being adsorbed and the ethanol concentration profile showing a roll-up. The slightly
larger cages and windows of ITQ-29 allow the adsorption of n-butanol. Competitive
adsorption leads to the displacement of ethanol in dynamic conditions, while acetone
is excluded due to its molecular size [59]. The effect of water on the selectivity of
ITQ-29 for biobutanol over ethanol and water was studied by performing ternary
ethanol/n-butanol/water separations. In this case, an increase in ethanol adsorption
and thus a decrease in selectivity for n-butanol with increasing water partial pressure
were observed. This increase in ethanol capacity with increasing water pressure
might be caused by water molecules serving as a seed for ethanol adsorption, as is
also observed the case of ZSM-5 and silicalite-1 (vide supra).
In more recent work, Miyamoto et al. synthesized core-shell structured chabazite
zeolite crystals [102]. Native aluminosilicate crystals with a Si/Al ratio of 39 were
used as a seed for external growth of an all-silica chabazite layer. From an equilibrium perspective, both the native crystals and coated zeolite crystals showed similar
n-butanol, ethanol, and water isotherms. However, a clear difference in uptake
kinetics was observed for ethanol, with the uptake of ethanol on the native zeolite
sample being much faster compared to the coated crystal. This slower diffusion of
ethanol into the coated chabazite crystals is also visible in the presented dynamic
ABE mixture separations, with the ethanol breakthrough profiles being significantly
broader compared to the native material. The effect of temperature on the separation
was also investigated, with the authors showing a slight increase in n-butanol
adsorption with higher temperature, possibly related to an increase in diffusivity of
n-butanol with higher experimental temperature.
3.5 Combining the Selectivity of CHA and LTA Zeolites
As shown in the previous paragraphs, the LTA-type ITQ-29 zeolite shows a larger
selectivity for n-butanol, compared to the other ABE mixture components. In
contrast, chabazite-type zeolites have a larger selectivity for ethanol and/or water
and exclude acetone and n-butanol, due to very slow diffusion of these components
into the material. Therefore, Van der Perre et al. proposed a dual-column process,
exploiting the selectivity of both types of materials to fully purify n-butanol (Fig. 11)
[59]. In a first step, a column packed with ITQ-29 pellets was used to adsorb
n-butanol from an ABE vapor mixture. Small amounts of ethanol and water
co-adsorb with n-butanol, so hence, the desorbing stream from the ITQ-29 column
was fed to a column packed with a chabazite-type zeolite (all-silica or SAPO-34),
allowing the removal of ethanol and water from the stream. When only using ITQ-29
as a single column, the final purity of the n-butanol was only 65.5 mol-% at a
recovery of more than 99.5%. By employing this dual-column process, the purity of
the produced n-butanol could be increased to 99.5 mol% at a recovery of more than
99.5%.
106
B. Claessens et al.
larger cages and windows of ITQ-29 allow the adsorption of n-butanol. Competitive
adsorption leads to the displacement of ethanol in dynamic conditions, while acetone
is excluded due to its molecular size [59]. The effect of water on the selectivity of
ITQ-29 for biobutanol over ethanol and water was studied by performing ternary
ethanol/n-butanol/water separations. In this case, an increase in ethanol adsorption
and thus a decrease in selectivity for n-butanol with increasing water partial pressure
were observed. This increase in ethanol capacity with increasing water pressure
might be caused by water molecules serving as a seed for ethanol adsorption, as is
also observed the case of ZSM-5 and silicalite-1 (vide supra).
In more recent work, Miyamoto et al. synthesized core-shell structured chabazite
zeolite crystals [102]. Native aluminosilicate crystals with a Si/Al ratio of 39 were
used as a seed for external growth of an all-silica chabazite layer. From an equilibrium perspective, both the native crystals and coated zeolite crystals showed similar
n-butanol, ethanol, and water isotherms. However, a clear difference in uptake
kinetics was observed for ethanol, with the uptake of ethanol on the native zeolite
sample being much faster compared to the coated crystal. This slower diffusion of
ethanol into the coated chabazite crystals is also visible in the presented dynamic
ABE mixture separations, with the ethanol breakthrough profiles being significantly
broader compared to the native material. The effect of temperature on the separation
was also investigated, with the authors showing a slight increase in n-butanol
adsorption with higher temperature, possibly related to an increase in diffusivity of
n-butanol with higher experimental temperature.
3.5 Combining the Selectivity of CHA and LTA Zeolites
As shown in the previous paragraphs, the LTA-type ITQ-29 zeolite shows a larger
selectivity for n-butanol, compared to the other ABE mixture components. In
contrast, chabazite-type zeolites have a larger selectivity for ethanol and/or water
and exclude acetone and n-butanol, due to very slow diffusion of these components
into the material. Therefore, Van der Perre et al. proposed a dual-column process,
exploiting the selectivity of both types of materials to fully purify n-butanol (Fig. 11)
[59]. In a first step, a column packed with ITQ-29 pellets was used to adsorb
n-butanol from an ABE vapor mixture. Small amounts of ethanol and water
co-adsorb with n-butanol, so hence, the desorbing stream from the ITQ-29 column
was fed to a column packed with a chabazite-type zeolite (all-silica or SAPO-34),
allowing the removal of ethanol and water from the stream. When only using ITQ-29
as a single column, the final purity of the n-butanol was only 65.5 mol-% at a
recovery of more than 99.5%. By employing this dual-column process, the purity of
the produced n-butanol could be increased to 99.5 mol% at a recovery of more than
99.5%.
106
B. Claessens et al.