React. Kinet. Catal. Lets Voi. 27, No. 1, 157~161 (1985)

GAS PHASE CONVERSION OF CHLOROBENZENE OVER SUPPORTED RHODIUM CATALYSTS

B. Coq, G. Ferrat and F. Figueras

Laboratoire de Chimie Organique Physique et Cin6tique Chimique Appliqu6e

E.R.A. 954 du C.N.I~.S.

Eeole Nationaie Sup6rieure de Chimie de Montpetlier 8, rue Ec01e Normale - - 3~t075 M0ntpellier Cedex, France

t~eceived April 25, 1984: Accepted May 7, 198~

The catalytic conversion of chlorobenzene has been investigated over I%h/ALOa and l~h/SiOe with metal dispersities varying from 7 to 80%. The rates of hydrogenation to ehlorocyelohexane and hydrodechlorination to benzene decrease when the dispersity increases, therefore, both reactions appear as structure sensitive. This behavior is interpreted by a weak interaction be- tween the aromatic ring and the metal surface. At similar particle sizes the activity is the same on Rh/A12Oa and Rh/SiO~, however, the selectivity for dechlorination exhibits a support effect.

KOHBepcH~ xJ~op6eH3oaa 6bl~a uccae~oBaHa Ha KaTannaaTopax Rh/AI203 H Rh/SiO2 c ~HcnepcHOCT~m MeTanaa OW 7 ~0 80 %. CKopocTb rHJIpoBaHun ~o xnop- t~u~(.nore~(caHa ~ rH~po~exaopHpoBaHg~ ~o 6eHaona yMeHbIIlaeTcfl C yBeJIHqe- HHeM ~IHcHepcHOCTH. TaKHM o6paaoM, o6e peaK~HH oKaaaJIHch YyBCTBHTeJ]B- HI)IMH K CTpyKType. ~TO goBe/leHI4e O~%flCHfleTC~I cJla6bIM B3aHMo2iefICTBI4eM Me)K- /Iy ap0MaT~qeCKHM KOJ'IbIIOM H /IoBepXHOCTLm MeTaJ~a. l-lpH 0/IgHaKOBBIX pa3- Mepax qacwm~ aKTHBH0CTb ;Ing Rh/AI203 u Rh/SiO 20;IHHaKoBbI, a ceneKTUB- HOCTb ~ex.nopnpoBaun~ rIORqI4H~IeTC~ Bnn~HrI~ HOCHTe:I~.

I N T R O D U C T I O N

The hydrogenolysis o f carbon-chlorine bonds is an impor t an t reaction in the manufac tu re of pesticides. This react ion is ca ta lyzed b y t ransi t ion metals. I n recent work on the cata lyt ic propert ies of rhodium a clear influence of par- ticle size and suppor t was repor ted on C - - C bond cleavage [1, 2] and sulfur re- sistance [3]. The mechanism of sulfur poisoning b y th iophene implies the hydro- genolysis of C - - S bonds. The common behavior of these two reactions suggested t o compare the results with those of C--C1 hydrogenolysis . This react ion has been scarcely invest igated and most of the results appear in the pa t en t liter- a ture. Previous works suggest t h a t hydrodechlor ina t ion could be a s t ructure :sensitive react ion: the select ivi ty depends on poisoning by sulfur [4], on alloying [5] and on the suppor t [6]. Much of this previous work has been carried ou t in the liquid phase, using complex molecules. I n order to check the influence

157

COQ et M.: CONVERSION OF CHLOROBENZENE

of the surface s t ructure of the metal, ~ simple model was investigated in the gas phase to avoid the competitive adsorption of the solvent.

E X P E R I M E N T A L

1) Cat~ly~t~'." The preparation and characterization of the catalysts have been described in detail previously [1, 2, 3, 7]. Well dispersed rhodium catalysts are obtained by using ion-exhange. The final dispersity is controlled by the temperature of reduction and the partial pressure of water in the reducing gas. The dispersity of the metallic phase has been determined by volumetric adsorp- tion of hydrogen at room temperature. Stoichiometric R h - - H was used, in agreement with previous results [7]. The results of chemisorption were checked by transmission electron microscopy and a good agreement was observed. Table 1 summarizes the main characteristics of the catalysts.

2) Catalytic measurements. A flow microreactor was used in the differentia[ mode. An aliquot of the sample used for dispersity measurements was activat- ed under flowing hydrogen at 473 K. The reactants were chlorobenzene from Prolabo (purity>99.99%) and hydrogen from Air liquide (purity>99.995%). Hydrogen was saturated with the partial pressure of ehlorobenzene at 273 K (2.53 Tort) and passed through the reactor. The reaction temperature was 353 K. The products were analyzed by on line gas chromatography using a column (3 reX 1/8") packed with 10% Carbowax 20 M on Chromosorb W (80--100 mesh).

The reaction yields benzene, chlorocyclohexane, cyclohexane and HC1;: HC1 is not detected by the flame ionization detector used.

RESULTS AND DISCUSSION

In the conditions of the reaction, all samples suffer a deactivation, whick is due to interaction of the catalyst with the HC1 produced in the reaction. With a partial pressure of reactant of 9 Tort, deactivation is noticed and a con- stant act ivi ty is reached after 20 h on stream. The initial activity cannot be. restored by a simple treatment under H~ even at the temperature of reduction. The initial activity is reobtained by pretreating the sample in air at 513 K and further reducing. This deactivation is therefore interpreted by the formation of a stable surface chloride. This deactivation depends on the partial pressure. of HC1 produced by the reaction and can be minimized by using a low partial pressure of reactant and low conversions. In the steady state a reversible in- hibition by HC1 was also noticed : the rate decreases as the conversion increases, The comparison of the rhodium catalysts was therefore made at zero conversion by extrapolating the results obtained at different contact times. In these con- ditions, in absence of ttC1, the reaction order relative to ehlorobenzene is close to zero.

158

COQ el al.: CONVERSION OF CHLOROBENZEN~.

Table 1 Physical characteristics of the rhodium catalysts

Reduction H / R h Sample Rhodium (%) Support temp. (K)

RAl l* 1.85 A120 a 320 m"/g 1103 7.5 RA12* 1.5 AI20 a 200 mX/g 1103 17 RA13* 2 AI~O~ 200 m2/g 873 42 RA14 1 AI~O a 200 m2/g 823 45 RA15 0.5 A120 a 200 m2/g 673 56 RA16 0.5 A120 a 320 m'~/g - - 80 RSi l 2 SiO s 320 m-"/g 803 49 RSi2 1 SiO 2 215 m2/g 773 75 RSi3 0.55 SiO 2 215 m2/g 773 S0

* reduced in presence of water (18 Torr)

The selectivity for chlorocyclohexane remains constant when the degree of con- version increases. I t may then be assumed that eyclohexane is formed by hy- drogenation of benzene and not from dechlorination of chlorocyClohexane. T he hydrodechlorination rate is therefore taken as the sum of the rates of formation of benzene and cyelohexane.

The results obtained using this procedure are reported in Table 2 ; the activi- ties are expressed as turnover frequencies o r activities per surface metal atom. Different supports were used as silica and alumina carriers. These supports of' different origins contain different amounts of impurities, bu t give similar results. Therefore, the results observed here are not artefacts induced by these, impurities, and have a real physical significance.

Table 2 Catalyt ic properties of the rhodium catalysts at 353 K

Turnover frequencies (h-') Sample Dispersity Bz

(%) total C--CI hydrogenation hydrogenolysis a CCIID

RAIl 7.5 150 142 8 17.8 RAI2 17 170 166 4 41.5 RA13 42 22 21 1 21 l~A14 45 28 27.5 0.25 110 RA15 56 44 43.5 0.5 87 RA16 80 60 59 0.5 120 RSil 49 32 31 I 31 RSi2 75 33 32 i 32 RSi3 80 40 38 2 19

a) ra te of formation of benzene + cyclohexane b) B Z / C C H : r a t i o of b e n z e n e to c h l o r o b e n z e n e

T h e i n f l u e n c e o f p a r t i c l e s i z e a p p e a r s o n R h / A l ~ O 3 : l a r g e p a r t i c l e s a r e m o r e a c t i v e t h a n s m a l l p a r t i c l e s f o r d e c h l o r i n a t i o n a n d f o r h y d r o g e n a t i o n o f t h e a r o m a t i c r i n g . T h e i n f l u e n c e o f t h e s u p p o r t i s n e g l i g i b l e o n t h e a c t i v i t y f o r

159

COQ et at.: CONVERSION OF CHLOROBENZENE

hydrogenolysis, but noticeable on the activity for hydrogenation. Tile selec- t iv i ty of the conversion (benzene/chlorocyelohexane ratio) depends on the par- ticle size and the support: small particles supported by silica behave as Iarge particles supported on alumina from the standpoint of selectivity.

The conclusion that the hydrogenation of the benzene ring of chlorobenzene is a structure sensitive reaction is surprising, since the hydrogenation of benzene to cyclohexane on the same catalysts was found to be insensitive to the disper- si ty of rhodium [7]. The high puri ty of the ehlorobenzene reactant allows the accurate determination of ehlorocyctohexane at conversions as low as 0.01% and these conditions ensure the isothermieity of the system. The 16-fold varia- tion observed here appears to be beyond experimental error.

The rate of chlorocyc]ohexane production is 100 to 1000 times lower than the rate of benzene hydrogenation. Harper and Kemball [8] reported that the rate of H~/De exchange on noble metals is much lower for chlorobenzene than for benzene. These two facts may be ascribed to a weaker interaction between the aromatic ring and the metallic surface in the ease of chlorobenzene. Burwell e t at. [9] have suggested that the lack of structure sensitivity in the hydrogen- ation of alkynes may be related to a partial extraction of the platinum atom above the plane of a face. This effect needs a strong adsorption of the reactant. The observation of some structure sensitivity in the ease of the hydrogenation of chlorobenzene could then be explained by the weaker interaction with t h e surface. The weakly adsorbed reactant could then be more sensitive to the small modifications of the electronic properties at the surface atom.

In the case of dechlorination, the adsorption may be represented as an oxida- t ive addition to rhodium:

Rh ~ + PhC1 ~ Ph--RhIx--C1

followed by the reductive elimination of tIC1. The poisoning effect of HC1 is then readily interpreted by a competition with chlorobenzene in this reaction.

C--C1 hydrogenolysis proceeds easily on larger particles supported on alu- mina. A similar behavior was observed with thiophene [3] and was interpreted b y the modifications of the electron density at the surface rhodium atoms when changing the particle size. These two reactions have the common character to induce a formal oxidation of rhodium. The interpretation finds some support in the theoretical calculations made on Ni clusters [10]. These calculations show that the local density of states is broader on large crystals than on small particles. This is equivalent to the creation of lower energy levels which would "tend to increase the reactivity towards electron acceptors.

In conclusion, the hydrogenation and hydrodechlorination of chlorobenzene appear as structure sensitive. I t is therefore expected that the selectivity of C1 elimination in more complex reactions depends on particle size, support, alloying or poisoning, as is usually observed with this class of reactions.

R E F E R E N C E S

1. G. A del Angel, t~. Coq, F. Figueras: Studies in Surface Science and Catalysis, vol. 11, p. 63. Elsevier, New York 1982.

160

COQ et al.: CONVERSION OF CHLOROBENZENE

2. G. A del Angel, B. Coq, t l . Dutar t re , F. F igureas : J. Catal. (1984), in press. 3. G. A del Angel, B. Coq, F. F igueras : Nouveau J. Chim., 7,173 (1983). 4. French 7106074; U.S. 3,920,743. 5. P. Fouil loux, G. Cordier, Y. Colleuille: Studies in Surface Science and Catalysis,

vol. 11, p. 369. Elsevier, New York 1982. 6. P. I)ini, J. C. J. Bar t , M. Giordano: J. Chem. Soe. Perkin I I , 14, 1479 (1975). 7. S. Fuentes , F. F igueras : J. Catal. , 61, 443 (1980). 8. 1~. J. Harper , C. Kemba l l : Trans. F a r a d a y Soe., 65, 2224 (1969). 9.1~. L. Burwell, H. K. I{2ung, 1%. J. Pel le t : Proe. gth Intern. Congr. Catal. , London

1976. Vol 1, p. 108. The Chemieal Society 1977. 10. M. B. Gordon, F. Cyro t -Lackmann, M. C. Desj0nqu6ires: Surf. Sci., 68, 359 (1977).

11 161