Highly active catalyst of NiO—ZrO2 modified with H2SO4 for ethylene dimerization

~ ELSEVIER

APPLIED CATALYSS I AG : ENERAL

Applied Catalysis A: General 128 (1995) 127-141

Highly active catalyst of NiO-ZrO2 modified with HzSO 4 for ethylene dimerization Jong R a c k S o h n a,,, H a e W o n K i m a, M a n Y o u n g Park a, Eun Hee Park a, Jong Taik Kim a, Sang Eun Park b ~Department of Industrial Chemistry, Engineering College, Kyungpook National University, Taegu 702-701, South Korea b Korea Research Institute of Chemical Technology, Taejon, 305-606, South Korea

Received 5 January 1994; revised 12 January 1995; accepted 20 February 1995

Abstract A series of catalysts, NiO-ZrO2/SO 2-, for ethylene dimerization were prepared by coprecipitation from a solution of a nickel chloride-zirconium oxychloride mixture followed by modifying with H2SO4. On the basis of the results obtained from X-ray diffraction and differential thermal analysis, the addition of nickel oxide to ZrO2 or modification with H2SO 4 shifted the transition of ZrO2 from amorphous to a tetragonal phase at higher temperatures due to the interaction between nickel oxide (or sulfate ion) and ZrO 2. Infrared spectra of the catalyst modified with [-[2804 showed bidentate sulfate ion coordinated to Zr 4+ or Ni 2+. NiO-ZrO2 without sulfate ion was inactive for the ethylene dimerization, but NiO-ZrO2/SO4 was found to be very active even at room temperature. The high catalytic activity of NiO-ZrO2/SO42- was closely correlated with the increase of acid strength by the inductive effect of sulfate ion. The decrease of catalytic activity above 450°C of evacuation temperature was due to the structural change from amorphous phase to crystalline and the sintering followed by the decrease of surface area. Keywords." Acid strength; Coprecipitation; Differential thermal analysis; Ethene dimerization; Nickel oxide-

zirconium oxide; Sulphate modification; X-ray diffraction

1. Introduction Nickel oxide on silica or silica-alumina is an effective catalyst for alkene dimerization and isomerization as well as for hydrocracking, hydrogenation and methanation, following reduction of the nickel component [1-10]. One of the remarkable features of this catalyst system is its activity in the relation to a series * Corresponding author. 0926-860X/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDIO926-860X (95) 00057-7

128

ZR. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

of n-alkenes. In contrast to usual acid-type catalysts, the nickel oxide on silica or silica-alumina shows a higher activity for a lower alkene dimerization, particularly for ethylene [ 1,2,5-7,11 ]. The catalyst is also active for the isomerization of nbutenes, the mechanism of which has been proved to be of a proton donor-acceptor type [ 12]. It was reported that the dimerization activities of such catalysts are related to the acid properties of the surface and low valent nickel ions. In fact, nickel oxide which is active for C2Ha-C2D4 equilibration acquires an activity for ethylene dimerization upon addition of nickel sulfate which is known to be an acid [ 13 ]. A transition metal can also be supported on zeolite in the form of a cation or a finely dispersed metal. Several papers have treated ethylene dimerization on transition-metal cation exchanged zeolite [ 14-16]. In previous papers [5,6], it was reported that the active sites on a NiO--SiO2 catalyst for ethylene dimerization are classified into two types, montmorillonite and antigorite, the optimum activation temperature for them being around 100°C and 600°C, respectively. It was also previously reported that ZrO2 modified with H2SO4 exhibits superacidic properties [ 17-20]. As an extension of the study on the ethylene dimerization, we tried to prepare other catalyst systems by combining nickel hydroxide to give low valent nickel after decomposition with ZrO2 modified with H2SO4 which is known to be an acid [ 19]. In this paper, characterization of NiOZrO2/SO] and catalytic activity for ethylene dimerization are reported.

2. Experimental 2.1. Catalysts The coprecipitate of Ni (OH)2-Zr(OH)4 was obtained by adding aqueous ammonia slowly into a mixed aqueous solution of nickel chloride and zirconium oxychloride at room temperature while stirring until the pH of the mother liquor reached about 8. The ratio of nickel chloride to zirconium oxychloride was varied. The coprecipitate thus obtained was washed thoroughly with distilled water until chloride ion was no longer detected, and then dried at room temperature. The dried precipitate was powdered below 100 mesh and then the catalyst was modified with sulfate ion by pouring 30 ml of 0.5 M H2SO4 onto 2 g of the powdered sample on a filter paper and drying the sample in air. The dry solid powder was used as catalyst after decomposing at different evacuation temperatures for 1.5 h. This series of catalysts are denoted by the mol percentage of nickel oxide. For example, 25-NiOZrO2 means the catalyst having the nickel oxide of 25 mol.-% and 25-NiO-ZrO2/ SO 2 means the 25-NiO-ZrO2 modified with H 2 S O 4.

2.2. Procedure The catalytic activity for ethylene dimerization was determined at 20°C by a conventional static system following pressure change from an initial pressure of

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

129

288 Torr ( 1 T o r t = 133.3 Pa). A fresh catalyst sample of 0.2 g was used for every run after evacuation to 10-4 Tort at different temperatures for 1.5 h and the catalytic activity was calculated as the amount of ethylene consumed during the initial 5 min. The isomerization of 1-butene was carried out at 20°C in a closed circulating system. Reaction products were analyzed by gas chromatography equipped with a VZ-7 column at room temperature. The specific surface area was determined by applying the BET method to the adsorption of nitrogen at - 196°C. The acid strength of the catalysts was measured qualitatively after the pretreatments using a series of Hammett indicators [ 21 ]. The catalysts were pretreated in glass tubes by the same procedure as for the reactions. They were cooled to room temperature and filled with dry nitrogen. The color changes of a series of indicators were observed for each catalyst by the spot test under dry nitrogen. The infrared spectra were recorded using a Mattson model GL6030 FTIR spectrometer. Usually 10 mg of catalyst was mixed with 100 mg of KBr and pressed into a disc (600 kg/cm2). However, for IR studies of pyridine and carbon monoxide adsorbed on catalyst, self-supported discs of about 10 mg cm -2 were used. X-ray diffractograms of catalysts were taken using a Jeol Model JDX-88 X-ray diffractometer with a copper target and nickel filter at 30 kV and 800 cps. The sulfur content remaining on the catalyst after the evacuation at 400°C was determined with a Rigaku X-ray fluorescent spectrometer. The content of sulfate ion in the catalysts was estimated to be about 4-6 wt.-%. X-ray photoelectron spectroscopy was performed on a Hitachi 507 photoelectron spectrometer equipped with a cylindrical mirror analyzer using a A1 anode (9 kV, 50 mA). Binding energies were referenced to the C ls level at 285.0 eV. The thermal analysis was carried out with a PL-STA 1500 HF analyzer in air. The heating rate was 10°C/rain and for each experiment 10-15 mg of sample was used. Temperature-programmed ammonia desorption experiments were performed as follows. Before introducing ammonia, the catalyst was activated at 400°C for 1 h in helium flowing at 50 cm 3 per rain. After adsorption of ammonia, the helium flow was continued for an additional 30 min to remove the excess ammonia. The reactor temperature was raised at a rate of 10°C rain 1 to 500°C.

3. Results and discussion

3.1. Infrared spectra The infrared spectra of the 25-NiO-ZrO2/SOl- (KBr disc) are given in Fig. 1. The catalyst showed infrared absorption bands 1230-1220, 1140-1130, 1060-1050 and 990 c m - ~ which are assigned to the bidentate sulfate ion coordinated to Zr 4 + or Ni 2+ as follows [22] :

130

J.R. Sohn et al./Applied Catalysis A: General 128 (1995) 127-14l

a

o

1140 1060 I

1400

I

1200 1000 Wavenumber, cm

Fig. 1. Infrared spectra of (a) 25-NiO-ZrO2, (b) 25-NiO-ZrOJSO~ NiO-ZrO2/SO] evacuated at 400°C.

-M-0

-1

800

evacuated at room temperature, (c) 25-

\/0 S

0/\0

Infrared bands of other samples were similar to those of 25-NiO-ZrO2/SO 2 . The u SO spectra from the adsorbed sulfate in the v~ and v3 frequency region (9001400 c m - 1) support a species of reduced C2, symmetry with four bands arising from v j and splitting of the triply degenerate u3 vibration [23]. Other catalysts modified with sulfate ion also showed similar infrared absorption bands. Even after evacuation at 400°C for 1.5 h, strong absorption bands of sulfate ion are left, indicating a very strong interaction between sulfate ion and the cations. For the sample modified with H2SO 4 followed by evacuation at 400°C, based on infrared results, the sulfur is considered to reach the highest oxidation state, S6+ of SO]-. To obtain further information on the oxidation state of sulfur, an X-ray photoelectron spectroscopic investigation was performed. As expected, the sample modified with H2SO4 gave the signal attributed to S6+ [24], suggesting the oxidation state of sulfur as SO 2- to be the main species. The S 2p binding energy of S 6+ was 168.8 eV, referenced to the C ls level at 285.0 eV.

3.2. Structure of catalyst The crystalline structure of catalysts calcined in air at different temperatures for 1.0 h was examined. ZrO2 was amorphous to X-ray diffraction up to 400°C with a

J.R. Sohn et al. /Applied Catalysis A: General I28 (1995) 127-141

• AA _XAU

•0

•ll

131

• I000~

0

•

900"C 0

0

800"C

700"C o

~

550"C --------'.-----500"C

I

l

I

I

!

20

30

40

50

60

20 Fig. 2. X-raydiffractionpatternsof 25-NiO-ZrO2 calcinedat differenttemperaturesfor I h: (©) tetragonalphase

ZrO2: (Q) monoclinicphase ZrO~.

tetragonal phase at 450°C, with a two-phase mixture of the tetragonal and monoclinic forms at 600-700°C, and a monoclinic phase at 800°C. Three crystal structures of ZrOz, tetragonal, monoclinic and cubic phases have been reported [25,26]. However, in the case of NiO-ZrO2 catalysts the crystalline structures of samples were different from that of pure ZrO2. For the 25-NiO-ZrO2, as shown in Fig. 2, ZrO2 was amorphous up to 500°C. In other words, the transition temperature from amorphous to tetragonal phase was higher by 100°C than that of pure ZrO2. X-ray diffraction data indicated a tetragonal phase of ZrO2 at 550-700°C, a two-phase mixture of the tetragonal and monoclinic ZrO2 forms at 800-900°C, and a monoclinic phase at 1000°C. It is assumed that the strong interaction between nickel oxide and ZrOz hinders the transition of ZrO2 from amorphous to tetragonal. The presence of nickel oxide strongly influences the development of textural properties with temperature in comparison with pure ZrO2. Moreover, as shown in Fig. 3, for the sample of 25-NiO-ZrO2/SO ] the transition temperature from amorphous to tetragonal phase was higher by 200°C than that of pure ZrO2. It is assumed that the strong bond formation between sulfate ion and ZrO2 prohibits the transition from amorphous to tetragonal [19]. 25-NiOZ r 0 2 / S O ] - was amorphous to X-ray diffraction up to 600°C, with a tetragonal phase of Zr02 at 650-700°C, a two-phase mixture of the tetragonal and monoclinic

132

J.R. Sohn et al. / Applied Catalysis A: General 128 (1995) 127-141

•

•

•

I~0~

•

V

~

,

~

~

~

~0~

o ~

~

~

.

7~

~

6 5 0 " C

, ~ . . - ~ " " ~ - ~ -

_

,

•

--

I

I

l

I

I

20

30

40

50

60

600"C

20

Fig. 3. X-ray diffraction patterns of 25-NiO-ZrO2/SO4z calcined at different temperatures for 1 h: (©) tetragonal phase ZrO2; ( • ) monoclinic phase ZrOz.

forms at 800-900°C, a monoclinic phase at 1000°C. The cubic phase of nickel oxide was not observed up to 25 mol-% of NiO, indicating a good dispersion of nickel oxide on the surface of ZrO2 due to the interaction between the two. However, for the sample that has a higher NiO content than 30 mol-%, a cubic phase of nickel oxide was observed in the samples calcined above 300°C.

3.3. Thermal analysis As observed from the X-ray diffraction patterns, the structures of the catalysts are different from each other depending on the calcination temperature. To examine the thermal properties of precursors of catalysts more clearly, their thermal analyses were carried out and are illustrated in Fig. 4. For pure ZrO2, the DTA curve shows an endothermic peak in the temperature range 30-180°C due to water elimination, and a sharp exothermic peak at 430-470°C due to the ZrO 2 crystallization [27 ]. In the case of ZrO2/SO2-, sulfate ion influences the phase transition of ZrO2 from amorphous to tetragonal. As Fig. 4 shows, the exothermic peak due to the phase transition appears at about 450°C for pure ZrO2, while for ZrO2/SO 2- samples it is shifted to higher temperature, about 640°C. Since nickel oxide also influences the phase transition of ZrO2, for 25-NiO-ZrO 2 the exothermic peak due to the phase

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

0

200

400

600

133

800

Temperature, °C Fig. 4. DTAcurvesof catalystprecursors. transition appears at about 550°C. These results are in good agreement with those of X-ray results described above. The shift increases and the shape of the peak becomes broadened with increasing nickel oxide content. Consequently, for 34NiO-ZrO2 the exothennic peak appears at 500-570°C, while for 34-NiO-ZrOz/ S O ] - it appears at 550-710°C. In the case of NiO-ZrO2 catalysts, an additional endothermic peak appears at 230-330°C due to the decomposition of Ni (OH)2. The decomposition of Ni (OH)z is known to begin at 230°C [28]. As shown in Fig. 4, in this work pure Ni(OH)2 decomposes at 230-330°C. In the case of samples modified with H2804, the endothermic peaks which appear at 700-740°C for Z r O J S O ] - , at 710-760°C for 25NiO-ZrO2/SO 2-, and at 760-840°C for 34-NiO-ZrO2/SO 2 are responsible for evolution of SO3 due to the decomposition of sulfate ion bonded on the surface of the catalysts. As shown in Fig. 4, the decomposition temperature of sulfate ion increases with increasing nickel oxide content. The decomposition temperature of pure nickel sulfate is reported to be 848°C [29]. These results for the modified samples are very similar to that for iron oxide treated with HzSO 4 [ 30]. It is relevant that the strong interaction between nickel oxide (or sulfate ion) and zirconia delays

134

J.R. Sohn et al./Applied Catalysis A: General 128 (1995) 127-141

Table 1 Specific surface areas of several samples ( m 2 / g ) Sample

Surface area

ZrO2

94

Sample

Surface area

ZrOJSO~-

137

5-NiO-ZrO2

132

5-NiO-ZrOJSO]

182

25-NiO-ZrOz

186

25-NiO-ZrO2/SO~

230

34-NiO-ZrO2

139

34-NiO-ZrO2/SO]

184

60-NiO-ZrO~

96

60-NiO-ZrO2/SO]

140

NiO

39

NiO/SO]

64

the transition of Z r O 2 from amorphous to tetragonal. A similar result has been observed by Sohn et al. for silica and chromium oxide additions [31,32].

3.4. Surface properties of catalysts It is necessary to examine the effect of nickel oxide and sulfate ion on the surface properties of the catalysts, that is, specific surface area, acid strength, and nature of acid centers (BrCnsted or Lewis type). The specific surface areas of some samples after evacuation at 400°C for 1.5 h are listed in Table 1. The surface area increases gradually upon addition of NiO to ZrO2, reaching a maximum at 25-NiO-ZrO2. As shown in Table 1, however, the surface area of modified samples increased compared with that of unmodified catalysts. It seems likely that the interaction between nickel oxide (or sulfate ion) and ZrO2 protects catalysts from sintering [ 19,32]. The acid strength of the catalysts was examined by a color change method, using a Hammett indicator [19] in sulphuryl chloride. Since it was very difficult to observe the color of indicators adsorbed on catalysts of high nickel oxide content, a low percentage of nickel content (5 tool-%) was used in this experiment. However, in the case of catalyst having high content of NiO (34 mol-%), a sample diluted with inert material, SiO2 was used. The results are listed in Table 2. In this table, ( + ) indicates that the color of the base form was changed to that of the conjugated acid form. ZrO2 evacuated at 400°C for 1 h has an acid strength Ho ~< + 1.5, while 5-NiO-ZrO2 was estimated to have a Ho ~< - 8.2, indicating the Table 2 Acid strength of several samples Hammett indicator

Benzeneazodiphenylamine Dicinnamalacetone Benzalacetophenone Antraquinone Nitrobenzene 2,4-Dinitro-fluorobenzene

pK~ value of ZrO2 5 - N i O indicator ZrO2

34-NIO-

+ 1.5

ZrO2/SO4

ZrO2

5-NiO-

34-NiO-

ZrO2/SO4

ZrO,/SO~

+

+

+

+

+

+

- 3.0

-

+

+

+

+

+

- 5.6

-

+

+

+

+

+

- 8.2

-

+

+

+

+

+

- 12.4

-

-

-

+

+

+

- 14.5

-

-

-

+

+

+

135

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

..=

I

I

I

I

I

I

I

I

I

I

100

200

300

400

0

100

200

300

400

500

Temperature, "(2 Fig. 5. Temperature-programmed desorption profiles of ammonia from (a) ZrO2, (b) ZrO2/SO42 , (c) 5-NiOZrO2, (d) 5-NiO-ZrO2/SO 2- , (e) 25-NiO-ZrO2, and (f) 25-NiO-ZrO2/SO 2-.

formation of new acid sites stronger than those of single oxide components. Regardless of NiO content, 8-NiO-ZrO2 and 34-NiO-ZrO2 were also estimated to have a Ho ~< - 8 . 2 , The acid strength of ZrO2/SO 2-, 5-NiO-ZrO2/SO 2-, 8-NiO-ZrO2/ SO 2- and 34-NiO-ZrO2/SO 2- was also found to be Ho ~< - 14.5. Acids stronger than Ho ~< -11.93, which corresponds to the acid strength of 100% H2SO4, are superacids [33]. Consequently, ZrO2/SO 2- and NiO-ZrO2/SO 2- catalysts would be solid superacids. The superacidic property is attributed to the double bond nature of the S---O in the complex formed by the interaction of ZrO2 with sulfate ion [ 19,20,34]. That is, the acid strength of samples modified with H2SO 4 becomes stronger by the inductive effect of S=O in the complex. Fig. 5 shows the TPD curves of ammonia adsorbed on catalysts as a function of temperature. As seen in Fig. 5, each curve displays a very broad peak, which is indicative of site heterogeneity. The curve was analyzed by appropriate curve fitting and the presence of two or three components was confirmed. Although pure Z r O 2 showed little acidity, the acidity increased upon the addition of nickel oxide to Z r O 2. Moreover, as seen in Fig. 5, the catalysts modified with H z S O 4 exhibited a remarkable increase in acidity and acid strength compared with unmodified catalysts. 25-NiO-ZrO2/SO] showed the highest acidity, explaining that 25-NIO-

136

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

O

a

1700

I

i

I

1600

1500

1400

Wavenumber,

cm

1300 -1

Fig. 6. Infrared spectra of pyridine adsorbed on (a) ZrO2/SO~ and (b) 5-NiO-ZrO2 afterevacuation at 400°C for I h, The gas phase was evacuated at 250°C for I h afteradsorption in (a) and (b).

ZrO2/SO42- exhibits the highest catalytic activity for ethylene dimerization described later. It has been established that BrCnsted and Lewis acid sites are distinguishable by infrared spectra of adsorbed pyridine [35]. Fig. 6 shows the infrared spectra of pyridine adsorbed on ZrO2/SO ]- and 5-NiO-ZrO2/SO42 evacuated at 400°C, after pyridine adsorption at room temperature, followed by desorption at 250°C. Both the pyridinium ion band at 1543 c m - ~ and the coordinated pyridine band at 1450 cm-1 are found with them. For the other samples modified with H2804, similar results were obtained. It is clear that both BrOnsted and Lewis acid sites exist on the surfaces ofZrO2/SO ] and N i O - Z r O J S O 2- evacuated at 400°C. 3.5. Ethylene dimerization over NiO-ZrOJS024 -

Over 25-NiO-ZrO2/SO 2 , ethylene was selectively dimerized to n-butenes. In the composition of the n-butenes analyzed by gas chromatography, 1-butene was found to predominate only initially during the reaction as compared to cis-butene and trans-butene. As shown in Fig. 7, however, the amount of 1-butene decreases with the reaction time, while the amount of 2-butene increases. Therefore, it seems likely that the initially produced 1-butene is also isomerized to 2-butene during the reaction. It was found that the activity decrease is accompanied by a decrease in the extent of isomerization of n-butene produced by the dimerization. The compositions of nbutenes obtained by dimerization runs on a series of catalysts for 30 min are plotted against the catalyst composition of the NiO-ZrO2/SO 2- system in Fig. 8. It is

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

137

I00

80

trans-2

1

60

40

t g

/ ~ q ~ Q zo

i

0

0

- - O "-----------C

I

I

I

30

60

90

1E0

Reaction time, rain Fig. 7. Composition change of produced n-butenes with reaction time, Reaction conditions: catalyst, 25-NiOZrOJSO] (0.2 g) evacuated at 400°C for 1.5 h; reaction temperature, 20°C; ethylene initial pressure, 185 Tort.

90

~ BO

m i

~ 70 60

250

5

150

0

I

ZrO2

20

I

I

40 60 NiO moW.

I

80

©

~50 NiO

Fig. 8. Variations of ethylene dimerization activity, product composition in ethylene dimerization, and specific surface area with the composition of NiO-ZrOJSO 2 system.

138

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

"7

8

T % X 6

ca 4 O °~ .,..a N

'E 2 O

0

Zr02

I

I

I

I

I

20

40

60

80

NiO

Ni0 mol~ Fig. 9. Variation of l-butene isomerization activity with NiO tool-% in NiO--ZrO2/SO4

system.

obvious that the selectivity to 2-butene runs parallel with its dimerization activity. Since the initial product of ethylene dimerization seems to be 1-butene [36], the above result suggests that the isomerization activity runs parallel with the dimerization activity. Thus the isomerization of 1-butene was carried out at 20°C on NiOZrO2/SO 2- in a circulating system. The time course of isomerization was in agreement with the first order rate law. Thus the isomerization activity is represented by the first order rate constant. As shown in Fig. 9, it is confirmed that the isomerization activity also gives a maximum at 25 tool-% of NiO as does the case of Fig. 8 for the dimerization activity. The above results are in conformity with the view that 1-butene is the initial product of ethylene dimerization. The effect of catalyst composition on the catalytic activity was examined, where the catalysts were evacuated at 400°C for 1.5 h. As shown in Fig. 8, the maximum activity is obtained with the catalyst of 25 mol-% of nickel oxide. This is due to the increase of specific surface area and the subsequent increase of active sites. In fact Fig. 8 shows that the specific surface area attained a maximum when the NiO content in the catalyst is 25 mol-%. In Fig. 10, the ethylene dimerization activities of 25-NiO-ZrO2/SO4z - are plotted against the temperature at which the catalyst was evacuated for 1.5 h. It can be seen that the activity appears above 200°C reaching a maximum at 400-450°C. In Fig. 4, the decomposition of coprecipitate was observed to begin at 230°C. The decomposition of nickel hydroxide is known to begin at 230°C [ 28 ]. Therefore, it is very likely that the activation of the catalyst above 200°C is related to the decomposition of the catalyst. Fig. 10 also shows the gradual decrease of activity when the evacuation temperature raises above 450°C. It would be worthwhile to discuss the activity decrease above 450°C. As shown in Figs. 2 and 3, the catalyst showed a nearly amorphous structure up to 450°C, while above 450°C the catalyst structure

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 127-141

Lf~

139

300 2 N

800 d L

1

© 0

loo '~ K ¢0

0 200

J 300

i 400

i 500

i 600

0

Evacuation temperature, °C Fig. 10. Variations of ethylene dimerization activity and specific surface area with evacuation temperature.

changed from amorphous to tetragonal. The activity decrease above 450°C is reasonably explained by the solid phase rearrangement or sintering of the catalyst as demonstrated by an irreversible decrease of specific surface area (Fig. 10). The NiO-SiO2 is known to be a typical solid catalyst for selective ethylene dimerization [5,6]. The NiO-SiO2 (NiO content= 24 mol-%, surface area= 491 mZ/g) gave maximum catalytic activity of 1.31 m m o l / g . 5 min, while 25-NiOZrO2/SO 2 (surface area = 186 m2/g) showed maximum catalytic activity of 2.1 mmol/g. 5 min. Taking into account the surface area differences between them, it can be concluded that NiO-ZrO2/SO 2 exhibits considerably higher catalytic activity for ethylene dimerization than NiO-SiO2. It is remarkable that the samples which were not modified with H2SO 4 were inactive as catalysts for ethylene dimerization, but the sample modified with H2SO 4 exhibited a high catalytic activity. The active site responsible for dimerization is suggested to consist of a low valent nickel ion and an acid as observed in the NiOSiO2 catalyst [6,36]. The term 'low valent nickel' originated from the fact that the NiO-SiO2 catalyst was drastically poisoned by carbon monoxide, since a low valent nickel is favorable to chemisorb carbon monoxide [36]. Elev et al. [ 15] concluded that Ni + ions in NiCaY zeolite are responsible for the catalytic activity, using EPR spectroscopy to identify low-valent nickel species. To obtain further information on the oxidation state of low-valent nickel species, infrared spectroscopy using carbon monoxide as a probe was used. Upon addition 20 Torr of CO to various NiO-ZrO2/SO42- evacuated at 400°C for 1.5 h, two bands appeared at 2116 and 2086 cm 1, which may be assigned to the stretching vibrations of CO bonded to Ni + and Ni °, respectively, by comparison with the spectra of carbon monoxide adsorbed on NiCa-X zeolite [37]. In view of the EPR results by Elev et al. [ 15], these results suggest that Ni + ions are responsible for the low-valent nickel species. In the system of NiO-ZrO2/SO] the catalyst was poisoned by 1 /xmol/g of

140

J.R. Sohn et al. /Applied Catah'sis A: General 128 (1995) 127-141

carbon monoxide for dimerization. The nickel oxide modified with H2SO 4 exhibited some activity for dimerization as shown in Fig. 8, but nickel oxide without sulfate ion was inactive. These results support more firmly the fact that the active site for dimerization is formed by an interaction of a low valent nickel ion with an acid. In fact ZrO2/SO42 alone without nickel ion was inactive for dimerization although it exhibited some activity for polymerization. It is well known that an acid catalyst is effective for the polymerization of alkenes [ 38 ].

4. Conclusions A series of catalysts, NiO-ZrO2/SO42- were prepared by coprecipitation and the following fact is demonstrated in present work. The interaction between nickel oxide (or sulfate ion) and ZrO2 influences the changes in physicochemical properties of the prepared catalysts with calcination temperature. The presence of nickel oxide and sulfate ion delays the phase transition of ZrO2 from amorphous to tetragonal, and the specific surface area of catalysts increases as compared with pure ZrO2. NiO-ZrO2/SO42 is very effective for ethylene dimerization, but NiOZrO2 without sulfate ion does not exhibit any catalytic activity. The high catalytic activity of NiO-ZrO2/SO4 is closely correlated with the increase of the acid strength by the inductive effect of sulfate ion coordinated to Zr 4+ or Ni 2+. The active site for dimerization is formed by an interaction of a low valent nickel ion with an acid, because the sample without either nickel ion or sulfate ion was inactive for dimerization. The decrease of catalytic activity above 450°C can be explained in terms of the transition of the catalyst from amorphous to crystalline and sintering followed by a decrease of surface area.

Acknowledgements This work was supported by the Research Center for Catalytic Technology, the Korea Science and Engineering Foundation.

References [ 1] [2] [3] [4] [5] [6] [7] [8] [9]

J.P. Hogan, R.L. Banks, W.C. Lanning and A. Clark, Ind. Eng. Chem., 47 (1955) 752. H. Uchida and H. Imai, Bull. Chem. Soc. Jpn., 35 (1962) 989. H. Uchida and H. Imai, Bull. Chem. Soc. Jpn., 35 (1962) 995. H. Uchida and H. lmai, Bull. Chem. Soc. Jpn., 38 (1965) 925. J. R. Sohn and A. Ozaki, J. Catal., 59 (1979) 303. J. R. Sohn and A. Ozaki, J. Catal., 61 (1980) 29. G. Wendt, E. Fritsch, R. Schollner and H. Siegel, Z. Anorg. Allg. Chem., 467 (1980) 5 I. R.A. Flinv, O.A. Larson and H. Beuther, Ind. Eng. Chem., 52 (1960) 153. G.F. Bemdt, S.J. Thomson and G.J. Webb, Chem. Soc. Faraday Trans. 1, 79 (1983) 195.

J.R. Sohn et al. /Applied Catalysis A: General 128 (1995) 12~141

141

[ 10] T.V. Herwijnen, H.V. Doesburg and D.V. Jong, J. Catal., 28 (1973) 391. [ 11 ] G. Wendt, D. Hentschel, J. Finster and R. SchoIlner, J. Chem. Soc. Faraday Trans. 1, 79 (1983) 2013. [ 12] A. Ozaki and K. Kimura, J. Catal., 3 (1964) 395. [ 13] K. Maruya and A. Ozaki, Bull. Chem. Soc. Jpn., 46 (1973) 351. [ 14] L. Bonneviot, D. Olivier and M. Che, J. Mol. Catal., 21 (1983) 415. [ 15] I.V. Elev, B.N. Shelimov and V.B. Kazansky, J. Catal., 89 (1984) 470. [ 16] A.K. Ghosh and L.J. Kevan, J. Phys. Chem., 92 (1988) 4439. [ 17] M. Hino and K. Arata, J. Chem. Soc. Chem. Commun., (1985) 112. [ 18] T. Yamaguchi, T. Jin and K. Tanabe, J. Phys. Chem., 90 (1986) 3148. [ 19] J.R. Sohn and H.W. Kim, J. Mol. Catal., 52 (1989) 361. [20] J.R. Sohn, H.W. Kim and J.T. Kim, J. Mol. Catal., 41 (1987) 375. [21 ] L.P. Hammett and A.J. Deyrup, J. Am. Chem. Soc., 54 (1932) 2721. [ 22 ] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, 3rd ed., Wiley, New York, 1978, p. 241. [23] R.L. Parfitt and R.St.C. Smart, J. Chem. Soc. Faraday Trans. 1, 73 (1977) 796. [24] Y. Okamoto, M. Oh-Hara, A. Maezawa, T. Imanaka and S. Teranishi, J. Phys. Chem., 90 (1986) 2396. [25] M.J. Torralvo, M.A. Alario and J. Soria, J. Catal., 86 (1984) 473. [26] A. Clearfield, Inorg. Chem., 3 (1964) 146. [27] J. Livage, K. Doi and C. Mazieres, J. Am. Ceram. Soc., 51 (1968) 349. [28] G.F. Hiitig and A. Pater, Z. Anorg. Chem., 189 (1930) 183. [29] R.C. Weast, CRC Handbook of Chemistry and Physics, 65th ed., CRC Press, Boca Raton, Florida, 1984. [30] T. Jin, M. Machida and K. Tanabe, Inorg. Chem., 23 (1984) 4396. [31 ] J.R. Sohn and H.J. Jang, J. Mol. Catal., 64 ( 1991 ) 349. [32] J.R. Sohn and S.G. Rhu, Langmuir, 9 (1993) 126. [33] F.G.A. Olah, G.K.S. Prakash and J. Sommer, Science (Washington), 206 (1979) 13. [34] J.R. Sohn, Y.I. Pae, H.J. Jang and M.Y. Park, J. Catal., 127 ( 1991 ) 449. [35] E.P. Parry, J. Catal., 2 (1963) 371. [36] K. Kimura, H. A-I and A. Ozaki, J. Catal., 18 (1970) 271. [37] M. Kermarec, D. Olivier, M. Richard and M. Che, J. Phys. Chem., 86 (1982) 2818. [38] K. Tanabe, Solid Acids and Bases, Kodansa, Tokyo, 1970, p. 55-106.