Coking of SO4=2?ZrO2 Catalysts during Isomerization of n-Butane and Its Relation to the Reaction Mechanism

VERA, C.; PIECK, C.L.; SHIMIZU, K.; QUERINI,C.A.; PARERA, J.M.

JOURNAL OF CATALYSIS

ACADEMIC PRESS INC ELSEVIER SCIENCE

Año: 1999 vol. 187 p. 39 - 49

0021-9517

A study of the deactivation of SO2¡SO2¡ 4 ?ZrO2 catalysts during isomerization of n-butane showed that a small amount of a carbonaceous deposit (1?1.2%) is formed. A high proportion of this deposit is produced during the first minutes of the reaction and it seems to be responsible for the initial short-term catalyst deactivation but it is doubtful whether it is connected to the long-term loss of activity. Activity in isomerization of n-C4 was very sensitive to activation conditions while coking was more consistent; in all cases, SZ catalysts had a limiting content of 1 to 1.2% coke when fully deactivated. In an oxygen flow, the coke deposit could be burned at 500±C, but it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=zup to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=zup to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z conditions while coking was more consistent; in all cases, SZ catalysts had a limiting content of 1 to 1.2% coke when fully deactivated. In an oxygen flow, the coke deposit could be burned at 500±C, but it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=zup to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=zup to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z deposit (1?1.2%) is formed. A high proportion of this deposit is produced during the first minutes of the reaction and it seems to be responsible for the initial short-term catalyst deactivation but it is doubtful whether it is connected to the long-term loss of activity. Activity in isomerization of n-C4 was very sensitive to activation conditions while coking was more consistent; in all cases, SZ catalysts had a limiting content of 1 to 1.2% coke when fully deactivated. In an oxygen flow, the coke deposit could be burned at 500±C, but it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=zup to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to be directly linked to the coking process 3 m=z it could also easily be removed by stripping with an oxygen-free inert gas at higher temperatures. In the latter case, the coke was oxidized by the catalyst surface groups, most likely sulfate. The coke was soluble only in polar solvents (methanol, pyridine) and was highly insoluble in benzene and hexane. An FD-mass spectrum of the soluble fraction of the coke indicated the presence of compounds with m=z up to 2700, but the distribution was concentrated aroundm=zD350 to 500. The latter value, when compared with the small amount of total carbon, indicates that only a small fraction of the surface sites was affected by the coke, or that deactivation proceeded by a mechanism other than site blocking. A surface reduction process by the reacting hydrocarbon may also be linked to the activity loss. ESR results indicated that the catalyst was also reduced during the reaction; the intensity of the Zr3C signal was higher on the coked catalysts. The probable concurrence of reduction could explain the long-term activity decline which seems not to