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