Methane silicalite

In the direct ammoxidation of propane over Fe-zeolite catalysts the product mixture consisted of propene, acrylonitrile (AN), acetonitrile (AcN), and carbon oxides. Traces of methane, ethane, ethene and HCN were also detected with selectivity not exceeding 3%. The catalytic performances of the investigated catalysts are summarized in the Table 1. It must be noted that catalytic activity of MTW and silicalite matrix without iron (Fe concentration is lower than 50 ppm) was negligible. The propane conversion was below 1.5 % and no nitriles were detected. It is clearly seen from the Table 1 that the activity and selectivity of catalysts are influenced not only by the content of iron, but also by the zeolite framework structure. Typically, the Fe-MTW zeolites exhibit higher selectivity to propene (even at higher propane conversion than in the case of Fe-silicalite) and substantially lower selectivity to nitriles (both acrylonitrile and acetonitrile). The Fe-silicalite catalyst exhibits acrylonitrile selectivity 31.5 %, whereas the Fe-MTW catalysts with Fe concentration 1400 and 18900 ppm exhibit, at similar propane conversion, the AN selectivity 19.2 and 15.2 %, respectively. On the other hand, Fe-MTW zeolites exhibit higher AN/AcN ratio in comparison with Fe-silicalite catalyst (see Table 1). Fe-MTW-11500 catalyst reveals rather rare behavior. The concentration of Fe ions in the sample is comparable to Fe-sil-12900 catalyst, as well as... 

Methane (continued) silicalite, 42 66-70 zeolite A, 42 64-65 vibrational spectra, 42 207-210 ZSM-5 self-diffusion coefficients, 39 369-371... 

Supported Fe-Mn Fischer-Tropsch Catalysts. A much more limited number of studies have dealt with supported Mn-promoted Fe F-T catalysts. In this respect, it is worthwhile to mention the work of Xu et al These authors added MnO to a Fe/silicalite catalyst and observed an enhanced selectivity towards light olefins. Meanwhile the yields for methane as well as for CO2 formation were almost unaffected by MnO addition. Moreover, the conversion of CO was also insensitive to the addition of the MnO promoter. 

There are experimental results that show the anisotropic nature of diffusion of methane in silicalite (24, 77). From a stochastic jump model of the diffusion process, Karger et al. (24) found that the ratio of the rate of diffusion in the direction of the two channel systems should not be less than 4.4 times that in the orthogonal direction ... 

Other flexible framework calculations of methane diffusion in silicalite have been performed by Catlow et al. (64, 66). A more rigorous potential was used to simulate the motion of the zeolite lattice, developed by Vessal et al. (78), whose parameters were derived by fitting to reproduce the static structural and elastic properties of a-quartz. The guest molecule interactions were taken from the work of Kiselev et al. (79), with methane treated as a flexible polyatomic molecule. Concentrations of 1 and 2 methane molecules per 2 unit cells were considered. Simulations were done with a time step of 1 fs and ran for 120 ps. 

Nowak et al. (63) presented a comparative study of the diffusivities of rigid models of methane, ethane, and propane in silicalite. (The details of the calculation are reported in the preceding section.) The calculated diffusion coefficients decreased as the length of the carbon chain increased, and the effect was found to be far more pronounced for ethane than propane. The calculated diffusivities, in units of 108 m2/s, were 0.62, 0.47, and 0.41 for methane, ethane, and propane, respectively. The ethane value is in satisfactory agreement with PFG-NMR measurements [0.38 (77), 0.3 (80), 0.4 (42) for silicalite. The value for propane, however, was calculated to be almost an order of magnitude larger than the NMR results of Briscoe et al. (80). [The agreement with the value of Caro et al. (71) is better, but still an overestimation.]... 

Nicholas et al. (67) have performed MD calculations of propane in sili-calite in which the propane molecule is given complete flexibility. The calculations, which have been detailed previously for methane diffusion, employed a large simulation box with multiple sets of adsorbates to ensure good statistics. The framework was kept fixed and data were collected over a 40-ps run. The results predict diffusion coefficients in very good agreement with the values of Caro et al. (71). The calculated values for a concentration of 4 and 12 propane molecules per silicalite unit cell are 0.12 and 0.005 X 10 8 m2/s, respectively. These values for propane are far lower than those of Nowak et al. (63), the reason for this is that Nicholas et al. used flexible adsorbate molecules, whereas Nowak et al. used rigid ones. 

In Silicalite. A variety of papers are concerned with sorption of methane in the all-silica pentasil, silicalite. June et al. (87) used a Metropolis Monte Carlo method and MC integration of configuration integrals to determine low-occupancy sorption information for methane. The predicted heat of adsorption (18 kJ/mol) is within the range of experimental values (18-21 kJ/ mol) (145-150), as is the Henry s law coefficient as a function of temperature (141, 142). Furthermore, the center of mass distribution for methane in silicalite at 400 K shows that the molecule is delocalized over most of the total pore volume (Fig. 9). Even in the case of such a small sorbate, the channel intersections are unfavorable locations. 

In an MD study of methane sorption and diffusion in silicalite, Nicholas et al. (67) identified favorable sites for sorption. From the MD calculations, the time-averaged position of the center of mass of the methane molecule was plotted. Energy minimization calculations were then performed, locating the methane molecule at positions where the MD calculations predicted they spent the most time. Each channel intersection region was found to contain two sites that are minima for methane-zeolite interactions. These two sites are separated by a translation parallel to the straight channel... 

Fig. 9. Center of mass distribution of methane in silicalite at 400 K. The light grid illustrates the pore volume, and the dark shows delocalization over most of the cell. Reprinted with permission from Ref. 87. Copyright 1990 American Chemical Society.

Adsorption isotherms of methane in silicalite have also been predicted in a number of calculation studies (62, 155, 156). Goodbody et al. (62) predicted a heat of adsorption of 18 kJ/mol and simulated the adsorption isotherm up to 650 bar. From the adsorption isotherm, they found that the sinusoidal pore volume contains more methane molecules at all pressures. Snurr et al. (155) performed GC-MC and MD simulations over a wide range of occupancies at several temperatures. The intermolecular zeolite-methane potential parameters were taken from previous MD studies (11, 87) and the methane-methane parameters from MD simulations were adjusted to fit experimental results for liquid methane (157). Electrostatic contributions were neglected on account of the all-silica framework, and methane was represented by a rigid, five-center model. 

Maginn EJ, Bell AT, Theodorou DN (1993) Transport diffusivity of methane in silicalite from equilibrium and nonequilibrium simulations. J. Phys. Chem. 97 4173—4181... 

Molecular mechanics (MM), molecular dynamics (MD), and Monte-Carlo (MC) methods were employed to simulate the adsorption of methane, ethane, propane and isobutane on silicalite and HZSM-5. The silicalite was simulated using the same cluster-model adopted in the diffusion calculations. The H-ZMS-5 structure was constructed according to the procedure suggested by Vetrivel et al. [32], which consists in replacing one atom at the channel intersection by and protonating the oxygen atom bridging the Ta and Tg sites in order to preserve the lattice neutrality. 

In this article we present an experimental method to measure the selectivity in binary gas adsorption under infinite dilution conditions over a range of pressures and thereby characterize its behavior with respect to both pressure and composition. The method is hist, efficient and robust. We use the methane-ethane binary gas mixture on silicalite as a demonstration. The experimental results obtained using this method are compared with the predictions from two models to check their validity. 

The major limitation of this technique is that it may not be suitable to use it for highly selective systems like methane-butane mixture on silicalite. For these systems the retention times may be long and a significant dispersion in the peak might occur. 

This result is in agreement with recent PFG NMR measurements with oriented ZSM-5 crystals, where the mean self-diffusivity of methane averaged over the straight and sinusoidal channels (i.e., in the x, y plane) has been found to be larger by a factor of 3-5 than the self-diffusivity in the c direction, perpendicular to those 42,43). For the self-diffusion of methane in ZSM-5/silicalite-I, the MD simulations were in very satisfactory agreement with the experimental PFG NMR and QENS self-diffusion data (cf. Refs. 34-38). Furthermore, the degree of a mass transport anisotropy in the MFI framework as predicted by MD simulations (34-38) is compatible with the experimental findings (42,43,49). 

In Fig. 2 the fluxes of several light hydrocarbons through a silicalite-1 membrane are shown as a function of their partial pressure on the feed side. The trend that can be deduced from this figure is that as the molecules get larger, their flux becomes lower. This decrea.se in flux is, however, smaller than expected on the basis of differences in diffusion coefficients [14]. The increase in the size of the molecule results in a lower mobility in the pores, but this effect is partly compensated by the higher concentration in the membrane, due to better adsorption of the larger molecules. This compensation effect is also the reason that at low partial pressures, ethane permeates faster through the membrane than does methane. 

Figure 2 Flux of methane ( ), ethane, ( ), propane ( ), n-buiane ( ), and iso>butane (A) through a silicalite-1 membrane as a function of partial pressure on the feed side (T = 298 K, = 100 kPa). Feed was composed of hydrocarbon and balance helium sweep gas used was helium. There was no absolute pressure difference across the membrane. (Adapted from Ref. 14.)...

The same model was applied to permeation of lighter hydrocarbons (C1-C3) through the silicalite-1 membrane [50]. In the case of methane, ethane, and ethene, some concentration dependence of the Maxwell-Stefan diffusivity was observed. This can be caused either by the importance of interfacial effects, which are not taken into account, or by the contribution of activated-gas translational diffusion to the net flux. The diffusivities calculated from these permeation experiments were, however, in rather good agreement with diffusivity values from the literature, which implies that these zeolitic membranes could also be a valuable tool for the determination of diffusion coefficients in zeolites. 

The temperature dependence of the methane permeation through a silicalite membrane, showing a maximum and a minimum as a function of temperature (Fig. 3 [14]), can not be predicted by using the Maxwell-Stefan description for surface diffusion only. Such a maximum and minimum in the permeation as a function of temperature can be predicted only when the total flux is described by a combination of surface diffusion and activated-gas translational diffusion (Fig. 15). 

Figure 19 Pore apertures of DOH (6-membered ring. 0.28 nm), zeolite-A (LTA, 8-membered ring, 0.41 nm), silicalite-1 (MFI, 10-membered ring, 0.52 X 0.55 nm), and a zeolite-X or -Y (FAU, 12-membered ring, 0.74 nm) and a methane molecule (kinetic diameter = 0.38 nm).

As an example. Figure 10.15 compiles the permeance evolution with temperamre for single hydrocarbons (from Ci to C3) over silicalite-1 membranes supported on stainless steel tubes. A specific interaction with hnear hydrocarbons appears, as it could be expected due to the organophilic character of SIL-1 membranes. For methane, the temperature used is too high to find... 

Jost S, Bar NK, Fritzsche S, Haberlandt R, and Karger J. Diffusion of a mixture of methane and xenon in silicalite A molecular dynamics study and pulsed field gradient nuclear magnetic resonance experiments. J Phys Chem B 1998 102 6375-6381. 

PFG (pulsed field gradient) NMR data determined the diffusion coefficients of methane in 3 MFI-type siliceous zeolite silicalite samples.640 The rotational motion of butane and pentane molecules adsorbed on zeolite ZK-5... 

---