Intracrystalline coking

Rollmann and Walsh (266) have recently shown that for a wide variety of zeolites there is a good correlation between shape-selective behavior, as measured by the relative rates of conversion of n-hexane and 3-methyl-pentane, and the rate of coke formation (see Fig. 24). This correlation was considered to provide good evidence that intracrystalline coking is itself a shape-selective reaction. Thus, the rather constrained ZSM-5 pore structure exhibits high shape selectivity, probably via a restricted transition-state mechanism (242b), and therefore has a low rate of coke formation. Zeolite composition and crystal size, although influencing coke formation, were found to be of secondary importance. This type of information is clearly... 

The effect of coke deposition on the MTO reaction is complex. Coke deposition influences either the formation of dimethyl ether (DME) or the DME conversion inside the pores during MTO. However, the effect of coke deposition on the dimethyl ether conversion to light olefins (the DTO process) catalyzed by SAPO-34 is much simpler and can allow us to focus on the effect of intracrystalline coke on the olefin formation from DME. 

The different coking behaviors of n-hexane and mesitylene on ZSM-5 can also be demonstrated by the ratios of Dmm/Dcff (Fig. 40). Because mesitylene coking produces mainly surfece coke, the intracrystalline mobility of methane remains unchanged, i.e., Dmin/D ff > 1. In the case of coke formation by n-hexane cracking, surfece coke (Z)intra/ >efr > 1) is formed after an initial period of intracrystalline coke deposition (Dimra/Deir = ) ... 

The self-diffusivity of propane in the coked polycrystalline grains unveils details of coke formation in polycrystalline particles. Figure 44 shows that coking reduces the translational mobility inside the grains. The effect of intracrystalline coke deposition on the translational mobility of propane is indicated in Fig. 43. After a coking time of 1 h, only a slight increase of Di ,ra with increasing observation times occurs. After 12 h, a decrease is ob-... 

The coke formation enhances the ethene formation. The ethene to propene increased with intracrystalline coke content, regardless of the nature of coke. 

It is seen that the intracrystalline MgO induces pore blockage in a fraction of the pore system and alters the porosity as well as D0, and/or r with the latter factors contributing most to the reduced diffusivity. In contrast, the coke modifier appears to affect mainly the surface-to-volume ratio and suggests that the effective surface area, number of available entrance ports, is reduced by two orders of magnitude. 

Water and organic molecules occluded during the synthesis were removed from the intracrystalline volume as follows. The solids were slowly heated (5°C/min) in a N2 flow up to 550°C and held at this temperature for 2h. The coke deposit resulting from the non-oxidative degradation of the organics was then... 

Dealumination of HM resulted in common features for both the isopropylation of biphenyl and naphthalene. The dealumination of HM decreases the blocking of the pores by coke deposition in the both reactions because of the decrease in acid sites at intracrystalline and external surfaces. It also reduced the non-regioselective isopropylation at the external surface, and enhanced the reaction to yield the least bulky product molecules. The pressure of propylene is also an important key factor for high yield of 4,4 -DIPB in the isopropylation of biphenyl, because the isomerization of 4,4 -DIPB occurred significantly under low pressure of propylene at the external acid sites. 

Pyrolysis of Sodium-Tetramethylammonium Zeolite Omega. Preliminary calcinations of the Q zeolites showed that intracrystalline diffusion restrictions interfered greatly with transport both of oxygen and of calcination products. Under mild conditions, coking was observed, and even under favorable conditions (550° C, thin beds, good venting) the reaction was slow. Some samples of zeolite were pyrolyzed under vacuum, and the products were identified by low resolution mass spectrometry. 

The intracrystalline pore volume of the catalysts was evaluated by n-hexane sorption as shown in Fig. 6. Sorption capacities for samples SI to S3 are comparable to that of the zeolite before Ga impregnation and correspond to the value expected for an unaltered ZSM-5 type material (S10). Sorption capacity decreases for samples S3, S4, S5, and S6, because of intracrystalline volume blockage by coke deposits and possibly also (silica)-alumina debris [6] in the aged catalyt S6. In addition, the sorption rate for S6 is about twice the rate observed for the other samples, suggesting that adsorption occurs mostly at the external surface of the S6 catalyst crystallites. Thus, it appears that coke deposited on S6, probably as polyaromatic species, has almost blocked the channel pore mouths and/or practically occupied the whole intracrystalline pore volume. It explains the poor catalytic performance of S6. 

The classical method of investigation of effects of diffusion on reactions is typically to run a reaction with catalyst particles of various sizes. For zeolites, the resistance of intracrystalline diffusion is normally much larger than that characteristic of molecular diffusion or Knudsen diffusion that could occur in the spaces between the zeolite crystals in a catalyst particle. Thus, the crystal size of the zeolite has to be varied instead of the particle size to determine the effects of diffusion on zeolite-catalyzed reactions. Kinetics of the MTO reaction has been measured with SAPO-34 crystals with identical compositions and sizes of 0.25 and 2.5 pm 89). The methanol conversion was measured as a function of the coke content of the two SAPO-34 crystals in the TEOM reactor. 

This case study clearly illustrates the usefulness of the ZLD-TEOM technique in determining intracrystalline diffusivities in zeolites, provided that effects of other transport resistances such as the surface barrier are eliminated by varying the crystal size of the zeolites. The measured steady-state diffusivity can be directly used for predicting effects of diffusion in reactions catalyzed by zeolites. More important, the TEOM makes it possible to distinguish the deactivation caused by blockage of the active sites and by increased diffusion resistance caused by blockage of cavities or channels by coke. 

Zeolites are solid acid catalysts which are widely used in hydrocarbon processing, such as naphtha cracking, isomerization, dispropornation and alkylation. During reactions carbonaceous materials called coke deposit on the zeolite and reduces its activity and selectivity. Coke deposited not only covers the acid sites of the catalyst, but also blocks the pores, and restrain reactants from reaching the acid sites, leading to the decrease in the apparent reaction rate (1, 2). Here, we will mainly deal with the intracrystalline diffusivity of zeolites, and will discuss the relationship between it and the change in catalyst selectivity. 

Figure 5) NMR intracrystalline self-diffusion coefficient Di ( ) and NMR desorption diffusivity Dd of methane sorbed in ZSM5 with increasing coking time, (reproduced with the permission from reference 18)...

A reduction of the intracrystalline self-diffusion coefficient, Dintra, of a probe molecule (e.g., methane) as measured by PFG NMR has to be regarded as proof of the existence of coke, modifiers, etc. in the volume phase of the zeolite crystal. 

The absence of additional diffusion barriers at the crystal surface (coke, modifiers, etc.) can be assumed if the experimentally determined mean intracrystalline lifetimes, Timra (measured by TD NMR), and the corresponding calculated data, rSitra [according to Eq. (10)], coincide. The rSim data are calculated by using the self-diffusion coefficients, Dimra (measured by PFG NMR), and crystal radii, R, assuming the adsorption/desorption process to be diffusion controlled. 

Flo. 37. NMR intracrystalline self-diffusion coefficient Dm, (a) and effective self-diffusivity Dcir ( ) of methane in HZSM-5 crystals that were coked for different times by n-hexane cracking (131-133). Before loading with methane (9.2 CHa per u.c.), the coked ZSM-5 crystals were carefully outgassed at 623 K and 10 Pa. The remaining carbonaceous residues were defined as coke. Amounts of coke after different times on stream 1 h, 0.8 wt% C 2 h, 1.3 wt% C 6 h, 3.2 wt% C 16 h, 4.8 wt% C. The starting self-diffusion coefficient is 8.1 x 10" m s . ... 

In contrast to HZSM-5 coked by n-hexane, for mesitylene-coked HZSM-5 (Fig. 39) the intracrystalline mobility, Dmtfa, of methane is nearly unaffected by coke deposits (42,131). On the other hand, the effective selfdiffusion coefficient, decreases continuously with increasing time onstream, i.e., Deff < Dintra. The conclusion is that mesitylene coking leads to the pronounced formation of a surface barrier, i.e., from the beginning of... 

The crystalline aluminosilicate-catalyzed aldol condensation of acetophenone to form dypnone has been reported (27). As shown in Table XXIII, hydrogen zeolites were the most effective catalysts for this conversion. Operation at low temperatures in the liquid phase is critical for this reaction, to avoid both coke formation and condensation with aromatic solvents. Catalyst aging was rapid, however. Only transient conversions of acetone to mesitylene were obtained over REX or H-mordenite at 315° owing to rapid intracrystalline self-condensation and coke formation. 

During mesityiene coking, the carbonaceous compounds are found to be exclusively deposited on the outer surface. For n-hexane, two stages of the coke deposition become visible At shorter coking times n-hexane is mainiy deposited in the intracrystalline space, tnus simultaneously affecting a retardation of intracrystaliine diffusion ana tracer desorption. In a second stage, similar to the behaviour observed with mesityiene, coke is predominantly deposited on the crystallite surface. 

Figure 6 Intracrystalline mean life times t,<. > and Tini,, "1" (<4 ) for methane at 296 K and a sorbate concentration of 3 molecules per channel intersection in H ZSM-5 coked by n-hexane (full symbols and mesitylene (open symbols in dependence on the time on stream (Reproduced with permission from Ref. 6. Copyright 1 >S7 Butterworth ...

Influence of the addition of silica, as a binder at a concentration of 10 or 50 wt%, to H-gallosilicate (MFI) zeolite on its inter- and intracrystalline acidity, initial activity, product selectivity and distribution of aromatics formed in the propane amortization (at 550°C) and also on its deactivation due to coking in the aromatization process has been thoroughly investigated. Silica binder caused an appreciable decrease in the zeolitic acidity (both external and intracrystalline acid sites) and also in the propane conversion/aromatization activity. Because of it, the deactivation due to coking of the zeolite in the propane aromatization is, however, decreased. The deactivation rate constant for the initial fast deactivation is decreased but that for the later slow deactivation is increased because of the binder. The aromatics selectivity for aromatics and para shape selectivity of the zeolite, particularly at lower conversions, are increased but the propylene selectivity and dehydrogenation/cracking activity ratio are decreased due to the presence of binder in the zeolite catalyst. 

Figure 26 compares values for the intracrystalline mean lifetime and Tin,ra " for methane in ZSM-5 type crystallites after different coking times and of the values of [145,187]. Depending on the applied coking compound,... 

Modification of H-ZSM-5 zeolites by impregnation with H3PO4 has become a common technique to improve their activity and selectivity [190,191]. In parallel with these changes, however, the transport properties of the zeolite crystallites are also changed. As in the cases of hydrothermal treatment and coking discussed above, a combined application of PFG NMR to study both intracrystalline diffusion and intercrystalline molecular exchange may provide information about... 

---