b Faculty of Science, University of Aleppo, Aleppo, Syria. Received 10 January 1995; revised 25 April 1995; accepted 27 April 1995. Abstract. ZnO-A&O3 ( 1 .
August 1995
ELSEVIER
Materials Letters 24 ( 1995) 3 19-325
Effect of thermal treatment on the structural, textural and catalytic properties of the ZnO-A1203 system Th. El-Nabarawy a,*, A.A. Attia a, M.N. Alaya b aNational Research Centre, Dokki, Cairo, Egypt b Faculty of Science, University of Aleppo, Aleppo, Syria Received 10 January 1995; revised 25 April 1995; accepted 27 April 1995
Abstract ZnO-A&O3 ( 1.O:1.011catalysts were prepared by mechanical mixing and co-precipitation. Four thermal products were obtained from each preparation by calcination, in air, at 300,500,700 and 1000°C. XRD and DTA indicated the absence of any crystalline phase upon calcination below 5OO”C, ZnAl,O, is obtained at 700°C. The crystallinity of this spine1 increases upon further rise of calcination temperature. The surface area decreases and the mean pore radius increases with the rise of the calcination temperature above 500°C. The conversion of 2-propanol proceeds through dehydration to propylene and dehydrogenation to acetone with the activity depending on the calcination temperature. Cracking of cumene depends on the chemical composition, method of preparation and the reaction temperature.
1. Introduction Mixed
oxides
are widely
used in the field of adsorp-
They are frequently superior to simple metal oxides with regard to the catalytic performance, long lifetime and resistance to sintering [ $61. Mixed oxides are prepared by mechanical mixing [7], co-precipitation [ 81, surface coverage [9], impregnation [ lo] or treatment of mixed precipitates [ 111. Mixed oxides are used in many catalytic reactions of great importance in chemical and petrochemical industries including cracking [ 121, hydrogenationdehydrogenation [ 131, reforming [ 141, dehydration [ 151 and others. Mor’e recently mixed oxides are used in pollution control via adsorption and/or catalysis [ 161. There are many factors contributing to the performance (activity and selectivity) of mixed oxide catalysts. These are namely, method of preparation, tion and catalysis
[ l-41.
* Corresponding author. 0167-577x/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved SSDZO167-577x(95)00101-8
chemical composition [ 171, textural properties [ 61, chemistry of the surface [ 91 and the conditions under which the reaction proceeds [ 181. The present investigation deals with the effect of calcination temperature on: (i) the structural and textural properties of ZnO-A1,03 catalysts and (ii) their catalytic properties towards conversion of 2-propanol and cracking of cumene.
2. Experimental 2. I. Materials Alumina, A, was prepared by the addition of the appropriate amount of N&OH to 1.0 M solution of A1,SOI *9H,O. The pH was adjusted at 8.0. The gel was left for decantation, filtered, washed with bidistilled water until it was free of SOZ- ions and then dried
320
Th. El-Nabarawy et al. /Materials Letters 24 (1995) 319-325
at 120°C. Zinc hydroxide, Z, was precipitated by the addition of 2.0 M NaOH to 1.O M ZnSO, at pH = 8.0. The precipitate was left for decantation, filtered, washed with bidistilled water until it was free of SO:- ions, and dried at 100°C. ZnO/Al,O, ( 1.O:1.O) was prepared by two different methods: M was prepared by mechanical mixing of alumina gel and zinc hydroxide using an electric mixer for 5 h and the mixture was dried at 90°C. C was prepared by co-precipitation as follows: 1.O M aluminium sulphate and 1.O M zinc sulphate were simultaneously dropwise added to distilled water. The pH was adjusted at 8.0. The solution was left for decantation, filtered and dried at 90°C. The dried precipitations were washed with bidistilled water until they were free from SO:- and then dried at 120°C. Four thermal products were obtained from each preparation by calcination of A, Z, M and C at 300, 500, 700 and 1000°C for 4 h. In designating the samples, A stands for A1,03, Z for ZnO, M for mechanically mixed ZnO-A1,03 and C for co-precipitated ZnOAl,O,. The number following any of these letters refers to the thermal treatment temperature (in degrees Celsius) .
2.2. Techniques
The differential thermal analyses (DTA) of the original preparations were obtained in the range 20-1000°C using a Netzsch DTA model 404 EP. Alumina crucibles were used together with a Pt 10% Rb-Pt thermocouple. For all the samples, the heating rate was lO”C/min. X-ray diffraction patterns (XRD) of M500, M700, Ml000 and C700 were obtained using a Philips diffractometer type PW 105 1. The patterns were run with nickel-filtered copper radiation (h = 1.5405 A) at 36 kV and 16 mA, the diffraction angle, 20, was scanned atarateof2”min’. The textural properties (surface area and porosity) of the different thermal products were determined from the adsorption of nitrogen at 77 K, using a conventional volumetric apparatus. The catalytic conversion of 2-propanol at 240-340°C and the catalytic cracking of cumene at 360-500°C were carried out in a reactor attached directly to a PYEUnicam chromatograph model 104 assembly.
3. Results and discussion 3.1. Differential thermal analysis Fig. 1 shows the DTA curves of the original preparations. Preparation A exhibited continuous endothermic effects between 120 and 350°C. In this temperature rage, physisorbed water is removed at low temperatures while specifically adsorbed water is removed at relatively high temperatures [ 191. The essential welldeveloped endotherm of preparation A shows its maximum at 560°C this may stand for the formation of aluminium oxide according to the reaction: 2Al( OH) 3 + A1203 + 3H,O. The broad endotherm convering the range 700875°C may be attributed to the removal of the hydroxyl groups held on the surface of A1203. The DTA of preparation Z shows four endothermic effects; the lowest is ill-developed and is due to the removal of physisorbed water, the second (at 340°C) stands for specifically adsorbed water. The endotherm located at 460°C is most probably attributed to the formation of the oxide phase as evidenced from the XRD. The endotherm with the minimum located at 960°C was previously ascribed [20] to the complex nature of the system ZnO-H,O. The DTA patterns of M and C preparations are completely similar. They both exhibit two endotherms at 330°C and 39wOo”C. These endothetms are attributed to the removal of physisorbed water and chemi-
I-
Q
T°C Fig. 1. DTA tracings of the original preparations.
321
Th. El-Nabarawy et al. /Materials Letters 24 (1995) 319-325
sorbed water respectively. The DTA curves of M and C showed exothermic effects with the maxima located at 640°C. The 640°C exotherm is due to the formation of zinc aluminate (ZnAl,O,) . XRD was used to explain the reason for this. The high temperature endotherms exhibited by M and C at 75%900°C may be attributed to the continuous removal of OH radicals held to the surface of the solid. 3.2. X-ray dzffractiorz XRD patterns of some selected samples are shown in Fig. 2. For the sample M 500, the lines shown indicate poorly crystalline ZnO. It is also evident from Fig. 2 that calcination at 700°C gave ZnAl,O, spinel. This is true for mechanically mixed oxides and coprecipitated oxides. Further rise of the calcination temperature to 1000°C was associated with an increase in crystallinity. 3.3. Textural properties The adsorption of nitrogen at 77 K on all the thermal products was followed until near saturation (Pl PO = 1.O), then the desorption was followed until the closure of the hysteresis loop. Representative adsorp-
0.2
0.4
0.6 PI
Fig. 3. Representative
0.0
PO
adsorption-desorption
of nitrogen at 77 K.
tion-desorption isotherms are shown in Fig. 3. The isotherms are typically type II of (BDDT) classification [ 211 or type IV of the classification of Sing et al. [ 221. The adsorption data were interperated by the application of the conventional BET equation [ 231 for the determination of the surface area S,, (m’/g) and by the application of the DR equation [24] for the calculation of the micropore volume V, (ml/g). The total pore volume V, (the volume of liquid nitrogen ml/g) adsorbed near saturation was read from the adsorption isotherm. The mean pore radius r (nm) for each sample was calculated from the relationship: r=
2v, x lo3 s BET
62
Fig. 2. XRD of some selected samples.
66
.
Table 1 summarizes the textural properties investigated samples. Inspection of Table 1 that A500,2500, M500 and C500 measure high areas. Calcination above 500°C decreases the
of the reveals surface surface
322
Th. El-Nabarawy et al. /Materials
Letters 24 (1995) 319-325
Table 1 Textural properties of the investigated catalysts Sample
Surface area
V, (ml/g)
VII (ml/g)
A A300 A500 A700 Al000
120 164 205 180 88
0.084 0.123 0.185 0.216 0.158
0.017 0.037 0.081 0.060 0.016
20 30 44 28 10
1.4 1.5 1.8 2.4 3.6
Z 2300 2500 2700 ZlOOO
24 30 40 22 12
0.034 0.045 0.108 0.088 0.058
0.061 0.126 0.032 0.007
18 28 30 8
2.8 3.0 5.4 8.0 9.6
M M300 M500 M700 Ml000
76 32 140 33 30
0.084 0.061 0.266 0.152 0.180
0.036 0.007 0.114 0.034 0.034
43 11 43 23 19
2.2 3.8 5.8 9.2 12.0
C c300 c500 c700 Cl000
75 40 148 77 27
0.090 0.064 0.370 0.366 0.176
0.046 0.003 0.207 0.066 0.025
52 5 56 18 14
2.4 3.2 5.0 9.5 13.0
area, however the decrease is more pronounced for mixed oxide catalysts. At one and the same calcination temperature, A measures the highest surface area and Z measures the lowest surface area. The mean pore radius gradually increases upon the rise of calcination temperature, this is more pronounced in mixed oxide samples. The formation of the spine1 ZnAlzOd may explain the significant changes in the textural properties. With the exception of the sample 21000, the micropore volume represents a fraction of the total porosity of the sample. Column 5 of Table 1 indicates that the contribution of microporosity expressed as V,/ VT depends on the chemical composition and calcination temperature. Thus for pure oxides (A or Z), VO/ VT increases upon calcination at 500°C and thereafter decreases with increasing calcination temperature. The effect of thermal treatment on the porosity of mixed oxides is different. For the mixed oxides V,/ VT is minimum at 300 and maximum at 500°C. In this temperature range, dehydration and dehydroxylation take place as evidenced from the DTA curves of mixed oxides.
VCJV, (%)
P
(m’/g)
(nm)
3.4. Conversion of 2-propanol
The catalytic activity measurements were followed on some selected samples, based on preliminary experiments that indicated their high performance. These samples are namely A500, M500 and C500. The heterogeneous decomposition of 2-propanol has been interpreted in terms of various mechanisms. The two main paths of conversion are: CH,COCH, + Hz dehydrogenation (DHG)
CH,CH=CH, + H,O dehydration (DHD) The products are either propene (due to acid centers) or acetone (due to redox or basic centers [ 25 I ) . For alumina, propene was the only conversion product indicating that alumina is a specific dehydration
Th. El-Nabarawy et al. /Materials
Letters 24 (1995) 31%325
323
Y
c
X 270
B-O.6
1.4
1.6
1.8
2.0
1000 T Fig. 5. Application of the Arrhenius equation to the DHD of 2propanol on C500.
log l/F Fig. 4. Representative linear Kiperman plots.
catalyst exhibiting 100% DHD selectivity. Pure zinc oxide exhibited dehydrogenation of 2-propanol at 28032O”C,its dehydrating efficiency was less than its dehydrogenating activity and started only at 300°C [ 261. Dehydration and dehydrogenation of 2-propanol are performed on mechanically mixed and co-precipitated oxides. Some qualitative results could be reached and are summarized as follows: (i) DHD and DHG decrease upon decreasing the conversion temperature from 340 to 260°C. (ii) ZnO-A1,03 exhibited very interesting behaviour which is different from the behaviour exhibited by other mixed oxide systems. Thus DHD and DHG and accordingly the total catalytic activities increase with the increase of the calcination temperature. This indicates that zinc aluminate is an active catalyst for the conversion of 2-propanol and possibly of other alcohols. (iii.) The selectivity of the ZnGA1203 catalyst depends on the calcination temperature and the method of preparation. M500 and C500 exhibited the same % DHD while % DHG of the former is higher than that of the latter. On the other hand, M700 and C700 show the same % DHG while C700 is more active in dehydration of 2-propanol than M700. It was also found that Cl000 is more active in the dehydration
of 2-propanol than MIOOO,the opposite is shown when dehydrogenation of this alcohol is considered. The dehydration of 2-propanol was followed at different flow rates of the nitrogen carrier gas to determine the kinetics of this reaction. It seems that the kinetics of alcohol dehydration depends on the catalyst used. Thus the dehydration of 2-propanol on A500 follows the first-order Bassett-Habgood equation [ 271. However this equation does not fit the results of M500 and C500 by the data of 2-propanol dehydration on mixed oxide. The application of the Kiperman equation [ 281 to this data proved to be satisfactory. Thus, the plot of log X,where x is the fraction of 2-propanol dehydrated, against l/F, where F is the flow rate of the nitrogen carrier gas (ml mm’), gave straight lines with the slope determining the order of the reaction. Representative linear Kiperman plots are shown in Fig. 4. The dehydration of 2-propanol at different temperatures also allowed the determination of the activation energy of the reaction, through the application of the Arrhenius equation. Representative Arrhenius plot is shown in Fig. 5; an activation energy of 24.6 kJ mol- ’ was calculated for dehydration of 2-propanol on C500. Two further issues should now be pointed out: (i) the dehydration and dehydrogenation of 2-propanol on ZnO/A1203 systems do not depend on the textural properties of the system. Thus although Ml000 and Cl000 measure very low surface areas, they exhibit
324
Th. El-Nabarawy et al. /Materials Letters 24 (1995) 319-325
Table 2 Cracking of cumene at different temperatures Sample
RT
on some selected catalyst
Propylene
Benzene
Toluene
500 450 400
64.26 52.15 21.56
55.64 38.51 19.41
_
MS00
500 450 400
43.99 33.22 21.90
33.13 24.18 14.66
3.54 3.29
c500
500 450 410 360
58.25 45.39 35.20 16.46
48.25 36.22 22.80 13.24
1.52 -
(“C) A500
Cumene
Ethyl toluene
30.12 41.32 62.19
-
4.90 3.14
26.31 40.23 56.19
6.08 -
12.41 16.88 21.31
2.22 3.23
5.83 3.72 _
18.01 36.38 51.02 69.04
2.13 -
6.19 11.90 13.18 14.50
2.00 2.60
Ethyl benzene
high conversion activities compared with M500 and C500. This indicates that the conversion of 2-propanol and possibly of other alcohols is a structural insensitive reaction. (ii) Many mixed oxide systems rapidly lose most of their conversion activity for alcohol when their calcination temperatures exceed 700-800°C. ZnOA1,03 catalyst do not exhibit this trend, and show high activity even upon calcination at 1000°C. This is very interesting in industry and may refer to the potential activity of this system in catalytic reactions proceeding at high temperatures. 3.5. Cracking of cymene The cracking of cumene is a model reaction that determines the performance of the catalyst in other cracking processes, meanwhile the activity of the catalyst in cumene cracking is frequently used as a measure of the strong surface acidity of this catalyst. Some investigators reported that propylene and benzene are the only products of cumene cracking [ 121 while others reported additional products due to side reactions [ 291. In the present investigation three catalysts were selected, namely A500, M500 and C500. The flow rate of nitrogen carrier gas was kept constant at 30 ml mm’, the temperature of cracking was changed between 360°C and 500°C. Table 2 summarizes the data obtained for cumene cracking on the selected catalysts. Inspection of Table 2 reveals that: (i) propylene and benzene are the main products of cumene cracking on
Cymene
Diisopropyl benzene
5.62 5.93 9.65
A500, however, cymene appeared as a minor product and increased with the decrease of the reaction temperature. (ii) For M500 and C500, the number of cracking products increased with the increase of the reaction temperature. Below 45O”C, the main products were propylene and benzene together with cymene as a minor product. At temperatures higher than 450°C toluene, ethyl benzene, ethyltoluene and diisopropyl benzene were detected for M500 and C500. The results for cumene cracking obtained in the present investigation, although limited may explain the discrepancies in the literature. It seems that cracking of cumene depends on the chemical composition of the catalyst and on the temperature of cracking. The differences between tbe results of cumene cracking on M500 and C500 (Table 2) refer to the importance of the method of preparation of mixed oxide catalysts in determining their catalytic performance.
References [ 11 K. Tanabe, Solid acids and bases (Academic Press, New York, 1910). [ 21 K. Tanabe, in: Catalysis, science and technology, Vol. 2, eds. J.R. Anderson and M. Boudart (Springer, Berlin, 198 1) p. 23 1. [3] G. Connell and J.A. Dumesic, J. Catal. 105 ( 1981) 285. [4] J.B. Uytter, L.G. Christner and W.K. Hall, J. Phys. Chem. 69 (1965) 3463. [5] A.M. Youssef and M.R. Mostafa, Bull. Chem. Sot. France (1981) 111. [6] A.M. Youssef, J. Hokkaido Inst. Catal. 28 ( 1980) 89.
325
Th. El-Nabarawy et al. /Materials Letters 24 (1995) 319-325 [7] B.S. Girgis, A.M. Youssef and M.N. Alaya, Bull. Chem. Sot. France 11-12 (1976) 1722. [ 81 V.A. Dzis’ko, Proc. 3rd. Inter. Congr. Catalysis, Amesterdam, Vol. 1 (1964) 1422. [9] R.Sh. M&hail, A.M. Youssef and Th. El-Nabarawy, J. Colloid Interface Sci. 70 (1979) 467. [lo] N.A. Youssef and A.M. Youssef, Bull. Sot. Chim. France 128 (1991) 864. [ 1 l] B.S. Girgis, A.M. Youssef and M.N. Alaya, Bull. Sot. Chim. France 9-10 (1977) 880. [ 121 B.S. Girgis, A.M. Youssef and M.N. Alaya, Surface Technol. 10 (1980) 105. [ 131 H. Vinek, 2. Physik. Chem. (N.F.) 120 (1980) 119. [ 141 H. Noller and G. Ritter, J. Chem. Sot. Faraday Trans. I 80 (1984) 275. [ 151 G. Ritter, H. Noller and J.A. Lercher, J. Chem. Sot. Faraday Trans. 178 (1982) 2239. [16] A.M. Youssef, A.M. El-Wakil, K. Tollan and Th. ElNabarawy, Afinidad XLVII ( 1990) 336. [ 171 A.M. Youssef, G.A. Fagal and Th. El-Nabarawy, Afinidad IL No. 440 ( 1992) 247. [ 181 A.M. Youssef and M.R. Mostafa, Bull. Sot. Chem. France (1988) 807.
[19] A.M. Youssef (1972)
and M.M. Selim, Fertilizer
Technology
12
237.
[ 201 W. Etel, Silicate science, Vol. 4 (Academic Press, New York, 1966) p. 617.
[ 211 S. Brunauer, L.S. Deming, W.E. Deming and E. Teller, J. Am. Chem. Sot. 62 (1940)
1723.
[22] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Noscou, R.A. Pierotti, J. Rouguerol and T. Siemieniewska, Pure Appl. Chem. 57 (1985) 603. [23] S. Brunauer, P.H. Emmett and E.E. Teller, J. Am. Chem. Sot. 60 (1938)
309.
[24] M.M. Dubinin, D. Zavarina Khim. 12 (1947) 1351.
and L.V. Radushkevich,
[25] M.M. El-Jamal, M. Forissier and A. Auroux, Faraday Trans. 184 (1988) 3169. [26] A.M. Youssef,
E. Elshereafy,
M. Khidir
Zh. Fiz.
J. Chem. Sot. and E. El-Said,
Silicates Industriels 7-8 ( 1994), in press. [27] D.W. Bassett and H.W. Habgood, 796.
J. Phys. Chem. 64 (1960)
[ 281 S.L. Kiperman, Introduction to the kinetics of heterogeneous catalysis
(Moscow,
Nauka, 1964) p. 543.
[29] D. Best and B.W. Wojciechowski,
J. Catalysis, 74 (1977)
11.
