Catalysts CoMo-alumina

A series of CoMo/Alumina-Aluminum Phosphate catalysts with various pore diameters was prepared. These catalysts have a narrow pore size distribution and, therefore, are suitable for studying the effect of pore structure on the deactivation of reaction. Hydrodesulfurization of res id oils over these catalysts was carried out in a trickle bed reactor- The results show that the deactivation of reaction can be masked by pore diffusion in catalyst particle leading to erro neous measurements of deactivation rate constants from experimental data. A theoretical model is developed to calculate the intrinsic rate constant of major reaction. A method developed by Nojcik (1986) was then used to determine the intrinsic deactivation rate constant and deactivation effectiveness factor- The results indicate that the deactivation effectiveness factor is decreased with decreasing pore diameter of the catalyst, indicating that the pore diffusion plays a dominant role in deactivation of catalyst. 

The second route is the hydrogenation in combination with hydrogenolysis. In this case, it also has the phenol as an intermediary (Figure E23.7). Viljava et al. (2000) identified the major products cyclohexane and cyclohexene with the catalyst CoMo supported on alumina. However, in our study, the cyclohexene was not observed, but only the cyclohexane. This observation is in agreement with the literature, where the NiMo and Ni catalysts supported on alumina have higher capacity of hydrogenation compared to Co catalysts (Senol et ah, 2007). 

The Nature of Exposed Co In CoMo Catalysts. Figure 1 shows typical original spectra for NO adsorbed on 2% Co/alumlna. Spectra were recorded after reduction at 600°C, after admission of 8 Torr of NO, and again after 5 min. evacuation at 50°C. With the pure alumina the main NO/Co bands were 10 to 20 wavenumbers higher In frequency than those observed when using alumina with sulfate Impurity. 

Figure 6 shows spectra of NO adsorbed on a similar series of catalysts having the same concentrations of metals on Filtrol 90 alumina which, as discussed, sulfides the catalyst during subsequent reduction. The effect of sulfide, as on Mo/alumlna catalysts (8,10) and as later seen on CoMo catalysts after more-usual sulfiding, was mainly a marked reduction In the Intensities of the NO bands. These were roughly a fourth as Intense as those on unsulflded catalysts on... 

Figure 9 shows CO adsorbed on CoMo catalysts made on two alumina supports. Most of the CO was weakly-held, being largely removed by evacuation for 5 min. at 50 C. Depending on the extent of sulfiding during reduction, various bands were obtained. In the absence of sulfide, as previously seen on Mo/alumlna catalysts (, only Mo (4+ ) or Co (2+) Ions appear to be exposed after reduction, giving... 

Mo/alumina ( ). Exposed Co Ions could also serve as members of active pairs, however. This possibility gains some support from the observation that optimum Co concentrations on CoMo catalysts... 

This work is a contribution to the understanding of the effect of spillover hydrogen in a type of catalyst of considerable industrial importance, namely that composed of transition metal sulfides and amorphous acidic solids. This is typically the case of sulfided CoMo supported on silica-alumina used for mild hydrocracking. 

Pure silica-aluminas are strongly deactivated, losing about 80% of their activity before reaching the steady-state. The loss in pure CoMo/Si02 catalyst is much less pronounced (about 15%). Mechanical mixtures represent an intermediate case they lose between 35% and 50% of their activity. 

In this series of experiments, the catalysts were used over five repeat contacts with fresh coal liquid. Point of Ayr coal liquid was supplied by the British Coal Corporation, Coal Research Establishment (CF ) one batch of this coal liquid was used in experiments with CoMo and NiMo catalysts and a further batch was used in experiments with ZnMo and ZnW. The catalysts were prepared as extrudates by the technique of incipient wetness which requires stirring the dry alumina support with a set volume of a pre-determined concentration of an appropriate soluble salt of the metal such that the pore space is just taken up by the metals at the required concentration. The alumina support was supplied by Akzo Chemie, The Netherlands and the catalysts were made up to contain 15% WO3 or M0O3 and 3% NiO, CoO, or ZnO, expressed as a weight percentage of the weight of support... 

The coals used were PSOC 1098 Illinois 6 and Beulah-Zap North Dakota lignite from the Argonne coal bank. The analytical data of these coals are shown in Table I. The ratio of catalyst to coal was approximately 0.6 mmoles of metal per gram of coal. The organometallic catalysts were molybdenum(II) acetate dimer, Mo2(OAc)4, obtained from Strem, molybdenum(II) allyl dimer Mo2(OAc)4, was prepared by die method of Cotton and Pipal (25). The NiMo supported catalyst was prepared by addition of bis(l,5-cyclooctadiene) Ni(0) (Strem) to sulfided Mo on alumina (. Cp2Mo2( l-SH)2(p.-S)2, referred to as MoS2(OM), was prepared by modification of method of Dubois et al. (26), and Cp2Mo2Co2( i3-S)2(li4-S)(CO)4, CoMo(OM) was prepared by the method of Curtis et al. (27). Pentacarbonyl iron was obtained from Aldrich,... 

Several authors have pursued this approach and indeed observed that desulfurization of 4.6-DMDBT was increased when acidic zeolites were used in combination with conventional HDS catalysts (30, 31, 33, 137, 149-151). Figure 34 shows that there can be a great acceleration in the conversion of 4,6-DMDBT through the use of a hybrid catalyst consisting of CoMo impregnated into a composite containing 5% NiY zeolite and alumina (137). Low temperatures had to be employed, as at temperatures exceeding about 340°C, severe color fluorescence occurred in the product. 

In 1959, H. Beuther et al. (8) of Gulf Oil Company published the first systematic study of the HDS activity of CoMo and NiMo supported on alumina as a function of the atomic ratio Co(Ni)/Mo. As a result, they showed what they called a promoter effect of the cobalt (or nickel) on the molybdenum for atomic ratios Co/Mo = 0.3 and Ni/Mo = 0.6. This publication was preceded by several patents proposing similar atomic ratios for cobalt by Union Oil of California (1948) (9) and Shell Oil Company (1954) (10) and for nickel by Union Oil of California (1954)(/7). Figure 1 shows a typical activity curve of NiMo/Al203 catalysts as a function of the value of the atomic ratio Ni/Mo (12). 

The first research group to propose a description of the structure of CoMo catalysts was led by Schuit and Gates (13). This group introduced the so-called monolayer model directly derived from the physical studies of CoMo oxide precursors supported on y-alumina carried out by J. T. Richardson (14) (Richardson first proposed the existence of a special Co/Mo entity.) In this model the upper or first layer contained only sulfur atoms, each bonded to a molybdenum atom of the second layer (below the first one), these molybdenum atoms being bonded to two oxygen atoms also located in this second layer. When a sulfur atom was removed by reduction (H2 flow) of Mo5+ to Mo3+, a vacancy was formed at the surface and became the preferential adsorption site of a sulfur atom in the organic gas phase. The presence of cobalt incorporated into underlying layers of the alumina... 

These tetrahedral distorted cobalt atoms can be observed by NMR as a pure phase on carbon supports in the absence of molybdenum and are thus stable these probably correspond to the Co sites observed by Topspe s group using Mossbauer spectroscopy because Craje et al. (93) found a similar Mossbauer doublet for both cobalt in CoMo catalysts and pure cobalt sulfide on carbon support. They are also active for HDS and confirm the findings of Prins and co-workers (94) and Ledoux (96). These different structures are in full agreement with the XANES experiments performed by Prins and co-workers (95) and Ledoux (96). These structures also led Ledoux et al. to an incorrect interpretation of the synergy effect (64). On poorly dispersed catalysts supported on silica or in bulk form, their presence and activity are large enough to explain the increase in activity when cobalt is added to molybdenum, but on well-dispersed catalysts i.e., on alumina or carbon support this interpretation is shown to be incorrect if the activity is carefully measured. 

Recently we have proposed an HDS catalytic treatment based on sodium-doped CoMo catalysts [Ref. 1-3]. Previous studies concerned essentially alumina-supported catalysts. As carbon was shown to be a good support for sulfided CoMo catalysts [Ref.4], we decided to investigate the performance of carbon-supported catalysts in terpene HDS. 

Impregnation of cobalt and molybdenum (without sodium) increases largely the isomerizing activity of the catalyst the /3-pinene is then completely converted. The catalysts prepared with sodium molybdate and sodium hydroxide (Co-Mo-Na and Na-Co-Mo-Na) have lower isomerizing activities while their HDS activities are significantly increased. As in the case of alumina supported catalysts the sulfided CoMo phase protected by a double layer of alkaline ions on the carbon support gives the best results in HDS of /3-pinene. The behaviour of this catalyst was examined in desulfurization of the turpentine oil (40% a-pinene, 25% /3-pinene, 25% A -carene and 10% camphene + dipentene + myrcene, 1500 ppm S). The results are recorded in Table 6. 

The most obvious choice to determine phases that may be present in the molybdena catalyst is XRD. Matching of diffraction lines obtained for the catalyst with those of pure bulk compounds gives unequivocal identification of phases present. This is one of the few techniques that yields positive results. The absence of matching diffraction lines, however, is not proof that the phase in question is not present in the catalyst. The XRD technique is limited to particle sizes of above approximately 40 A for oxides or sulfides, lower sized particles giving no discernible pattern over that of the broad alumina pattern. Thus, the presence of a highly dispersed phase, either as small crystallites or as a surface compound of several layers thickness will not be detected. Also, if the phase is highly disordered (amorphous), a sharp pattern will not be obtained, although some broad structure above that of the alumina may be detected. It is a moot point as to whether such a case is considered as a separate phase or a perturbation of the alumina structure. Ratnasamy et al. (11) have examined their CoMo/Al catalyst from the latter point of view, with particular emphasis on the effect of calcination temperature. 

Brown and Makovsky (77) have studied the Raman spectra of CoMo/Al commercial catalyst containing 5% Si02. The oxidized catalyst showed bridged Mo—O—Mo and terminal Mo=0 structures with no evidence for bulk Mo03. The sulfided catalyst showed spectra similar to MoS2 (also detected by XRD). There is reason to believe that Mo interaction with silica-alumina is weaker than with A1203 (28) thus, sulfiding may more easily destroy the surface interaction complex present in the oxidized catalyst. 

CoMo-124 Alumina was impregnated first with cobalt stepwise. The sample was dried at 120°C and calcined at 650 °C after each impregnation step. Afterwards the catalyst was impregnated with molybdenum. Final calcination temperature 650°C. The composition was the same as for MoCo-124. 

Some bulk cobalt aluminate formation is expected to take place for the boehmite based catalyst, owing to a Hedvall effect (24).The spectrum of adsorbed pyridine on CoMo-124 B shows indeed a weaker 1612 cm l band, comparable with the intensity of this band for the MoCo-123 catalyst. This indicates that about 25 % of the cobalt ions has disappeared in the bulk of the alumina. (Figure lid). 

---