Branched hexane

A (Figure 4.9). The diameter of such a neck, 2.3 A, is sufficiently large for a linear C-C chain to pass, but too small to also be an equilibrium adsorption position. The largest compound allowed inside the pores is a linear molecule limited in length to four carbon atoms due to the distance between two subsequent necks [103]. Another example of shape-selective behavior is found in a Zn-based MOF able to encapsulate linear hexane while branched hexanes are blocked [104]. 

Figure 13.4 CBMC simulations for adsorption of linear and branched hexanes over MFI [6],...

Figure 13.6 CBMC simulations for linear and branched hexane adsorption over AFI [6].

Of the branched paraffins, all of the four branched hexanes, ail but two of the eight branched heptanes, and all but two of the seventeen branched octanes are included. 

Reactions of 3-methylhexane were found to occur on the support alone which complicated the above studies. Besides aromatization to toluene (A), isomerization to cyclopentane derivatives (B), isomerization to heptane and other branched hexanes (C) and cracking to Ci to C6 hydrocarbons (D) were all observed as depicted in Eq. 5.23... 

S. J. Miller (Chevron) published results from early work that highlighted the selectivity of the platinum form of SAPO-11 catalyst compared to a number of others. These others were amorphous silica-alumina, from which one would expect little or no selectivity, ZSM-5, HY, and Na-Beta zeolites. All the catalysts carried 1 wt. % platinum and the feed employed was n-octane. He found that at 30% conversion, only SAPO-11, the amorphous silica-alumina, and the HY catalysts exhibited better than 94% selectivity for feed isomerization to isooctanes. ZSM-5 and Na-Beta catalysts behaved poorly in this regard. Selectivity for dimethylhexanes was low. SAPO-11 also produced equal quantities of 2- and 3-methyl heptanes, whereas the other catalysts favored 3-methyl heptane, with a ratio close to that favored by thermodynamics. SAPO-11 also produced one of the lowest levels of doubly-branched hexanes (Table 10.1646) and the predominant ones formed were those separated by more than one carbon—only minor amounts of the less thermally stable (bond breaking here can produce tertiary carbonium ions) geminal-dimethyl (2,2 and 3,3-) ones were formed. Noble metal presence was a key to success since replacement of the hydrogenation metal platinum by pallodium did not alter the isomeri-zation selectivity much, but replacement by nickel led to very poor isomerization. 

Figure 13 displays the self-diffusivities of n-hexane and 2-methylpentane in silicalite-1 and H-ZSM-5 as a function of the ratio of the hydrocarbons. The self-diffusivities of both hexanes linearly decrease with increasing gas-phase fraction of the branched hexane in the gas phase for the non-acidic and acidic zeolite. In H-ZSM-5, the mobility of alkanes is approximately two times slower than in silicalite-1. Obviously, the presence of acid sites strongly affects the molecular transport due to stronger interactions with the n-hexane molecules. A similar effect of Bronsted sites on the single component diffusion of aromatics was observed in MFI zeolites with different concentration of acid sites [63-65]. The frequency response (FR) technique provided similar results... 

Summarizing, we conclude that for binary mixtures of a linear and branched hexane in H-ZSM-5 and silicalite-1 two factors influence the respective diffusivities (i) the strong interaction with acid sites preferentially decreases n-hexane diffusivity and (ii) the blocking of intersection adsorption sites by 2-methylpentane decreases n-hexane diffusivity. At high loadings of the branched isomer the latter effect is dominating, and Anally the diffusivity of the linear hexane is totally determined by its branched isomer. 

A comparison between sihcalite-1 and H-ZSM-5 teaches that acid sites have a profound influence on the self-diffusivity of alkanes. The self-diffusivities of both components decrease strongly, and we observe a significant preferential adsorption of the linear over the branched hexane. This is caused by the relatively stronger interaction of the linear hexane with the acid sites. On the contrary, 2-methylpentane loadings in mixtures in sihcahte-1 and H-ZSM-5 are very close. In H-ZSM-5, the diffusivity of the hnear alkane in mixtures with the branched alkane is influenced by two factors... 

Hexane isomers and benzene-p-xylene mixtures have been separated by PV through an MFl-type zeolite membrane by Matsufuji et al. (2000). The PV tests for n-hexane, 2-methylpentane (2-MP), and 2,3-dimethylbutane (2,3-DMB) were performed using an MFl-type zeolite membrane at 303 K. -Hexane preferentially permeated through the MFI membrane in the PV tests for binary mixtures of n-hexane-2-MP and n-hexane-2,3-DMB. The separation factors (a( -hexane-2-MP) and a(n-hexane-2,3-DMB)) were always greater than their ideal selectivi-ties. The ideal selectivities of n-hexane-2-MP and -hexane-2,3-DMB were 37 and 50, respectively. It was observed that separation factor a( -hexane/2,3-DMB) was as high as 270 when the feed concentration of n-hexane was 10 mol%. Matsufuji et al. (2000) claimed that the MFI membranes have promising potential to separate n-hexane and branched hexane-isomer mixtures. 

Huddersman, K., and Klimczyk, M., Separation of branched hexane isomers using zeolite molecular sieves, AlChE J., 42(2), 405-408 (1996). 

---