappan & Srivastava (1980) applied LangmuirHinshelwood kinetics to determine the mechanism of parallel dehydrogenation and dehydration of cyclohexanol on Ni0-Al203 systems and found that the model with ‘dual site’ surface reaction with cyclohexanol adsorbed’ found to fit for dehydrogenation and ‘dual site surface reaction with cyclohexene not adsorbed’ for dehydration. However, Gut and Jaeger (1982) discriminated the models by fitting the data to different kinetic models and suitable model was found to be a Langmuir-Hinshelwood model, involving associative adsorption of cyclohexanol, cyclohexanone and hydrogen on the same of kind of sites. 3.2 DECOMPOSITION OF CYCLOHEXANOL ON MIXED METAL OXIDES As mentioned in section 1.1, the noble metals and some metal oxides are susceptible to sintering and capable of forming volatile metal oxides at high temperatures.. Hence, stabilization of the metal ions in a suitable environment becomes necessary. Different ways may be adopted for stabilizing the metal ions. A2B04 oxides stable complex oxides with A and./ B sites partially substituted. It seems meaningful to understand first the significance of such partial substitution in the light of solid state chemistry. If the cations are replaced by divalent A' cation, the charge compensation will be achieved by the stoichiometric formation of either trivalent B cations (or positive holes) or oxide ion vacancies. However, these are two extremes and a mixture of these possibilities seems to appear in practice.

39

1. A2-x3+ A'x2+ B2+ B3+x O4 2. A2-x3+ A'x2+ B2+ Ofrx/2) Vcxx/2)

(Trivalent B3+ ion creation) (Oxide ion vacancies)

3. A2-x3+ Ax'2+ B2+!.2y B3+2y 0(4^2) + y Vo(x/2-y)

(Both)

Introduction of abnormal valence causes a change in electronic properties, while that of oxide ion vacancies increases the mobility of oxide ions and the ionic conductivity. Since oxide ion vacancies can act as accommodation sites for extra oxygen, the latter also leads to a variation of non-stoichiometric compositions depending on temperature and oxygen partial pressure. This chapter discusses the decomposition of cyclohexanol by mixed metal oxides of the types A2B04 and AB03. As both of them possess catalytically active B site ion, they show catalytic activity towards the decomposition of cyclohexanol. In addition to the above mentioned features, these oxides are extremely stable even at high temperature. Decomposition of cyclohexanol has been carried out on rare earth mixed metal oxides of the following types (a)

(i) Metal oxides of A2B04 type with K2NiF4 structure Ln2Cu04 (Ln = La - Gd,) and La2Ni04 (ii) Srx La2.x Cu04_y (x = 0, 0.2, 0.3, 0.5, 0.7 and 1.0)

(b)

Metal oxides of ABQ3 type with perovskite structure LaM03 (M = Co, Ni & Mn).

Both K2NiF4 and AB03 type oxides were prepared by ceramic technique. They were prepared by the reaction between the intimate mixtures of the constituent binary oxides or their precursors (nitrates, carbonates, etc.) as described in section 2.2.2. XRD patterns of the above mentioned compounds established the crystal structure and the single phase. The d-spacings were calculated from 20 values. The observed and calculated d values for the representative samples are

4D

Table 3.1 Calculated and observed d-spacings of A2BO4 & ABO3 type compounds

d(cal)(f)

d(obs)Cfo

Intensity

d(cal)
3.6281 3.2632 2.8578 2.6900 2.1826 1.6793 1.5869 1.4290

3.7000 2.8062 2.0466 1.9636 1.7416 1.5900 1.3948 1.2493 1.1557

3.7365 2.8335 2.0589 1.9603 1.7368 1.6120 1.6937 1.2409 1.1543 Sm2Cu04

s

3.6773

3.7090

m

vs

2.7791

2.7840

vs

m

2.0205 1.9314 1.6145 1.5861 1.2210

2.0269 1.9533 1.6163 1.5993 1.2212

s

2.8587 2.6584 2.2138 2.1694 2.0753 1.8790 1.7012 1.5718

2.8630 2.6639 2.2034 2.1639 2.0742 1.8840 1.6882 1.5729

m w vs s

w m

w w

Nd2Cu(>4 3.7250 2.7700 2.0361 1.9592 1.6224 1.6052 1.4156

3.7325 2.7787 2.0220 1.9670 1.6357 1.6143 1.4116 Gd2CuC>4

m s s w

---------------- 1---------------- 1----------------

3.6769 2.7332 2.0146 1.9327 1.6074 1.5850 1.2256

3.6281 2.7384 2.0180 1.9328 1.5974 1.5856 1.2301

Intensity

Pr2Cu(>4

La2Cu04 3.6670 3.2689 2.8731 2.7150 2.1790 1.6778 1.5855 1.4366

d(obs)^

m vs s s s s

m

m vs s vs m m m m w

s s s

m Lai.8Sr0.2CuO3.9 m ms m m m s

m vs

41

Table 3.1 (continued) d(caI)W

d(obs)(&)

Intensity

d
La2NiC>4

d(obs)(A)

Intensity

LaCo03

3.6707

3.6929

m

3.8200

3.7939

ms

2.8463

2.8494

vs

2.7110

2.7064

vs

2.7203

2.7291

vs

2.6790

2.6739

vs

2.1188

2.1188

m

2.2130

2.2077

s

2.0670

2.0684

s

1.9090

1.9058

vs

1.9236

1.9298

s

1.5630

1.5606

ms

1.7040

1.7104

m

1.5420

1.5420

ms

1.6704

1.6701

m

1.3560

1.3576

ms

1.5937

1.5979

s

1.2249

1.2205

m

LaNiQ3

LaMn03

2.7320

2.7246

vs

2.7390

2.7204

vs

1.9200

1.9277

s

1.9510

1.9357

s

1.7080

1.7086

m

1.5960

1.5843

s

1.5765

1.5964

m

1.3700

1.3762

ms

1.3640

1.3637

w

Fig.3.1

0)

s>Q)CP

■a

a>

cvi

SdO Pow der XRD p attern of L a2C u04

42

t

O

Q_

co

-

30 -

10

6CH

-06

120

150

Fig.3.2

25

Powder XRD pattern- of Sro.2Lai.8Cu03.9

2 0 (degrees)

40

55

70

43

150

Fig.3.3

Sd 0

( degrees)

Powder XRD pattern of Sm 2 CuC>4

2 0

44

120

150

SdD

Fig.3.4

©

{degrees )

55

Powder XRD pattern of Gd2Cu04

2

40

45

150

10

25

Fig.3.5

SdO 2 0 (

degrees )

55

Powder XRD pattern of Nd2Cu04

40

70

46

R g. 3.6

6 (

degrees )

Powder XRD pattern of La2Ni04

2

47

Fig.3.7

Pow der XRD p attern of LaCoOa 48

SdO

t

49

IR spectra for Ln2Cu04 (Ln = La to Gd) showed an absorption band at 690 cm*1 which can be assigned to the stretching vibration of the short La-Oj bond in La2Cu04 which confirmed the crystal symmetry to be orthorhombic (Singh and Ganguly 1984) whereas for the other system Ln2Cu04 (Ln = Pr to Gd), the absence of absorption bands in this region indicated that the compounds were tetragonally distorted. The appearance of two prominent bands in the regions (i) 505 - 515 cm*1 and (ii) 360-370 cm'1 have been assigned to the stretching vibrations of Cu-0 and the deformative bending mode of square planar Cu04 group in the D4h symmetry. The band occurring at 320 cm*1 has been assigned to the internal modes of vibration of Ln202 layers. In the La2_x Srx CuO^ (x = 0.2 to 1.0) system, absence of IR absorption band at 690 cm*1 indicated that except for the parent compound La2Cu04, all the compounds of SrxLa2.xCu04l/ (x = 0.2 to 1.0) were tetragonally distorted due to the substitution of strontium at the A site. The change in the orthorhombic to tetragonal symmetry may be attributed to the change in the position of O, atom from a linear La-0|-Cu linkage along the c axis in the orthorhombic phase to a tetragonal interstice of the lanthanum bilayers of the tetragonal phase (Ogitaeta/ 1987). The IR absorption bands around 510-520 cm*1 and 315-390 cm*1 have been assigned to the asymmetric stretching and bending modes respectively of strong Cu-On linkages in the basal planes. Three IR absorption bands around 650, 510 and 375 cm*1 observed for La2Ni04 have been attributed to the shortest La-O stretch of the La-O-Ni bridges, short Ni-0 bonds in the equatorial position and superposition of several vibrations contributed by the deformation of La-O-Ni bridges respectively (Odier

et al 1986). IR spectra of LaM03 (M = Co, Ni & Mn) perovskites show atleast two strong absorption bands in all the cases, one around 600 cm*1 and the another around 400 cm*1 which have been assigned to M06 vibration. IR spectra representative samples are given in Figures 3.8 and 3.9. Thus, both XRD patterns and IR spectral data have confirmed the phase purity of the compounds prepared in the present study.

Transm ittance (V.)

Transmittance (•/•>

51

Wavenumber ( cm-1)

Fig.3.9

IR spectra of (a) La2NiC>4; (b) LaCo03

52

3.3 SURFACE AREA OF THE CATALYSTS The surface areas of the catalysts were determined by BET method using nitrogen as adsorbate at liquid nitrogen temperature. The data are shown in Table 3.2. The surface areas for all the compounds are relatively low in the range of 0.97-3.41 m2/g. The observed low values for these catalysts may be attributed to the high temperature employed in the synthesis of these materials. Table 3.2 Surface area values of the mixed metal oxide catalysts S.No.

Catalyst

1.

Surface area <m2/g) 1.36

2.

Pr2Cu04

1.40

3.

Sm2Cu04

1.21

4.

Nd2Cu04

0.97

5.

Gd2Cu04

1.24

6.

Sr0 2Lai.8Cu03 9

1.42

7.

Sro.3La1.7CuO3.g5

1.40

8.

^r0.5^-a1.5^'u®3.75

1.34

9.

Sr0 7Laj sCuO^g

1.50

10.

SrLaCu035

1.40

11.

La2Ni04

2.07

12.

LaCo03

3.41

13.

LaMn03

2.86

14.

LaNiOg

3.05

53

3.4 CATALYTIC ACTIVITY AND KINETIC STUDIES ON K2NiF4 AND ABOs TYPE OXIDES 3.4.1

Flow reactor technique

The catalytic activities of K2NiF4 and AB03 type oxides towards the dehydrogenation of cyclohexanol have been investigated in vapour phase by the flow reactor technique. The catalyst samples were first pretreated with hydrogen at 673 K for 4 hours. The reaction was investigated for the conversion of cyclohexanol in the temperature range of 473 - 823 K with constant flow rate (W/F = 53.0 g.hr./g.mole). The flow rate of cyclohexanol was controlled by a syringe model infusion pump. A constant weight (0.5 g) of the catalyst of mesh size 60-80 was packed in the reactor zone of the pyrex reactor which was then plugged tightly with the help of asbestos wool. The products were collected by condensing them and subjected to gas chromatographic analysis. The quantification of the product analysis was achieved from the peak areas calculated by the integrator attached to gas chromatograph. Among the catalysts synthesised and tested, only La2Cu04 and Sr^La^CuO^ were found to be selective towards the cyclohexanone formation in the temperature range of 553 - 648 K and all others in the series Ln2Cu04 (Ln = Pr - Gd), La2Ni04 and LaM03 (M = Co, Ni & Mn) gave a mixture of products. Hence, kinetic studies were carried out only for La2Cu04 and SrxLa2_xCu04^ (x = 0.2, 0.3, 0.5, 0.7 and 1.0). 3.4.2

Activity studies on l^CuO^

The effect of temperature on the conversion of cyclohexanol on La2Cu04 (0.5 g pretreated) is shown in Figure 3.10. It was found that the selectivity towards the formation of cyclohexanone on La2Cu04 was maximum in the temperature range from 553 - 613 K. Beyond 653 K, the formation of cyclohexene was found to increase. Therefore kinetic studies were limited to the temperature range of 553 - 613 K. The kinetics of catalytic vapour phase dehydrogenation of cyclohexanol to cyclohexanone was studied quantitatively by varying the (i) contact time (by varying the reactant flow rate) and (ii) temperature. Fresh, pretreated catalyst (0.5 g) was used for every run.

< T >

O

?

Conversion (%)

O

CD

O O

O

C V J

550

630

710

790

Temperature ( K )

Fig.3.10

The plot of the effect of temperature on the decomposition of cyclohexanol over La2Cu04; a : Cyclohexene; b : cyclohexanone; c : unreacted cyclohexanol; d : percentage selectivity towards cyclohexanone.

55

The percentage selectivity of cyclohexanone was calculated by using the formula Percentage selectivity of cyclohexanone = Percentage of Cyclohexanone Percentage of cyclohexanol converted 3.4.2.1

^qq

Effect of contact time

The effect of contact time on the catalytic conversion of cyclohexanol was investigated by varying the factor W/F (where W is the weight of the catalyst in grams and F is the flow rate of the reactant cyclohexanol in g. mole/hr). The conversion of cyclohexanol to cyclohexanone was found to increase with increasing W/F for all the compositions in La2Cu04 system. The particle size of the catalyst and the flow rates of cyclohexanol were chosen from preliminary trial experiments so as to eliminate the mass transfer effects. Considering the fact that the surface area of these mixed oxide catalysts were only about 1 - 3.4 m2/g (60 - 80 mesh), the diffusion effect may be neglected. These low values of surface area also indicate that the catalysts prepared are nonporous. 3.4.2.2

Effect of temperature

The effect of temperature on the conversion of cyclohexanol to cyclohexanone at constant catalyst weight was carried out in the temperature range 573 - 613 K for each of the catalyst compositions. The conversion in the above temperature range was found to increase with increasing temperature as expected. For heterogeneous catalytic reactions, the overall apparent effect of temperature is both due to the effect of temperature on adsorption and also the effect on the reaction velocity. The adsorption of the reaction generally decreases with increasing temperature. It is not possible to deduce independently the relative magnitude of these two factors on the basis of simple experimental studies. However, the advent of computational methods has paved the way for determining the relative importance of these factors. 3.5 ANALYSIS OF EXPERIMENTAL DATA The experimental data of conversion as a function of temperature at constant W/F value and as a function of W/F were analysed at fixed temperature in order to calculate the kinetic parameters of the reaction.

56

The experimental data have been analysed by two different approaches: i) Integral method and ii) Method based on reaction mechanism.

3.5.1

Analysis using integral method

The design equation for the differential flow reactor (catalytic) may be written in the form

W F

dx' r

(3.1)

where W is the weight of the catalyst in gram; F is the flow rate of the cyclohexanol in g.mole/hr. and r is a function of concentration of the reactant Ca which may be expressed as

r

dx' d (W/F)

= kCAn

(3.2)

where n is the order of the reaction The value of ‘n’ can be evaluated by plotting r vs CA or by assuming a value for n, solving the equation (3.2) and then testing whether the value of K remains constant. In this study, an integral method was used assuming a first order reaction for the dehydrogenation of cyclohexanol. The concentration of cyclohexanol can be calculated from the knowledge of the percentage conversion by the following method. Let the molar flow rate of cyclohexanol be F g mole/hr. Let the fractional conversion be x'. The molar flow rates of the components were evaluated by using the following equation.

( 3.3 )

57

The molar flow rates of different components at the exit end of the reactor will then be Na

(l-x')F

(3.4)

Nb

-

x' F

(3.5)

Nc

-

x' F

(3.6)

The total flow rate at any time will be (3.7)

(1 + x') F

NTotal

Assuming that the ideal gas law is applicable at the reaction conditions, the concentration at any time is Na

c

- —

y

Na

Pt

NTotel

RT

where, V = Volume of cyclohexanol PT = Pressure of cyclohexanol at the temperature of the reaction R = Gas constant T = Temperature of the reaction, F (1 - xO PT (3.8)

F (1 + x') R T The rate of the. first order reaction can be expressed as dx * . n r ~ d (W/F) " k Ca

(3.9)

substituting for CA in equation (3.2) from (3.9) d x' d (W/F)

_ , "

(1 - x') (1 + x')

*t RT

Integrating the equation within the limits 0 to x'

(3.10)

58

x'

» d x'

q-*).

J d (W/F)

(l+x)

(3.11)

0

gives the equation

(3.12)

-x' - 2 In (1 -x')

The value of the rate constant was determined from a plot of -x' - 2 In (1-x') vs W/F for the first order reaction.

3.5.1.1

Evaluation of kinetic and thermodynamic parameters The rate constant values were obtained from the slope values of the

linear plot of -x'-21n (1-xO vs W/F. The linear dependance of In k us 1/T gave the activation energy (Ea). By knowing rate constant and activation energy values, Arrhenius frequency factor values were calculated from the equation

log k = log A -

(3.13)

2.303 R T

The values of heat of activation (AH) were also found out from the general equation AH = Ea - RT

(3.14)

The validity of the integral equation was checked by substituting the values of the frequency factor (A) and activation energy (Ea) in the equation. -E

___a

p

-x' - 2 In (1-xO = AeRT (W/F) ^

(3.15)

and back calculating W/F values as a function of x' at a given temperature. The calculated W/F and experimental W/F values fall on the same curve for a given temperature which confirmed the validity of the integral equation.

59

3.5.2

Analysis based on reaction mechanisms

The integral rate equation has limitations in that it assumes a very simple expression for the concentration dependence of the reaction rate. However, catalytic reactions are quite complex and take place in a number of steps, both physical and chemical in nature. In addition, the temperature dependence is a combined effect of temperature on the adsorption equilibrium constant and the reaction velocity constant. Hence, to obtain the rate equations which could be applied over a wider range of conditions, all the steps through which the reaction takes place on the surface of the catalyst must be given due consideration. The general scheme for a surface catalysed reaction involves five stages of physical and chemical processes for the conversion of reactant A into a product B (Langmuir - Hinshelwood mechanism). 1. 2. 3. 4. 5.

Diffusion to surface Adsorption of reactant (A) on to surface Chemical conversion on surface Desorption of product (B) from surface Diffusion from surface

In general, the rate of any one of these stages can create a bottleneck, such that the kinetic dependences of that stage are imposed on the overall reaction. The chemical steps involved are considered to be the three consecutive steps based on Langmuir - Hinshelwood mechanism, namely adsorption, surface reaction and desorption. The experimental conditions were so chosen as to eliminate the physical steps. Internal diffusion was negligible because of the small surface area of the catalyst and absence of pores. The dimension of the reactor and the catalyst bed height precluded the influence of external diffusion and transport phenomena on the rate of reaction. Assumptions have also been made that all active centres (or catalyst sites) function identically and that each molecule is adsorbed regardless of interaction between molecules of the same and different kinds.Under these conditions, it was fair to assume that the rate of the reaction was influenced only by any one of the chemical steps.

60

The five possible rate

controlling

chemical steps and the

corresponding rate expressions are given below : B+C

A ^ 1.

Adsorption rate controlling

r

_ "

______ k(CA-CB Cc /Keq) KaCbCc 1+ K + KBCB + KcCc

(3.16)

*\q

2.

Surface reaction - single site rate controlling Adsorbed A reacts to give B (adsorbed) and C (not adsorbed) ^C /Keq)

r 3. 3a.

"

(3.17)

1 + KaCa + KbCb

Surface reaction - dual site rate controlling. Adsorbed A forms a transition state complex with a neighbouring free, active centre which reacts to give adsorbed B and adsorbed C (without dissociation of C) »
kbcb

+ Kcty2

(3.18)

3b. Adsorbed A forms a transition state complex with a neighbouring free, active centre which reacts to give adsorbed B and dissociatively adsorbed C k ^a (^a ~ CB Cc/Keq) r

3c.

(3.19) (1 + KaCa+KbCb + Kc

On an active centre, z, adsorbed A forms with a neighbouring free,, active centre, s, of other species a transition state complex which reacts further to give B (on z adsorbed) and C (on s adsorbed without dissociation) k KA (CA r

CB Cc/KCg)

(1 + KA CA + KB CB) (l+KcCc)

(3.20)

61

3d.

On an active centre, z, adsorbed A forms with a neighbouring free, active centre, s, of other species a transition state complex which reacts further to give B (on z adsorbed) and C (on s adsorbed with dissociation) k

r

4.

( CA ~ CB CC/^eq)

(1 + KaCa + KbCb) (1 + KcC(i)2

Surface reaction - dual site with negligible adsorption of products rate controlling. ^a(CA-CB CC /Keg) (i + kaca)2

5.

(3.21)

(3.22)

Desorption rate controlling

r

=

^eg ^ ( CA / CB - CB / K^)

l + KACA + KeqKc^

(3.23)

The true rate expression based on the reaction mechanism that governs the heterogeneously catalysed reaction was determined by statistical methods of parameter estimation and model selection. Two methods of modelling possible are : (i) from the integral data without extracting point rates, which involves estimation of the parameters in a system of first order differential equations that may be linear or non linear and (ii) from rate data obtained directly from a differential reactor or by differentiation of integral data (concentration vs W/F), which involves estimating the parameters in a system of linear or non­ linear algebraic equations. In the present study the second method was followed. A broad outline of the method involved the following steps. (a)

estimation of'initial rates

(b)

estimation of thermodynamic equilibrium constant for the overall reaction.

(c)

estimation of gas phase concentrations of reactant and products.

62

(d)

Numerical analysis of the algebraic equation of the rate expression for determining rate constant and the adsorption equilibrium constant.

(e)

Discrimination of the models.

Detailed methodology of these steps is given in the following subsections.

3.5.2.1 Estimation of initial rates The rate of the reaction was obtained from the experimental conversion vs W/F curve by regression analysis of the second degree polynomial of the form x' = a(W/F) + b(W/F)2

(3.24)

r = a + 2b (W / F)

(3.25)

and therefore

the zero power term of (W/F) was excluded from the polynomial to satisfy the boundary conditions, that at (W/F) = 0, x' = 0. A model calculation of the initial rates using the above polynomial by least squares method and the computer program adopted to solve the co-efficients of polynomials are given in Appendices

1 & 2. 3.5.2.2 Estimation of thermodynamic equilibrium constant The equilibrium constant for the uncatalysed reaction was calculated from specific heat data. The methodology involved the following steps. The specific heat data for the cyclohexanol dehydrogenation reaction was calculated by using the group contribution method (Lyman et al 1982). The data so obtained was used for calculating the free energy change of the reaction using the equation. AG° = AH° - Aa T (InT) -1 Ab T2 -1 Ac T3 + I.T 4

(3.26)

6

The AH° value in the above equation was obtained from the known values of AHT, Aa, Ab and Ac at the standard temperature of 298°K. The

63

thermodynamic equilibrium constant, K was calculated from AG° value by using eq the equation 3.28. AG° = - RT In K

(3.27)

eq

A model calculation is given in Appendix 3.

3.5.2.3 Estimation of gas phase concentrations The gas phase concentrations of cyclohexanol CA, cyclohexanone CB and hydrogen Cc were calculated from conversion data. (1-x )

PT

(1+X)

RT

t

X

(1

+ X)

PT RT

X

PT

(1+X)

RT

/

(3.28)

(3.29)

(3.30)

3.5.2.4 Numerical analysis of the rate expression The conventional method of analysing Hougen-Watson models (Yang and Hougen 1951) consists of transforming the non linear rate expression into a linear equation and solving it by the standard regression method. However, not all the Hougen-Watson models are linearised especially those involving dual site surface reaction mechanism. Further, linearisation results in a rate equation in which the concentration term becomes simultaneously dependent as well as an independent variable. Estimation based on such model violates the assumptions for which linear regression is valid. In linear methods, the results of conventional unweighted analysis can be quite misleading. Models can be rejected on the basis of negative adsorption constants even though these apparent inadequacies can be caused by the type of data analysis rather than by the true behaviour of the experimental data. Lastly, the linear least square analysis allows to fit the data to Hougen-Watson mechanism at isothermal conditions only. Non linear estimation methods allow the analysis of the data even at non isothermal conditions. A non linear estimation of the rate data can result in the estimation of the parameters which are compatible with the problem, because

the sum of squares of residual rates are minimised directly instead of minimising the residuals of a combination of variables involved in the rate. A comparison of the results obtained on the basis of the linear least squares analysis of isothermal data and those of non linear least squares analysis of both isothermal and non isothermal data has shown that the non linear least squares procedure was better for a rational selection of an acceptable model and estimation of its parameters (Kittrell et al 1965). The non linear estimation method has been used for the kinetic analysis of dehydrogenation of ethanol and isomerisation of n-butene reactions as reported in literature (Peterson & Leon Lapidus 1966). In the present study, the rate data have been analysed by non linear least square method of analysis through a computer program. The program consisted of three sections. 1)

Input section - for feeding the reaction model equation, and experimental data.

2)

Refinement section - consisting of an algorithm for solving the non linear objective function by iterative method from given initial estimates to generate a sequence of new estimates, each of which represents an improvement over the previous one.

3)

Regression section - algorithm for obtaining least square estimates of the parameter.

The pattern search technique proposed by Hooke and Jeeves (Swann 1972) was used in the program. This method attempts to define and pursue a direction which lies along the local principal axis of the contours of the function using a combination of exploratory moves and patterns moves to generate a sequence of improving approximations. Exploratory moves examine the local behaviour of the function and seek to locate the direction of any sloping valley present. Pattern moves utilise the information yielded by the exploratory moves for further progressing along such valleys. The procedures continue until convergence is achieved. A list of the program is included in Appendix 4.

65

3.5.2.5

Model discrimination The selection of the reaction mechanism was based on the following

criteria. 1.

That all the coefficients must be positive and significantly different from zero.

2.

That the reaction velocity constant and adsorption equilibrium constant must be temperature dependent.

3.

If more than one mechanism satisfied the conditions given above, then whichever mechanism gives the best fit between experimental and calculated reaction rates was selected as the correct mechanism.

3.6 KINETIC STUDIES ON La2Cu04 To find the effect of contact time( W/)F on the conversion of cyclohexanol, four different values of W/F were taken and the variation of conversion of cyclohexanol with W/F is shown in Figure 3.11. The flow reactor data for the dehydrogenation of cyclohexanol over La2Qi04, calculated by integral method is given in Table 3.3. The plots of -x'-21n (1-x') versus W/F at different temperatures are shown in Figure 3.12. Straight lines passing through origin were obtained indicating the reaction to be first order with respect to the concentration of cyclohexanol. The values of rate constants k, at different temperatures were obtained from the slopes of the straight lines in the Figure 3.12. The activation energy Ea is obtained from the slope of the plot of log k vs 1/T (Figure 3.13) and the activation energy was found to be 39.8 K.J/mole. The rate constant and activation energy values were substituted in the equation 3.12 and the Arrhenius factors for all the three temperatures were calculated. From the equation 3.13 AH values were also calculated and tabulated in Table 3.3. The Ea and frequency factor (A) were substituted in the equation 3.14 and W/F values were back calculated. The values of (W/F)^ and the experimental (W/F)exp values fall in the same curve. The effect of time on stream on the conversion of cyclohexanol over

La2Cu04 was studied in the temperature range of 573 - 613 K for 3 hrs. Figure 3.14 shows the effect of time on stream at W/F = 53 g. hour/g. mole.

W/p ( g r hr / g. mole )

Fig.3.11 The plot of effect of contact time on the conversion of cycloliexanol over La2Cu04 (1st hour) at different temperatures. a : 613K; b: 593K; c : 573K.

67

Table 3.3 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over La2Cu04 W/F (g-hr/ g. mole

k -x'-2ln(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH (kJ/mole)

Temperature 613 K

21.2

0.227

0.2879

26.5

0.280

0.3770

37.9

0.373

0.5606

53.0

0.436

0.7094

0.6539 2.289xl03

36.5

Temperature 593 K

21.2

0.200

0.2463

26.5

0.238

0;3056

37.9

0.294

0.4023

53.0

0.405

0.6334

0.5157 2.378xl03

36.7

Temperature 573 K

17.7

0.112

0.1256

26.5

0.175

0.2055

37.9

0.242

0.3121

53.0

0.310

0.4321

0.3709 2.296xl03

Activation energy Ea = 39.8 kJ/mole. Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream.

36.8

(

2 In 1-X*)

68

Fig.3.12 Kinetic plot for the dehydrogenation of cyclohexanol over La2Cu04 at different temperatures a : 613K; b : 593K; c : 573K

log k

( lit. atm./g.hr)

69

Fig.3.13 Arrhenius plot for the cyclohexanol over La2Cu04

dehydrogenation

of

Fig.3.14 Plot of the effect of time on stream on the conversion of cyclohexanol over La2CuC rel="nofollow">4 at W/F = 53 g.hr. / g. mole. a : 573K; b : 593K; c : 613K.

71

All the possible mechanisms based on the Langmuir-Hinshelwood and Hougen-Watson models namely, 1.

adsorption

2.

surface reaction - single site B adsorbed and C not adsorbed

3.

Surface reaction - dual site A adsorbed on same types of active centres 3a. Both B and C adsorbed without dissociation 3b. Both B and C adsorbed but C dissociatively adsorbed A adsorbed on different types of active centres (z and s) 3c. B adsorbed on z and C adsorbed on s without dissociation 3d. B adsorbed on z and C dissociatively adsorbed on s

4.

surface reaction with negligible adsorption of products

5.

desorption

were considered as rate controlling steps for the catalytic conversion of cyclohexanol to cyclohexanone catalysed by La2Cu04. The rate constant k and adsorption equilibrium constant of the reactant KA as well as of the products KB and Kq were calculated by using the computer program. The flow reactor data used for the model descrimination and the parameter values obtained for different models for the dehydrogenation of cyclohexanol to cyclohexanone catalysed by La2Gi04 are given in Tables 3.4 and 3.5

3.7 KINETIC STUDIES ON La1<8Sr0'2 Cu03 9 The kinetic studies of the dehydrogenation of cyclohexanol was investigated in the temperature range of 573 - 613 K. The experimental data is given in Table 3.6. Figures 3.15 to 3.17 show various plots for the evaluation of kinetic parameters. The initial rates derived from W/F us x' plot and the gas phase concentration derived from the conversion data (Table 3.7) were used for deducing the rate determining step. The computed values for all the possible mechanisms are listed in Table 3.8.

72

Table 3.4 Flow reactor data used for the model descrimination for the dehydrogenation of cyclohexanol on La20uO4 W/F

S*hr/ g.mole

t

X

CA

cB

cc

g. mole

g. mole

g. mole

r

Temperature - 613 K

21.2

0.227

0.01252

0.00368

0.00368

0.00922

26.5

0.280

0.01118

0.00435

0.00435

0.00829

37.9

0.373

0.00908

0.00540

0.00540

0.00631

53.0 .

0.436

0.00781

0.00604

0.00604

0.00369

Coefficients for the polynomial: a = 1.29 x 10'2 b = -8.69 x 10'5 Temperature = 593 K

21.2

0.200

0.01370

0.00343

0.00343

0.00802

26.5

0.238

0.01265

0.00395

0.00395

0.00753

37.9

0.294

0.01121

0.00467

0.00467

0.00646

53.0

0.405

0.00870

0.00592

0.00592

0.00506

Coefficients for the polynomial: a = 9.995 x 10'3

b =-4.6582 xlO'5

Temperature = 573 K

17.7

0.112

0.01699

0.00214

0.00214

0.00633

26.5

0.175

0.01493

0.00317

0.00317

0.00587

37.9

0.242

0.01298

0.00414

0.00414

0.00529

53.0

0.310

0.01120

0.00503

0.00503

0.00451

Coefficients for the polynomial: a = 7.237 x 10'3

b = -2.572 x 10'5

73

Table - 3.5 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on La2Cu04

Model No.

1

2

3a

3b

3c

Tempe­ rature K

CONSTANTS k

KA

Kb

Kc

613

0.650

-1.7138

-1.8952

-1.9525

593

0.600

3.0327

3.4657

2.9250

573

0.425

5.6294

5.6310

5.6000

613

1.200

0.550

-2.1475

-

593

1.200

0.500

4.4560

-

573

1.025

0.400

-4.8560

-

613

1.200

0.550

-0.8314

-0.9000

593

1.150

0.500

-1.3377

-1.3000

573

1.025

0.400

-1.6727

-1.6500

613

1.100

0.550

-0.6918

-0.4906

593

1.100

0.500

-0.3708

-0.4136

573

0.979

0.400

-0.0760

-0.2107

613

1.200

0.5020

-0.7438

-0.2002

593

1.122

0.5000

-3.8831

-3.8237

573

1.127

0.4000

-1.2046

-1.1516

74

Table - 3.5 (Continued)

Model No.

Tempe­ rature K

CONSTANTS k

ka

kb

Kc

613

1.116

0.5983

0.5032

-.0.3500

593

1.100

0.5000

0.0064

-0.5398

573

1.014

0.3996

-1.0578

613

1.156

0.5750

-

-

593

1.183

0.5000

-

-

573

0.981

0.4250

-

-

613

0.022

1.4540

-

3.2580

593

0.022

1.5280

-

3.0634

573

0.019

1.5420

-

3.3204

3d

4

5

1.0000

-

Model No.l

Adsorption rate controlling

Model No.2

Surface reaction - single site rate controlling

Model No.3

Surface reaction - dual site A adsorbed on same types of active centres 3a. Both B and C adsorbed without dissociation rate controlling 3b. Both B & C adsorbed but C dissociatively adsorbedrate controlling A adsorbed on different types of active centres 3c. B and C adsorbed on different types of active centre (C - without dissociation) - rate controlling 3d. B and C adsorbed on different types of active centre (C - with dissociation) - rate controlling

Model No. 4

Surface reaction with negligible adsorption of products rate controlling

Model No.5

Desorption - rate controlling

75

Tabic 3.6 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over Lai.sSr0.2CuO3.9 W/F (S-br/ g. mole

k -x'-21n(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH (kJ/mole)

Temperature 613 K. 15.1

0.206

0.2553

17.7

0.226

0.2864

26.5

0.275

0.3682

35.3

0.351

0.5136

0.8047 6.020xl03

40.37

Temperature 593 K 15.1

0.171

0.2041

21.2

0.203

0.2508

26.5

0.259

0.3408

35.3

0.295

0.4041

0.5839 5.901xl03

40.54

Temperature 573 K 17.7

0.155

0.1818

18.9

0.165

0.1956

26.5

0.220

0.2769

35.3

0.268

0.3559

0.4278 5.966x103

Activation energy Ea = 45.47 kj/mole. Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream.

40.71

Fig.3.15 The plot of effect of contact time on the conversion of cyclohexanol over Lai.8Sr0.2CuO3.9 (1st hour) at different temperatures, a : 613K; b: 593K; c : 573K.

77

Fig.3.16 Kinetic plot for the dehydrogenation of cyclohexanol over Lai.8Sro.2Cu03.9 at different temperatures a : 613K; b : 593K; c : 573K

tog

k (

tit. atm /g.hr

)

7B

Fig. 3.17 Arrhenius plot for the dehydrogenation cyclohexanol over Lai.8Sr0.2CuO3.9*

of

79

Table 3.7 Flow reactor data used for the model descrimination for the dehydrogenation of cyclohexanol on Sro^Lai.sCuOa.g W/F g.hr/ g.mole

9

X

CA g. mole

CB

g. mole

Cc g. mole

r

Temperature = 613 K

15.1

0.206

0.01309

0.00340

0.00340

0.01053

17.7

0.226

0.01255

0.00366

0.00366

0.00970

26.5

0.275

0.01130

0.00429

0.00429

0.00690

35.3

0.351

0.00955

0.00516

0.00516

0.00410

Coefficients for the polynomial a = 1.5339 x 10'2

b =-1.5925 x 1O'4

Temperature = 593 K

15.1

0.171

0.01455

0.00300

0.00300

0.00902

21.2

0.203

0.01356

0.00347

0.00347

0.00750

26.5

0.259

0.01209

0.00409

0.00409

0.00617

35.3

0.295

0.01119

0.00468

0.00468

0.00396

Coefficients for the polynomial a = 1.2806 x 10'2

b =-1.2524 xlO4

Temperature = 573 K

17.7

0.155

0.01556

0.00285 . 0.00285

0.00761

18.9

0.165

0.01525

0.00301

0.00301

0.00745

26.5

0.220

0.01360

0.00384

0.00384

0.00641

35.3

0.268

0.01227

0.00450

0.00450

0.00520

Coefficients for the polynomial a = 1.00346 x 10'2 b = -6.8484 x 10~5

an

Table - 3.8 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on Sro^LaigCuC^g

Model No.

1

2

3a

3b

Tempe­ rature . K

CONSTANTS k

ka

KB

Kc

613

0.650

-1.1411

-1.1594

-1.1250

593

0.525

4.5575

4.5750

4.5500

573

0.475

2.5512

2.4249

2.4000

613

1.175

0.5500

-3.0349

—

593

1.025

0.5000

-5.2513

-

573

1.075

0.4500

5.3615

-

' 613

1.200

0.5500

0.1871

0.1750

593

1.000

0.5000

-1.7530

-1.8000

573

1.075

0.4500

0.8354

0.8500

613

1.163

0.5409

-0.2039

-0.4898

593

1.063

0.4369

-0.5265

-0.5478

573

1.002

0.4000

-0.8632

-0.3068

B1

Table - 3.8 (Continued)

Model No.

3c

3d

4

5

Tempe­ rature K

CONSTANTS k

ka

KC

KB

613

1.146

0.5957

-3.9275

-3.9744

593

0.999

0.4984

-2.6724

-2.6519

573

1.137

0.4279

2.1831

2.1048

613

1.187

0.5802

-1.0988

-0.5884

593

1.063

0.4776

-0.4079

-0.4433

573

1.177

0.4000

-0.9507

-0.3500

613

1.195

0.5500

-

-

593

1.025

0.5000

-

-

573

1.002

0.4750

-

-

613

0.022

1.3590

-

2.8999

593

0.022

1.5200

-

3.3179

573

0.022

0.2990

-

3.2252

B2

3.8 KINETIC STUDIES ON LaL7 Sr0 3 Cu03>85 x' vs W/F, kinetic plot and Arrhenius plots for the system Lai 7Sr03CuO3 85 are Siven in Figures 3.18 to 3.20 and Table 3.9 gives the experimental data for the cyclohexanol decomposition over Lax 7 Sr0 3 Cu03 85. The values of CA, CB, Cc and r used to descriminate the models are given in Table 3.10 and the parameter values obtained from different models for the catalyst Lai 7 $ro 3 Cu03 85 are given in Table 3.11.

3.9 KINETIC STUDIES ON Laj 5 Sr0 5 Cu03 75 The corresponding plots for the system Lai 5 ^r0 5 Cu03 75 are given in Figures 3.21 to 3.23 and Table 3.12 shows the experimental data for the cyclohexanol decomposition over Lai 5 ^ro 5 Cu03 75- The values of initial reaction rates obtained from the polynomials and the gas phase concentrations derived from the conversion data are given in Table 3.13. These values were used as input for solving the reaction rate expression based on mechanisms with the help of the computer program. The adsorption equilibrium constants for reactants and for the products are given in Table 3.14.

3.10

KINETIC STUDIES ON Lax 3 Sr0 7 Cu03 65

Figures 3.24 to 3.26 shown for the system Lai 3 SrQ 7 Cu03 65 were used for evaluating kinetic parameters and Table 3.15 shows the experimental data for the cyclohexanol decomposition over Laj 3 Sr0 7 Cu03 65. The values of the initial rates of reaction derived from the polynomials are given with the corresponding concentrations of reactant and products in Table 3.16. The values of the rate constant (k), adsorption equilibrium constant (KA, KB and Kc) calculated from various mechanistic steps are given Table 3.17.

3.11

KINETIC STUDIES ON LaSrCu03 5

Kinetic studies were carried out in the temperature range of 593 to 643 K. the experimental data pertaining to LaSrCu03 5 is shown in Table 3.18. The corresponding x'vs W/F, x' vs -x' - 21n (1-x'), log k vs 1/T are shown in Figures 3.27 to 3.29. The values of CA, CB, Cc and r are given in Table 3.19 and Table 3.20 shows the values of k, KA, KB and Kc for all the possible reaction mechanisms.

83

Fig.3.18 The plot of effect of contact time on the conversion of cyclohexanol over Lai.7Sro.3CuC>3.8 (1st hour) at different temperatures, a : 603K; b: 593K; c : 583K1

84

Fig.3.19 Kinetic plot for the dehydrogenation of cyclohexanol over Lai.7Sro.3Cu03.8 at different temperatures a : 603K; b : 593K; c : 583K

log k

( lit. atm /g.hr

5

05

Fig.3.20 Arrhenius plot for the dehydrogenation cyclohexanol over Lai.7Sro.3CuO3.8S

of

86

Table 3.9 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over La1.7Sro.3CuO3.s5 W/F (s-W g. mole

k -x'-21n(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH

(kJ/mole)

Temperature 603 K 17.7

0.215

0.2691

26.5

0.285

0.3859

35.3

0.368

0.5497

53.0

0.470

0.7997

0.6926 0.839xl03

30.58

Temperature 593 K 21.2

0.213

0.2661

26.5

0.250

0.3253

35.3

0.305

0.4226

53.0

0.415

0.6572

0.6082 0.830xl03

30.67

Temperature 583 K 17.7

0.171

0.2041

21.2

0.185

0.2241

35.3

0.280

0.3770

53.0

0.370

0.5540

0.5381 0.832xl03

Activation energy Ea = 35.6 kj/mole.

Reaction conditions : Catalyst Weight - 0.5 g. Data taken after 1 hr. of time on stream.

30.75

87

Tabic 3.10 Flow reactor data used for the model descrimination for the dehydrogenation of cyclohexanol on Sro.3La1.7CuO3.85

W/F

g.hr/ g.mole

9

X

eA

g. mole

CB

g. mole

g. mole

r

Temperature = 603 K

17.7

0.215

0.01305

0.00358

0.00358

0.01033

26.5

0.285

0.01125

0.00448

0.00448

0.00887

35.3

0.368

0.00934

0.00544

0.00544

0.00741

53.0

0.470

0.00729

0.00646

0.00646

0.00447

Coefficients for the polynomial a = 1.3265 x 10'2

b = -8.295 x 105

Temperature 593 K

21.2

0.213

0.01333

0.00361

0.00361

0.00844

26.5

0.250

0.01233

0.00411

0.00411

0.00779

35.3

0.305

0.01094

0.00480

0.00480

0.00671

53.0

0.415

0.00850

0.00602

0.00602

0.00455

Coefficients for the polynomial a = 1.1026 x 10'2

b =-6.1132 x lO’5

Temperature 583 K

17.7

0.171

0.01480

0.00305

0.00305

0.00805

21.2

0.185

0.01438

0.00326

0.00326

0.00769

35.3

0.280

0.01176

0.00457

0.00457

0.00586

53.0

0.370

0.00961

0.00565

0.00565

0.00366

Coefficients for the polynomial a = 1.0246 x 10'2

b = -6.216 x 10'5

aa

Table-3.11 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on Sro.3La17CuO3.s5

Model No.

1

2

3a

3b

Tempe­ rature K

CONSTANTS k

Ka

Kb

Kc

603

0.750

0.5282

0.5161

0.5500

593

0.550

-5.1926

-5.2349

-5.2500

583

0.475

-4.1762

-4.2101

-4.1750

603

1.225

0.6000

-4.6220

-

593

1.075

0.5750

4.2181

-

583

1.050

0.4750

4.7491

-

603

1.225

0.6000

-1.4929

-1.5000

593

1.075

0.5750

0.6603

-6.5000

583

1.050

0.4750

0.7199

-0.8000

603

1.170

0.5691

-0.4015

-0.5293

593

1.091

0.5000

-0.5361

-0.7144

583

1.016

0.4375

-0.6961

-0.4862

89

Table - 3.11 (Continued)

Model No.

3c

3d

4

5

Tempe­ rature K

CONSTANTS k

Ka

Kg

KC

603

1.101

0.6000

-1.5904

-2.5128

593

1.093

0.5972

5.8800

2.3916

583

1.025

0.4467

2.9757

3.0821

603

1.195

0.5954

0.0660

-0.7064

593

1.089

0.5000

0.0598

-0.9091

583

0.969

0.4694

-0.4675

-0.4250

603

1.260

0.6000

-

-

593

1.064

0.5750

-

-

583

1.036

0.4750

-

-

603

0.022

1.3080

-

2.8381

593

0.022

2.1020

-

3.1475

583

0.022

4.8010

-

3.5049

V

90

Fig.3.21 The plot of effect of contact time on the conversion of cyciohexanot over Lai.5Sro.5CuO3.75 (1st hour) at different temperatures, a : 648K; b: 623K; c : 613K.

91

Fig.3.22 Kinetic plot for the dehydrogenation of cyclohexanol over Lai.5Sro.5CuO3.75 at different temperatures a : 648K; b : 623K; c : 613K

•

3

(

lit. atm ./g.h

)

92

Fig.3.23 Arrhenius plot for the dehydrogenation cyclohexanol over Lai.5Sro.5CuO3.75

of

93

Table 3.12 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over La1.5Sro.5CuO3.75 W/F (g.hr/ g. mole

k -x'-21n(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH (kJ/mole)

Temperature 648 K 26.5

0.298

0.4096

35.3

0.377

0.5694

53.0

0.475

0.8137

0.8188 0.999xl03

32.9

Temperature 623 K 26.5

0.266

0.3525

35.3

0.356

0.5241

53.0

0.451

0.7480

0.6645 1.079xl03

33.1

Temperature 613 K 26.5

0.248

0.3220

35.3

0.331

0.4729

53.0

0.402

0.6263

0.5684 1.042xl03

Activation energy Ea = 38.3 kJ/moIe. Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream.

33.2

94

Tabic 3.13 Flow reactor data used for the model descrimination for the dehydrogenation of cyclohexanol on Sro.5La1.5CuO3.75 W/F g.hr/ g.mole

X/

CA

g. mole

CB g. mole

CC

r

g. mole

Temperature = 648 K

26.5

0.298

0.01017

0.00432

0.00432

0.00898

35.3

0.377

0.00851

0.00515

0.00515

0.00739

53.0

0.475

0.00669

0.00606

0.00606

0.00420

Coefficients for the polynomial a = 1.3761 x 10'2

b = -9.0249 x 10'5

Temperature = 623 K

26.5

0.266

0.01134

0.00411

0.00411

0.00856

35.3

0.356

0.00929

0.00514

0.00514

0.00734

53.0

0.451

0.00740

0.00608

0.00608

0.00489

Coefficients for the polynomial a = 1.223 x 10'2

b =-6.9264 xlO5

Temperature = 613 K

26.5

0.248

0.01198

0.00395

0.00395

0.00763

35.3

0.331

0.00999

0.00494

0.00494

0.00623

53.0

0.402

0.00848

0.00570

0.00570

0.00340

Coefficients for the polynomial a = 1.186 x 10'2

b = -7.9835 x lO:5

95

Table - 3.14 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on Sro.5La1.5CuO3.75

Model No.

1

2

3a

3b

Tempe­ rature K

CONSTANTS k

Ka

KC

Kb

648

0.800

0.7063

0.7423

0.7667

623

0.767

4.6289

4.4670

4.4667

613

0.533

-3.8420

-3.7409

-3.8000

648

1.167

0.6670

-3.2450

-

623

1.230

0.6000

0.6660

-

613

1.033

0.5330

-0.9047

-

648

1.167

0.6670

-1.1648

-1.0670

623

1.233

0.6000

-0.1578

613

1.067

0.5330

-0.5492

-0.5330

648

1.274

0.5869

-0.7927

-0.5256

623

1.093

0.5464

-0.8074

-0.7653

613

1.081

0.4806

-0.5894

-0.6172

0.1000

-

?6

Table - 3.14 (Continued)

Model No,

3c

3d

4

5

Tempe­ rature K

CONSTANTS k

Ka

KC

Kb

648

1.278

0.6932

2.1408

2.2372

623

1.110

0.6000

-0;3704

-2.0000

613

1.022

0.4934

-0.6168

-0.5900

648

1.295

0.5947

-1.6852

0.6888

623

1.195

0.6000

-1.7657

-0.5660

613

1.012

0.4120

-1.1504

0.6667

648

1.239

0.6670

-

-

623

1.196

0.6000

-

-

613

1.046

0.5330

-

-

648

0.021

1.3710

-

3.0418

623

0.021

1.3550

-

2.9510

613

0.021

1.6430

-

3.6439

97

Fig.3.24

The plot of effect of contact time on the conversion of cyclohexanol over Lai.3Sro.7CuC>4 (1st hour) at different temperatures, a : 633K; b: 613K; c : 593K.

98

-1 ) UI2 Fig.3.25 Kinetic plot for the dehydrogenation of cyclohexano! over Lai.3Sro.7CuC>4 at different temperatures a : 633K; b : 613K; c : 593K

log k

(

lit.

atm /g.hr )

99

1.58

1.62

1.66

1.70

T Fig.3.26 Arrhenius plot for the dehydrogenation cyclohexanoi over Lai.3Sr0.7CuO4

of

1DD

Table 3.15 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over Lai.3Sro.7CuO3.65 W/F (g.hr/ g. mole

k -x'-21n(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH (kJ/mole)

Temperature 633 K 13.3

0.375

0.5650

17.7

0.451

0.7483

26.5

0.575

1.1363

53.0

0.772

2.1848

2.0775 1.153xl06

64.36

Temperature 613 K 14.1

0.292

0.3986

26.5

0.405

0.6334

35.3

0.500

0.8862

53.0

0.622

1.3237

1.2574 1.074xl06

64.52

Temperature 593 K 14.1

0.180

0.2168

17.7

0.240

0.3089

26.3

0.357

0.5262

35.3

0.415

0.6573

53.0

0.540

1.0131

0.9147

Activation energy Ea = 69.62 kJ/mole. Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

1.239xl06

64.69

101

Table 3.16 Flow reactor data on the dehydrogenation of cyclohexanol on Sro.7La1.3CuO3.g5 W/F g.hr/ g.mole

9

X

Ca.

g. mole

Qb g. mole

Cci g. mole

r

Temperature = 633 K

13.3

0.375

0.00875

0.00525

0.00525

0.02249

17.7

0.451

0.00728

0.00598

0.00598

0.01983

26.5

0.575

0.00519

0.00703

0.00703

0.01450

53.0

0.772

0.00248

0.00839

0.00839

-0.01755

Coefficients for the polynomial a = 3.053 x 10'2

b = -3.024 x 104

Temperature = 613 K

14.1

0.292

0.01089

0.00449

0.00449

0.02711

26.5

0.405

0.00842

0.00573

0.00573

0.00417

35.3

0.500

0.00663

0.00663

0.00663

-0.01210

53.0

0.622

0.00463

0.00762

0.00762

-0.04483

Coefficients for the polynomial a = 5.3172 x 10'2

b = -9.243 x 104

Temperature = 593 K

14.1

0.180

0.01428

0.00314

0.00314

0.01260

17.7

0.240

0.01259

0.00398

0.00398

0.01190

26.5

0.357

0.00974

0.00541

0.00541

0.01020

35.3

0.415

0.00850

0.00603

0.00603

0.00851

53.0

0.540

0.00614

0.00721

0.00721

0.00509

Coefficients for the polynomial a = 1.5319 x 10'2

b = -9.6466 x 10'5

102

Table-3.17 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on Sro.7La1.3CuO3.65

Model No.

1

2

3a

3b

Tempe­ rature K

CONSTANTS k

Ka

Kb

KC

633

0.250

-1.2575

-1.7972

-1.8000

613

-2.025

-0.7694

-0.6186

-0.5844

593

0.920

-1.7742

-1.9150

-1.8980

633

2.025

0.3250

-5.5166

-

613

1.725

-0.750

-1.9634

-

593

1.360

0.7000

-1.5705

-

633

2.000

0.3000

-1.9528

-1.9500

613

1.650

-0.7500

-1.9547

-2.2750

593

1.520

1.1750

-0.6855

-1.7250

633

1.800

0.9796

-0.7119

-1.0296

613

1.415

0.9760

-1.1020

-1.0688

593

1.302

0.6954

-0.6260

-0.3452

103

Table - 3.17 (Continued)

Model No.

3c

3d

4

5

Tempe­ rature K

CONSTANTS k

Ka

Kb

KC

633

1.9302

0.7906

-2.3781

-2.3412

613

2.6933

3.3858

-2.4485

-2.3880

593

1.4060

0.6976

2.8239

2.9042

633

1.7692

0.8295

-1.4041

-1.3560

613

1.3974

1.1342

-1.7340

-1.0099

593

1.3483

0.6411

0.0994

-0.4780

633

1.9900

1.4000

-

-

613

1.7620

1.3000

-

-

593

1.3660

0.7000

-

-

633

0.6670

92.2200

-

3.4600

613

0.0280

-82.1500

-

1.6600

593

0.0320

-14.6640

-

3.1837

104

Table 3.18 Flow reactor data obtained by using integral method for the dehydrogenation of cyclohexanol over LaSrCu03.5 W/F (g-hr/ g. mole

k -x'-2in(l-x') (lit. atm./ x' at (1st hour) g. hr.)

A (1/h)

AH (kJ/mole)

Temperature 643 K 21.2

0.410

0.6450

26.5

0.492

0.8773

29.4

0.534

0.9930

35.3

0.600

1.2330

44.2

0.611

1.2770

0.2576 2.868xl016

194.16

Temperature 633 K 21.2

0.274

0.3664

26.5

0.320

0.4513

35.3

0.401

0.6240

53.0

0.505

0.9014

-0.0551

2.540xl016

194.24

3.330xl016

194.4

Temperature 613 K 21.2

0.119

0.1344

26.5

0.128

0.1459

35.3

0.172

0.2055

53.0

0.334

0.4789

-0.4746

Activation energy Ea = 199.5 kJ/mole.

Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

0.75

a

Fig.3.27 The plot of effect of contact time on the conversion of cyclohexanol over LaSrCu04 (1st hour) at different temperatures. a : 643K; b: 633K; c : 613K.

1D6

Fig.3.28

Kinetic plot for the dehydrogenation of cyclohexanol over LaSrCu04 at different temperatures a : 643K; b : 633K; c : 613K

log

k (

lit. atm

/ g.hr

)

107

Fig.3.29 Arrhenius plot for the cyclohexanol over LaSrCuC>4

dehydrogenation

of

108

Table 3.19 Flow reactor data used for the model descrimination for the dehydrogenation of cyclohexanol on SrLaCuOs.s W/F g.hr/ g.mole

t X

g. mole

CB g. mole

Cc.

g. mole

r

Temperature = 643 K

21.2

0.410

0.00793

0.00551

0.00551

0.01455

26.5

0.497

0.00637

0.00629

0.00629

0.01174

29.4

0.534

0.00576

0.00660

0.00660

0.01019

35.3

0.600

0.00473

*0.00711

0.00711

0.00708

44.2

0.611

0.00458

0.00719

0.00719

-0.00237

Coefficients for the polynomial a = 2.5776 x 10"2

b = -2.6486 x 10-4

Temperature = 633 K

21.2

0.274

0.01097

0.00414

0.00414

0.01061

26.5

0.320

0.00992

0.00467

0.00467

0.00952

35.3

0.401

0.00823

0.00551

0.00551

0.00773

53.0

0.505

0.00633

0.00646

0.00646

0.00412

Coefficients for the polynomial a = 1.4929 x 10'2

b = -1.0192 x 10"4

Temperature = 613 K

21.2

0.119

0.01565

0.00211

0.00211

0.00455

26.5

0.128

0.01536

0.00226

0.00226

0.00416

35.3

0.172

0.01404

0.00292

0.00292

0.00353

Coefficients for the polynomial a = 6.0920 x 10'2

b = -3.634 x 10'5

109

Table - 3.20 Parameter values obtained for different models for the dehydrogenation of cyclohexanol on SrLaCuC>3.5

Model No.

1

Tempe­ rature K 643

CONSTANTS k

Ka

1.460

Kb

KC

-1.2465

-1.7172

-1.7600

y

2

3a

3b

633

0.900

2.7317

2.6642

2.6473

613

0.300

18.5468

21.4050

21.4660

643

1.540

0.9400

-0.0826

-

633

1.300

0.6750

-0.1896

-

613

0.899

0.3000

-1.9766

-

643

1.520

0.9400

-0.6855

-0.6800

633

1.300

0.6750

-0.3274

-2.8750

613

0.899

0.3000

-1.1398

-1.1667

643

1.481

0.9233

-0.9620

0.9199

633

1.293

0.6943

-0.8584

-0.4750

613

0.829

0.3000

-0.8284

-0.2000

110

Table - 3.20 (Continued)

Model No.

3c

3d

4

5

Tempe­ rature K

CONSTANTS k

Ka

Kb

Kc

643

1.6976

1.0932

0.5434

0.5866

633

1.1303

0.6135

0.3995

0.3804

613

0.8856

0.3000

2.8736

2.8079

643

1.5413

0.9892

-0.0824

-0.9480

633

1.0842

0.5562

-1.3373

-1.1915

613

0.8831

0.3000

-1.1136

-0.0667

643

1.5150

0.9600

-

-

633

1.3110

0.6750

-

-

613

0.9120

0.3000

-

-

643

0.0260

-86.6680

-

2.8293

633

0.0220

3.8440

-

2.8028

613

0.0120

1.2590

-

4.1008

111

3.12

SUMMARY AND COMPARISON OF RESULTS OF THE CATALYTIC ACTIVITY OF La2.x Srx Cu04.y SERIES (x=0.0, 0.2, 0.3 0.5 0.7 & 1.0)

The investigation made for the different compositions of the series La2_x Srx Cu04.y (x=0.0, 0.2,0.3 0.5 0.7 & 1.0) by flow reactor technique have indicated that all the compositions are selective towards the dehydrogenation of cyclohexanol in the temperature range investigated (573 to 648 K). However, beyond 648K, all the catalysts favour dehydration of cyclohexanol yielding cyclohexene with decreasing percentage of cyclohexanone. All the compositions require identical activation and pretreatment procedures involving evacuation and treatment with hydrogen at 573 K for 4 hours. Integral method of analysis was adopted which was based on the assumption of first order kinetics. By plotting - x' - 21n (1-x') vs W/F, the rate constants of the catalytic reactions were obtained and the rate constants found to obey the Arrhenius equation for all compositions. The Ea values obtained for the different compositions in the series La2_x Srx CuO^ are given in Table 3.21. A comparison of the catalytic activity in the series La2_xSrx Cu04_y (x=0.0, 0.2, 0.3 0.5 0.7 & 1.0) shows interesting features. These include (i)

Existence of a compensation effect

(ii)

Variation of activation energy Ea with strontium concentration.

(iii)

Variation of heat of activation with strontium concentration.

(iv)

Variation of rate constant with strontium concentration.

3.12.1

Compensation effect

Linear variation of Ea with logarithmic Arrhenius frequency factor has been observed for La2.x Srx Cu04^ (x=0.0, 0.2, 0.3 0.5 0.7 & 1.0) series. This compensation effect is shown in Figure 3.30. The linearity in the figure indicates that the active sites responsible for the catalytic conversion of cyclohexanol to cyclohexanone are same for the all the compositions in the temperature range investigated.

112

Table 3.21 Ea, AH and InA values for the system SrxLa2-xCu04^,

Compound

^LT^OjCliOg g5

LaSrCuOs 5

Energy of Heat of activation (Eq) activation (AH) kJ/Mole kJ/Mole

Logarithmic frequency factor (InA)

41.60

36.67

7.75

45.47

40.54

8.69

35.60

30.06

6.73

38.30

33.06

6.95

69.60

64.52

13.96

199.50

194.26

37.90

In A

113

Fig.3.30 Plot of compensation effect on the dehydrogenation of cyclohexanol over La2-xSrxCu04-y

114

3.12.2

Variation of activation energy Ea with strontium concentration (x)

The relative activities of the composition can be compared on the basis of the variation of the energy of activation as a function of strontium concentration (Figure 3.31). From the figure, it is observed that the activation energy is found to be minimum for the composition x - 0.3 in the series La2-xSj'xCu04 and thereafter it increased with x. This increase in the activation energy with strontium concentration indicates that the different compositions in the series La2-xSrxCu04 are energetically different or it can be attributed to the change in surface composition i.e., surface enrichment of A site cations (Sr) which was also supported by increase in coke deposition as strontium content increased.

3.12.3

Variation of AH values with strontium concentration (x) The variation of AH with strontium concentration is also plotted in

Figure 3.31. A similar trend is observed for AH values vs strontium concentration as that of activation energy. The magnitude of AH of a reaction on a given catalyst depends on a)

The affinity of the adsorbate for the active site which in turn depends on the nature of the active site and

b)

On the quantity of reactant adsorbed i.e., the concentration of surface adsorbed-spedes. Since the nature of the sitesandhence the affinity towards the adsorbate are almost constant within the series La2-xSrxCu04-y, it is apparent that increase in AH values may be attributed to the increased concentration of surface adsorbed species.

3.12.4

Variation of rate constants with respect to strontium concentration

The rate constants for the different compositions in the series La2-xSrxCu04^ which catalysed the dehydrogenation of cyclohexanol is presented in Table 3.22. The rate constants increased with temperature indicating the assumption made (1st order) was correct. However, a plot of rate constant vs strontium concentration at 593 K (figure 3.32) shows a trend in which the increase in rate constant is observed from the compositions x = 0 to x = 0.7 and rate constant decreases for the composition x =1. A similar trend

(0

O

o) Activation energy (Ea) KJ/mole l

)

• (

/mole

-

O

KJ

KJ / mole

• AH KJ / mole Is

Ah

O Ea

Strontium concentration ( X ) Flg.3.31 Plot showing the variation of activation energy (Ea) and heat of adsorption AH with strontium concentration.

116

Tabic 3.22 Variation of rate constants with temperature for the different compositions in the series La2-x®rx^u^4-y (x - 0.0, 0.2, 0.3, 0.5, 0.7 & 1.0) Compound

La2Cu04

Lal^Sr0.2CuO3.9

*-ai.7Sro.3<-'u03.85

La1.5Sr0.5^uO3.75

^l.S^OjCllOa 65

LaSrCu035

Temperature (K)

Rate constants (k) lit. atm. / g.hr.

613

0.6539

593

0.5157

573

0.3709

613

0.8047

593

0.5839

573

0.4278

603

0.6926

593

0.6082

583

0.5381

648

0.8188

623

0.6645

613

0.5684

633

2.0775

613

1.2574

593

0.9147

643

1.8096

633

0.8809

613

0.3353

(

Rate constant ( k ) lit atm /g.hr )

Fig.3.32 Plot showing the variation of rate constant with strontium concentration at 593K.

11B

is observed at other temperatures also. The increase in rate constant values upto x - 0.7 can be attributed to the increase of the oxide ion vacancies (or) trivalent copper ions at the catalyst surface. Increase in x brings about an increase in oxidising power (dehydrogenating) but a decrease in reoxidation ability. A maximum activity at x - 0.7 would thus be obtained for which the redox cycle proceeds most readily (Nakamura et al 1981 and 1982). The third explanation possible is increase in oxide ion mobility with increasing x, facilitates the supply of active oxygen from the bulk to the surface catalytic sites. Thus increasing availability of oxygen, coupled with decreasing specific reactivity of oxygen brings about maximum activity at x = 0.7. 3.12.5

Conclusions based on reaction mechanisms

The values of k, KA, KB & Kc obtained for different module for different catalysts were used to arrive at the best model which fits the rate of the overall reaction. The models namely adsorption and desorption steps as rate controlling (model Nos. 1 & 5) were discarded because the k values obtained in most cases were not temperature dependant. The models 2, 3a, 3b, 3c and 3d were not accepted because of both negative values of KB and Kc (in some cases Ka also) and discrepancy in the trend of KA values with temperature. The model ‘surface reaction - dual site with negligible adsorption of products’ was selected as the best model on the basis of following criteria. 1)

The k & Ka values obtained for all the catalysts at all the temperatures is positive and significantly greater than zero and

2)

The reactant velocity constant and the adsorption equilibrium constant values are temperature dependant.

3.13

DECOMPOSITION OF CYCLOHEXANOL CATALYSED BY Ln2Cu04 (Ln = Pr to Gd) AND La2 Ni04

The vapour phase decomposition of cyclohexanol was also investigated for the following rare earth cuprates and nickelate. Pr2Cu04 Nd2Cu04 Sm2Cu04 Gd2Cu04 and La2Ni04.

119

They were investigated in the temperature range of 593 - 773 K using the flow reactor described in section 2.6. Invariably all the catalysts of the type Ln2Cu04 (Ln = Fr - Gd) yielded cyclohexene and cyclohexanone as the main products. However cyclohexene is the predominant product obtained which increased with temperature and low levels of dehydrogenated product was obtained. The percentage conversion and the distribution of products is given in Table 3.23. The corresponding figures for all the catalysts are given in Figures 3.33 to 3.37. 3.14

DISCUSSION ON THE CATALYTIC ACTIVITIES OF CUPRATES OF LANTHANIDES The different behaviour of La20uO4 with that of cuprates of other

lanthanides may be explained as follows. The compound La2Cu04, alone adopts the K2NiF4 (orthorhombic) structure in the true sense of it whereas for the other cuprates of lanthanides exhibit slight distortion. In Nd2Cu04 and Gd2Cu04, Cu2+ has been shown to have square planar coordination. Further, in Ln2Cu04 series, La2Cu04 alone crystallises in orthorhombic symmetry which is known to involve two long Cu-0 distances (2.4A) and four short Cu-0 distances (1.9A) whereas the Ln2Cu04 (Ln = Pr - Gd) system have four long and two short Cu-O distances (Shaplygin et al 1979). The short Cu-0 distances (1.9 A) are found to be responsible for the dehydrogenation of cyclohexanol. The same is true in the case of copper oxide in which the Cu-0 distance is 1.95A. Hence, it can be concluded that since there are more number of short Cu-O bonds in La2Cu04 than the other cuprates, this may be the reason for the effective dehydrogenation of cyclohexanol over La2Cu04. e

o

However, an acceptable reason for the different behaviour of La2Cu04 is explained by Wolkenstein Theory (Wolkenstein 1960). According to his electronic theory of dehydrogenation and dehydration, dehydrogenation is an acceptor reaction and dehydration is the donor reaction and the reaction path mainly depends on the mode of attachment of alcohol molecule to the surface. According to him the scission of O-H bond favours dehydrogenation and C-0 scission favours dehydration. Here the catalytic activity depends on the adsorption of cyclohexanol on copper. This adsorption will take place in a facile manner only when the electron density around copper is low. The fact that there are no outer most electrons in La3+ makes the net availability of copper for cyclohexanol adsorption through O-H scission favouring cyclohexanone formation.

120

Table 3.23 Percentage conversion of cyclohexanol catalysed by cuprates of lanthanides (Pr, Nd, Sm and Gd) and La2Ni04 Percentage of Products Compound

Pr2Cu04

Nd2Cu04

Sm2Cu04

Temperature

(K)

%of Cyclohexene

%of Cyclohexanone

573 593 613 633 673 698 723 748 573 593 633 673 698 723 748 773 573 593 613 633 673 698 723 748 773

1.8 6.9 8.9 14.3 41.5 70.2 77.6 89.9 3.4 7.1 22.3 62.2 73.2 76.2 79.4 88.4 3.0 8.0 9.0 21.8 43.5 42.3 71.3 91.5 93.4

1.8 5.8 8.2 10.6 10.2 6.4 3.2 3.2 0.2 0.8 3.9 4.4 5.5 7.7 7.5 5.3 0.3 0.7 0.2 1.1 3.3 5.5 3.4 5.0 3.5

121

Table 3.23 {Continued) Percentage of Products Compound

Temperature
Gd2Cu04

La2Ni04

%of Cyclohexene

%of Cyclohexanone

593

6.8

0.4

633

22.2

7.9

673

37.1

5.3

698

47.9

7.8

723

72.9

4.5

748

90.5

5.4

773

95.5

1.3

593

7.3

0.8

613

18.8

633

46.5

3.7

673

69.5

3.7

698

70.9

7.4

723

89.9

2.4

748

90.8

5.7

773

93.2

4.1

Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

.

Flow rate = 2 ml/hr.

2.2

1

570

—I--------- 1----------- 1--------- L______ 1_______ l______

610

650

690

1

730

Temperature ( K )

Fig.3.33 Plot of effect of temperature on the decomposition of cyclohexanol over Pr2Cu04 a : cyclohexene; b : cyclohexanone; c : cyclohexanol

0

________i_____________ i_____________ i__

600

680

760

Temperature ( K ) Fig.3.34 Plot of effect of temperature on the decomposition of cycloliexanol over Nd20uO4 a : cyclohexene; b ; cyclohexanone; c : cyclohexanol

124

O O 01

O

O

Conversion (%

00

O

20

-

600

680

760

Temperature ( K ) Fig.3.35 Plot of effect of temperature on the decomposition of cyclohexanol over Sm2Cu04 a : cyclohexene; b : cyclohexanone; c s cyclohexanol

____ 1_________I_________L_ 680

760

Temperature ( K)

Fig.3.36 Plot of effect of temperature on the decomposition of cyclohexanol over Gd2CuC>4 a : cyclohexene; b : cyclohexanone; c : cyclohexanol

Conversion {%)

12G

600

640

680

720

760

Temperature (K ) Fig.3.37 Plot of effect of temperature on the decomposition of cyclohexanol over La2Ni04 a : cyclohexene; b : cyclohexanone; c : cyclohexanol

127

Whereas, as the rare earth ion is changed to Pr3+, Sm3+, Nd3+ and Gd3+, the f-Pz orbital interaction between the rare earth ion and the oxide ion of the lattice would increase and consequently, the electron density around copper (Cu ) increases which retards the effective adsorption of cyclohexanol through O-H scission (Sivakumar and Sivasankar 1996). Hence the cuprates of other lanthanides favour C-O bond scission and thus dehydration. 3.15

DECOMPOSITION OF CYCLOHEXANOL BY ABOs TYPE OXIDES

The following catalysts of the type AB03 were tested for the catalytic activities with respect to the cyclohexanol conversion. 1. 2. 3. 3.15.1

LaCo03 LaNi03 LaMn03 Catalytic activity of LaCoOg

The catalyst (0.5 g) was pretreated with hydrogen at 400°C for 4 hrs. and packed in the flow reactor. The details of the flow reactor is described in section 2.6. The effect of temperature on the catalytic conversion of cyclohexanol was studied in the temperature range of 593 K to 748 K. The contact time (W/F) was maintained at 26.5 g. hr./g.mole. The percentage conversion is given in Table 3.24 and Figure 3.38 shews the selectivities towards cyclohexanone and cyclohexene. It is clear from the Figure 3.38 that LaCo03 favoured both dehydration and dehydrogenation equally. However, among the catalysts studied (of the type ABO3), LaCo03 is found to give maximum conversion to cyclohexanone (as high as 42%). 3.15.2

Catalytic activity of LaNiOg *

The effect of temperature for the pretreated LaNiOg catalyst on the conversion of cyclohexanol is given in the Table 3.25. The contact time is 26.5 g. hr./g. mole.

128

Table 3.24 The effect of temperature on the decomposition of cyclohexanol catalysed by LaCo03

Tempe­ rature (K)

Selectivity %of %of %of towards Cyclohexene Cyclohexanone Cyclohexanol Cyclohexanone

593

8.1

0.7

91.0

7.7

633

25.4

18.3

55.9

41.5

673

27.7

27.4

44.9

49.7

698

27.5

29.2

41.4

49.8

723

27.6

31.6

39.1

51.9

748

41.0

42.1

8.3

45.9

Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

Flow rate = 2 ml/hr.

Conversion (%!

129

01wLI;;I 590

650

710

770

Temperature ( K ) Fig.3.38 Plot of effect of temperature on the decomposition of cyclohexanol over LaCo03 a ; cyclohexene; b : cyclohexanone; c : cyclohexanol

130

Table 3.25 Effect of temperature on the decomposition of cyclohexanol catalysed by LaNiOa Temperature (K)

%of Cyclohexene

%of Cyclohexanone

% of unreacted Cyclohexanol

593

5.9

4.4

89.6

633

14.2

6.5

79.0

673

43.3

12.1

43.5

698

80.2

10.0

8.6

723

84.2

10.6

2.4

748

90.6

7.3

0.8

Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

How rate = 2 ml/hr.

131

Dehydrated product is predominant in the products. On increasing the temperature, all catalysts of the type AB03 give dehydration product in the temperature range studied between 593 and 823 K. LaNi03 catalyst gives maximum yield of cyclohexanone (12.1%) at 673 K.

3.15.3

Catalytic activity of LaMn03

The experimental data for the conversion of cyclohexanol catalysed by LaMn03 is given in Table 3.26. From the conversion values, it is concluded that LaMn03 did not favour dehydrogenation reaction significantly. Only 10.3% of cyclohexanone was obtained at 748 K.

3.15.4

Summary and comparison of the results on the catalytic activity of ABO3 type compounds.

The studies made for the mixed oxide catalysts of the type AB03 by flow reactor technique have indicated that all catalysts lack selectivity. These AB03 type catalysts yielded both cyclohexene and cyclohexanone as products of which only cyclohexene predominates. However, LaCo03 favours both dehydrogenation and dehydration equally.

132

Table 3.26 Effect of temperature on the decomposition of cyclohexanol catalysed by LaMn(>3 Temperature (K)

%of Cyclohexene

% of Cyclohexanone

% of unreacted Cyclohexanol

, 593

8.6

0.6

90.6

633

33.8

0.7

64.9

673

22.9

1.8

74.7

698

32.6

7.0

59.5

723

71.2

4.8

21.4

748

75.6

10.3

10.6

773

84.8

7.2

2.2

Reaction conditions : Catalyst Weight = 0.5 g. Data taken after 1 hr. of time on stream

Flow rate = 2 ml/hr.

133

3.16

CATALYTIC DECOMPOSITION OF CYCLOHEXANOL BY MODIFIED ZEOUTE-Y

As discussed in Chapter 1, impregnation and / or exchange of metal ions over zeolite-Y is also one of the methods of stabilizing the metal ions from sintering. In the present investigation, both metal impregnated zeolites and metal exchanged zeolites were taken. They were characterised by different physico chemical techniques and their catalytic activities were evaluated for the conversion of cyclohexanol. 3.16.1

Characterisation of metal impregnated and metal exchanged zeolites

The sodium form of zeolite prepared by hydrothermal synthesis was found to have Si02/Al203 ratio of 2.57. The phase formation and the crystallinity were established by X-ray diffractometry using CuKa radiation and the patterns were compared with those of Linde-Y sample (Fig. 3.39). The BET surface areas for NaY, metal impregnated and exchanged zeolites were measured at liquid nitrogen temperature using nitrogen as adsorbate (Table 3.27). Using the flame photometer, thermo gravimetric analyser, inductively coupled plasma and wet chemical analyses, the unit cell formula for the synthesised zeolite-Y was arrived Na54 (A102)54 (Si02) 138 The moisture content of sodium form of zeolite found out by using thermo gravimetric analyser was 21.8%. Acidity of zeolites can be investigated by a number of techniques such as IR, DSC, TFD (using probe molecules such as ammonia, pyridine or amines) and more recently by solid state NMR spectroscopy (Kenaston et al 1994, Gil et al 1994, Pfieter et al 1991). In the present investigation, acidities of both metal impregnated and metal exchanged zeolites were determined by Differential Scanning Calorimetric technique (DSC) which involved pyridine adsorption, followed by thermal desorption adopting Aboul Gheit’s (1987) nullifying technique as described earlier in section 2.4.6.

SdO

0

2.00

4.00

2 0 (degrees

30 )

40

Fig.3.39 XRD pattern* of NaY

20

50

60

134

135

Table 3.27 BET surface area of metal impregnated and metal exchanged zeolite-Y S.No. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Catalyst

Surface area m2/s

NaY

494.2

Cu(22)NaY

446.6

Cu(26)NaY

438.5

Cu(29.3)NaY

425.8

Cu(34.8)NaY

420.8

Cu - 5 - NaY

362.7

Cu-10-NaY

323.9

Cu -15 - NaY

314.1

Co - 5 - NaY

385.1

Co -10 - NaY

372.4

Co -15 - NaY

363.3

Ni - 5 - NaY

393.6

Ni -10 - NaY

369.1

Ni -15 - NaY

351.5

The numbers in the bracket indicate the percentage of exchange.

iis 3.16.1.1 Acidity characterisation of copper impregnated and exchanged zeolites Pyridine adsorbed on copper impregnated / exchanged zeolites (12 to 14 mg.) were taken in the sample cell and an equal weight of pyridine-free samples were taken in the reference cell of DSC. Samples were investigated in the temperature range of 50-600°C and no purge gas was used. «

Figure 3.40 shows the DSC thermograms of 5,10 & 15 % of copper impregnated zeolites and the Figure 3.41 shows the DSC thermograms of different levels of copper exchanged zeolites. The DSC data for copper impregnated and exchanged zeolites are given in Tables 3.28 and 3.29.

3.16.1.2 Acidity studies on cobalt and nickel impregnated zeolites Cobalt and nickel impregnated zeolites were also characterised for their acidity using DSC. Figure 3.42 shows the DSC thermograms of cobalt impregnated zeolites and Figure 3.43 for nickel impregnated zeolites and their corresponding data are given in Tables 3.30 and 3.31.

3.16.1.3 Summary and discussion on the acidity of zeolites The DSC thermograms of Cu, Co and Ni impregnaed zeolites and copper exchanged zeolites indicate that pyridine had been chemisorbed over the acidic sites of these zeolites. Acidic sites should have been introduced during impregnation / exchange of metal ions which can be explained by considering the equations 3.31 and 3.32.

nM(H2OpM +

===^1

(M(H20)X + IM(H20)2+]v Nax.2yZ+(Na+)2y

(M(H20)2+)n^, + [M(H20)2+1v Nax.2yZ+ (Na+)2y (MO)m + [M(OH)]+y Bronsted acidic sites (y « n)

+ Na20

(3.31)

A ^ •«?

...... ....

(3.32)

Fig.3.40

DSC thermograms showing pyridine desorption from 5, 10, 15% copper impregnated zeolites

Heat flow

exo

138

IOO

300

500

Temperature ( °C) Fig.3.41 DSC thermograms showing pyridine desorption from copper exchanged zeolites a : Cu(22)NaY; b : Cu(26)NaY; c : Cu(29.3)NaY; d : Cu(34.8)NaY

139

Table 3.28 Differential scanning calorimetric data and their assignments for copper impregnated zeolites . S

Sample code

Cu-5-NaY

Cu-10-NaY

Cu-15-NaY

PPY

-

Peak Temperature (°C)

(J/s)

86.7

12.4

PPY + WPY

197.7

85.7

HPY

403.5

995.3

CuPY

80.4

9.7

PPY + WPY

197.7

58.1

HPY

396.0

859.0

CuPY

63.3

10.9

PPY + WPY

213.3

73.3

HPY

387.8

540.8

CuPY

AH

Pyridine desorption from physisorbed pyridine

WPY -

Pyridine desorption from weak acidic sites

HPY -

Pyridine desorption from hydrogen bonded pyridine

CuPY -

Assignment

Pyridine desorption from copper - pyridine complex

140

Tabic 3.29 Differential scanning calorimetric data and their assignments for copper exchanged zeolites

Sample code

Peak Temperature (°C)

(J/s)

Cu(22)NaY

159.3

135.5

HPY

230.0

27.3

WLPY

454.9

3308.0

SLPY + CuPY

162.0

164.0

HPY

230.0

20.2

WLPY

448.0

3252.0

SLPY + CuPY

167.0

160.0

HPY

224.0

13.0

WLPY

448.0

3164.0

SLPY + CuPY

167.0

142.0

HPY

234.0

12.0

WLPY

447.0

3124.0

SLPY + CuPY

Cu(26)NaY

Cu(29.3)NaY

Cu(34.8)NaY

AH

Assignment

HPY -

Pyridine desorption from hydrogen bonded pyridine

WLPY-

Pyridine desorption from weak Lewis acidic sites

SLPY -

Pyridine desorption from strong Lewis acidic sites

CuPY -

Pyridine desorption from copper - pyridine complex

exo Heat flow

Temperature { C) Fig.3.42 DSC thermograms showing pyridine desorption from 5, 10, 15% cobalt impregnated zeolites

143

Table 3.30 Differential scanning calorimetric data and their assignments for cobalt impregnated zeolites

Sample code

Co-5-NaY

Co-10-NaY

Co-15-NaY

PPY

-

Peak Temperature

AH

Assignment

<°q

(J/g)

89.5

25.2

PPY + WPY

203.7

51.2

HPY

407.5

1206.3

CoPY

71.5

18.3

PPY + WPY

190.0

38.3

HPY

400.6

1183.0

CoPY

66.3

11.7

PPY + WPY

182.4

35.9

HPY

379.7

1006.6

CoPY

Pyridine desorption from physisorbed pyridine

WPY -

Pyridine desorption from weak acidic sites

HPY -

Pyridine desorption from hydrogen bonded pyridine

CoPY-

Pyridine desorption from cobalt - pyridine complex

144

Table 3.31 Differential scanning calorimetric data and their assignments for nickel impregnated zeolites

Sample code

Ni-5-NaY

Ni-10-NaY

Ni-15-NaY

Peak Temperature (°C)

AH (J/g)

63.5

27.4

PPY + WPY

198.9

70.9

HPY

444.6

1229.6

NiPY

79.8

15.2

PPY + WPY

196.9

65.3

HPY

450.5

988.2

NiPY

81.6

11.9

PPY + WPY

184.3

61.5

HPY

438.0

774.8

NiPY

PPY -

Pyridine desorption from physisorbed pyridine

WPY -

Pyridine desorption from weak acidic sites

HPY -

Pyridine desorption from hydrogen bonded pyridine

NiPY -

Pyridine desorption from nickel - pyridine complex

Assignment

145

According to Equation (3.31), during impregnation, there could be the possibility of exchange of some of impregnating metal ion with sodium ions present in zeolite and upon calcination, all the metal ions that are impregnated except those of exchanged would have been converted into oxides. Thus, during aqueous impregnation of metal ions, a smaller amount of metal ions (say y) (y « n) is exchanged for sodium ions and (n-y) quantity of metal ions and 2y sodium ions remain on the surface of the zeolite. Equation (3.32) denotes the introduction of Bronsted acidic sites in to the lattice upon impregnation. These exchanged metal ions (although they are small in number), due to their polarisation power over the adsorbed water, could promote the formation of H+ ions which could interact with the framework oxygens producing OH groups bound to the structure, introducing the Bronsted acidity in the zeolite. This was supported by Ward as early as 1968 and also by Pietri de Garcia et al (1989). At the calcination temperature, due to dehydroxylation, Lewis acidic sites were also generated. As far as exchange of metal ions is concerned, M(H20)Z2+ Nax.2zZ is obtained in Equation (3.31) where z » y when compared. However, M(H20)n_z2+ and (Na+)2z are washed during exchange. Hence the metal ions will be available only in their exchanged form. Scheme 3.3 explains the exchange of metal ions and subsequent formation of Bronsted and Lewis acidic sites. Hence, when pyridine was adsorbed over these zeolites, it could adsorb over the different acidic sites of zeolites that were generated during impregnation followed by calcination and also over the exchanged metal ions which form metal-pyridine complex. To confirm the exchange of metal ions further, thermo gravimetric analyses of pyridine adsorbed, metal impregnated zeolites were made. The Figure 3.44 shows the TGA thermograms of pyridine adsorbed, copper, cobalt and nickel impregnated zeolites. Figure 3.45 shows the TGA thermogram of copper exchanged sample. In the thermogram, the weight loss around 400°C is due to decomposition of pyridine from metal pyridine complex. This was further confirmed by complexing metal ion with pyridine and investigating its decomposition using TGA. This metal could be from that of exchanged and the remaining metal ions are converted to metal oxides which cannot form complex with pyridine. The heterogeneity of acidic sites of both metal impregnated / exchanged zeolites discussed above can also be determined by titrimetry, IR spectral and DSC studies. However, quantitative information about the strength and the density of acidic sites could be obtained only by DSC technique. In the

146 © Na 0 Ov / \0/\

\ Si

At

© Na

0 / \

0 / \o/

Si

Si

Al

/ \/ \/ \/ \/ \ Cu (NOj )2 * 2H20

Scheme I

[CulH^iQ v

0

Ov

2+

0

0

l

✓

\ / \o/ \ / \ / V./

^ Si Al Si Si M /\/\/\/\/\

Scheme II [Cu ( OH ^ H

I

/0

\

s

Si

0 /\ Al

0 /\ Si

0 /°\o/ Si

Al

/ \/ \/ \/ \/ \ ( Bronsted )

A Scheme 111

v \

Al

450 C

Dehydroxylation

0 / \

© Si

A

0 / \ Si

O , / v>/ Si

Al

*

/\/\/\/ \/ \ l Lewis )

Scheme 3.3

Exchange of metal ions and subsequent formation of Bronsted and Lewis acid sites

147

4)

>

O

a> > i

200

400 600 Temperature (°C)

800

Flg.3.44 TGA thermograms showing pyridine desorption from metal impregnated (5%) zeolites a : copper; b : cobalt; c : nickel

DTG

axis

WEIGHT GAIN

+

DTG axis

148

__|__________ j__________ |__________ |_

200

400

600

800

Temperature (°C) Fig.3.45 TGA thermograms showing pyridine desorption from copper exchanged zeolites a : 22%; b : 26%; c : 29.3; d : 34.8

149

DSC thermograms, the number of peaks indicate the number of acidic sites, the peak temperature indicates the strength of acidic sites, and the magnitude of AH gives information about the density of acidic sites. The DSC thermograms of Cu, Co and Ni impregnated zeolites (Table 3.28, 3.30 and 3.31) show three distinct desorption signals in the temperature ranges of (i)

63.3 —89.5°C

(ii)

182.4 —213.3°C

(iii)

379.7 — 450.5°C

The desorption peak observed at 63.3 — 89.5°C is a weak signal which may be due to desorption of pyridine from very weak acidic sites or from physisorbed pyridine, the peak obtained at 182.4 — 213.3°C may be due to the desorption of pyridine from hydrogen bonded pyridine which is also an indirect measure of Bronsted acidic sites and the sharp signal observed at 379.7 — 450.5°C is due to pyridine desorption from metal-pyridine complex. The third signal clearly indicates that some amount of metal ions had been exchanged in zeolite and these exchanged metal ions are responsible for the formation of acidic sites (Sivakumar and Sivasankar 1996a). However, the DSC thermograms of copper exchanged zeolites (Fig. 3.4-1) are quite different from that of impregnated zeolites. The DSC thermograms of copper exchanged zeolite show two distinct peaks in the temperature ranges (i)

159 — 167°C (small) and

(ii)

224 — 234°C (small)

.

and a broad but unresolved peak, the first one approximately corresponding to 320°C and the second corresponding to the temperature range of 447 — 455°C are observed. The desorption peak obtained in the temperature range of 159 — 167°C may be assigned to the desorption of hydrogen bonded pyridine (Sivakumar and Sivasankar 1995). The peak observed in the temperature range of 224 — 234°C can be explained as follows. Since the zeolite has higher Si /

150

AI ratio (equal to 2.57), there may be the possibility for the presence of different aluminium environments. Literature shows evidences that isolated aluminium frame work atoms (having no next nearest aluminium neighbours) have the highest strength and as the number of next nearest aluminium atom increases, the acid strength is likely to decrease (Barthomeuf 1987, Barthomeuf and Beaumont 1973, Dempsey 1974 & 1975). There could be four such different acid strengths corresponding to aluminium atoms 0, 1, 2, or 3 Al nearest neighbours (Mikovsky and Marshall 1976, Peters 1982). This study has indicated that there are two such different aluminium topologies existing in the samples/ of which one is weak from which pyridine desorbs in the temperature range of 224-234°C and the slightly stronger one from which pyridine desorbs in the temperature range of 447 - 455°C along with desorption from copper pyridine complex. The third peak obtained at 447-455°C is the unresolved broad peak which may be the mixture of two desorption signals comprising of strong Lewis acidic site as discussed above and the pyridine desorption signal from copper pyridine complex (Sivakumar and Sivasankar 1996b). The desorption signal from the strongest Bronsted acidic site is not observed in the temperature range investigated as it requires very high temperature (>600°C). The thermograms of both metal impregnated and exchanged zeolites show a decreasing trend in their AH values as the percentage of the metal ion increases. This could be explained pn the basis that, as far as impregnation is concerned, the exchange of metal ions into the lattice is small. Therefore, they are not readily accessible to pyridine molecule. The lack of exposure to pyridine molecule would be more, when more and more metal ions are converted to metal oxides which mask the exchanged metal ions. In the case of copper exchanged zeolite also, the same trend in AH values is obtained, despite the fact that there is no copper oxide on zeolites. This could be explained as follows. The third thermogram obtained in the case of copper exchanged zeolite is actually a mixture of two signals namely from strong Lewis acidic sites and copper - pyridine complex. The copper ions occupy the inaccessible sites at low levels of exchange and as exchange level increases, they occupy the accessible sites (super cages) within the zeolite where it can interact with the acidic sites and decrease the acidity. This decrease in acidic strength is more inspite of increasing copper ions at high levels of exchange. The trend of decreasing acidic strength

151

is also observed by Ward (1975) with various levels of nickel exchanged hydrogen zeolites. Thus, the DSC studies of metal impregnated and metal exchanged zeolites confirm the presence of different types and densities of acidic sites. The confirm this further, IR spectra of pyridine adsorbed, copper exchanged samples were recorded using IR spectrophotometer (Perkin-Elmer 983G). IR spectra of a representative sample is given in Fig. 3.46 and the observed wave numbers along with their corresponding assignments are given in Table 3.32. The conversion of cyclohexanol to cyclohexene is a model reaction most used to study the zeolite acidity particularly Bronsted sites. Since no information could be obtained about the Bronsted acidity from the DSC studies, cyclohexanol dehydration was studied over copper exchanged zeolites at 190°C (W/F = 53 g.hr/g.mole). Cu(22)NaY yielded cyclohexene as much as 90% with 98% selectivity and as the percentage of copper increased from 22% to 34.8%, the conversion decreased to 82% with same percentage of selectivity. Hence, DSC, IR spectral and catalytic activity studies confirmed the presence of different types of acidic sites with varying strengths.

3.17

EVALUATION OF CATALYTIC ACTIVITIES OF METAL IMPREGNATED AND EXCHANGED ZEOLITES

The catalytic decomposition of cyclohexanol was investigated for the following catalysts. i)

5, 10, 15% of copper impregnated zeolites,

ii)

5, 10, 15% of cobalt impregnated zeolites,

iii)

5, 10, 15% of nickel impregnated zeolites and

iv)

copper exchanged zeolites.

T ransm ittance

(%)

80

40

1800

1600

1400

Wavenumber (cm~*} Fig.3.4 6 IR spectra of pyridine adsorbed copper exchanged zeolites.

153

Table 3.32 IR wavenumbers and their corresponding assignments of pyridine adsorbed copper exchanged zeolite Observed IR wave number (cm"*)

Assignment

1446

HPY + LPY

1490

HPY + LPY + BPY

1550

BPY

1599

BPY

1631

LPY + BPY

HPY

-

Hydrogen bonded pyridine

LPY

-

Pyridine adsorbed on Lewis acid sites

BPY

-

Pyridine adsorbed on Bronsted add sites.

154

3.17.1

Catalytic decomposition of cyclohexanol catalysed by copper impregnated zeolites

The catalytic activities of 5, 10, 15% copper impregnated zeolites have been investigated for the dehydrogenation of cyclohexanol by flow reactor technique. The copper impregnated zeolites were first pretreated with hydrogen at 673K for 4 hours. The catalytic activities were investigated by taking 0.5 g of catalyst in the temperature range of 473K to 823K. Samples collected at the first hour were subjected to chromatographic analysis. The temperature effect on the catalytic conversion of cyclohexanol over 15% copper impregnated zeolite-Y is given in Table 3.33. From the Table, it was found that the yield of cyclohexanone was found to be low at low temperature whereas the dehydrated product cyclohexene which was insignificant at low temperatures increased abruptly and reached its maximum at 773K and decreased from 773 to 823K. At high temperatures, in addition to the predominent product cyclohexene, small amounts of methyl cyclo pentene isomers and benzene were also form©!. At 823K, the dehydrogenated product cyclohexanone which was low (8%) at the initial hours found to increase and reaches the maximum of 31.2% at the 8th hour (but at low temperatures the % conversion decreased with time). Hence, for the other samples (5% and 10% copper impregnated zeolites), investigations were made only at 823K. The effect of time on stream on the conversion of cyclohexanol catalysed by 5,10,15% copper impregnated zeolites at 823 K is given in Table 3.34.

3.17.2 Catalytic decomposition of cyclohexanol catalysed by cobalt impregnated zeolites The effect of temperature on the catalytic conversion of cyclohexanol over 15% cobalt impregnated zeolite is given in Table 3.35. The effect of time on stream on the conversion of cyclohexanol catalysed by 5, 10, 15% cobalt impregnated zeolite is given in Table 3.36.

155

Table 3.33 Effect of temperature on the decomposition of cyclohexanol catalysed by 15% copper on sodium form of zeolite Temperature K Products 523 4-Methyl cyclo pentene 3-Methyl cyclo pentene 1-Methyl cyclo pentene Cyclohexene Benzene

573

598

623

673

723

773

0.8

-

-

-

-

-

-

-

-

-

-

-

-

1.1

4.3

-

-

-

-

0.8

1.0

1.2

1.1

3.1

-

823

12.9

10.1

14.6

39.0

72.0

87.6

70.8

—

—

*

0.9

1.34

2.5

10.6

5.7

9.8

2.7

8.0

■-

.

Cyclohexanone

9.6

Catalyst Wt. =

0.2 g

Row rate

2 ml /hr.

12.1

16.8

4.1

156

J

Tabic 3.34 Effect of time on stream on the conversion of cyclohexanol catalysed by copper impregnated zeolites at 823 K % Yield of Cyclohexanone Time in hours 5%

10%

15%

1

3.6

1.1

8.0

2

9.9

8.7

15.3

3

15.2

18.6

15.1

4

18.8

26.0

21.3

5

24.8

30.3

27.8

6

26.2

33.4

29.3

7

29.7

35.2

29.4

8

31.5

35.8

31.2

Catalyst Wt. =

0.2 g

Flow rate

2 ml /hr.

157

Table 3.35 Effect of temperature on the decomposition of cyclohexanol catalysed by 15% cobalt on sodium form of zeolite Temperature K Products 523 4-Methyl cyclo pentene 3-Methyl cyclo pentene 1-Methyl cyclo pentene Cyclohexene

573

623

673

723

773

823

-

-

-

-

-

0.8

2.3

-

-

-

-

1.2

2.3

4.7

~.

-

-

-

1.7

3.7

7.8

3.21

8.1

77.6

93.7

88.6

79.0

68.7

—

—

—

—

6.9

8.4

12.7

1.6

18.3

1.4

1.2

0.96

2.1

3.0

Benzene Cyclohexanone

Catalyst Wt. =

0.2 g

Flow rate

2 ml /hr.

150

Table 3.36 Effect of time on stream on the conversion of cyclohexanol catalysed by cobalt impregnated zeolites at 823 K

% Yield of Cyclohexanone Time in hours 5%

10%

15%

1

1.3

1.7

3.0

2

8.0

12.7

18.3

3

11.3

21.0

24.6

15.4

22.8

25.3

5

17.2

23.9

27.8

6

23.3

24.9

26.3

7

29.7

23.4

28.1

8

28.9

26.3

27.9

4

,

Catalyst Wt. =

0.2 g

Flow rate

2 ml /hr.

159

3.17.3 Catalytic decomposition of cyclohexanol catalysed by nickel impregnated zeolites Tables 3.37 shows the corresponding effect of temperature for 15% nickel impregnated zeolites and Table 3.38 shows the effect of time on stream on the conversion of cyclohexanol catalysed by 5, 10, 15% nickel impregnated zeolites at 823K. 3.17.4 Catalytic decomposition of cyclohexanol catalysed by copper exchanged zeolites The exchange of copper in the sodium form of zeolite has been carried out as described in section 2.3.4. The effect of temperature on the catalytic conversion of cyclohexanol catalysed by a representative copper exchanged zeolite is given in Table 3.39. But the copper exchanged zeolite found to favour dehydration and carbon skeletal isomerisation. A small increase in the percentage of cyclohexanone was observed on increasing time on stream. 3.18

SUMMARY AND DISCUSSION ON THE CATALYTIC DECOMPOSITION OF CYCLOHEXANOL BY METAL IMPREGNATED AND METAL EXCHANGED ZEOLITES

The metal impregnated zeolites catalysing the decomposition of cyclohexanol yielded both cyclohexene and cyclohexanone at low temperatures and the percentage of cyclohexene increased from a low value at 553K to a high value at 673K. This can be explained by considering the fact that the acidic sites introduced during impregnation are not easily accessible at 553K as they were masked by metal oxites and as temperature increased to 673K, the acidic sites became accessible and hence favours dehydration. Whereas, for the copper exchanged zeolites, since there is no coverage of metal oxides, the acidic sites are accessible even at 473K which favours dehydration to cyclohexene. However, both metal impregnated (673K onwards) and metal exchanged zeolite (473K onwards) show a decreasing trend in the percentage of cyclohexene. This is because of diminution of Bronsted acidic sites observed as temperature increased. This decrease in Bronsted acidic sites is more and more on increasing time on stream. Hence, at high temperature, on increasing time on stream the percentage of cyclohexanone increased and reached its maximum.

160

Table 3.37 Effect of temperature on the decomposition of cyclohexanol catalysed by 15% nickel on sodium form of zeolite Temperature K Products 523 4-Methyl cyclo pentene 3-Methyl cyclo pentene 1-Methyl cyclo pentene Cydohexene

573

623

0.8

-

-

1.4

1.9

2.3

-

-

0.9

2.3

3.6

12.8 61.03

78.9

76.3

71.8

76.3

13.4

14.3

15.1

0.8

1.2

1.4

-

-

-

-

Catalyst Wt. =

0.2 g

Row rate

2 ml /hr.

823

-

-

0.8

773

-

-

.

Benzene Cyclohexanone

723

-

-

7.07

673

13.3

3.3

1.3

161

Table 3.38 Effect of time on stream on the conversion of cyclohexanol catalysed by nickel impregnated zeolites % Yield of Cyclohexanone Time in hours 5%

10%

15%

1

0.89

1.09

1.4

2

1.3

3.1

16.3

3

7.3

11.8

27.9

4

13.4

23.6

29.4

5

19.3

21.4

32.1

6

23.6

27.8

33.3

7

24.2

29.8

38.1

8

24.8

31.4

38.9

9

29.38

38.3

41.2

Catalyst Wt. =

0.2 g

Flow rate

2 ml /hr.

162

Table 3.39 Effect of temperature on the decomposition of cyclohexanol catalysed by copper exchanged zeolite (34.8% exchange) Temperature K Products 473

523

573

623

673

723

773

823

4-Methyl cyclo pentene

0.35

2.54

8.4

10.3

5.68

4.13

2.01

1.01

3-Methyl cyclo pentene

0.26

0.58

2.17

5.2

9.53

8.71

9.07

4.49

1-Methyl cyclo pentene

0.93

2.77

8.9

17.2

24.3 21.57 21.41

10.4

Cyclohexene

90.6 87.94 77.77

62.9 56.84

Benzene

0.06

0.18

0.57

0.57

0.6

0.45

0.32

0.5

1.27

Cyclohexanone 0.43

0.47

Catalyst Wt. =

0.2 g

Flow rate

2 ml /hr.

62.4

57.4 54.01

4.6

7.01

2.52 15.92