Furthermore, the amount of these Al sites can be tuned, leading to materials with a
different Si/Al ratio. Milestone and Bibby studied the effect of the Si/Al ratio on the
n-butanol adsorption capacity, measuring single solute isotherms of alcohols in
water, using H-ZSM-5 with Si/Al ratios of 24, 46, and 99, and also studied the
effect of cations on the adsorption equilibria of linear C1–C4 alcohols [104]. They
observed that an increase in cation size leads to a decrease in saturation capacity for
n-butanol, linked to the decrease in accessible pore volume. The variation of Si/Al
ratio led to an interesting effect on the amount adsorbed of alcohols at low concentration. For n-alcohols containing a long aliphatic tail (i.e., propanol and n-butanol),
the amount adsorbed was observed to increase with increasing Si/Al ratio. However,
for the shorter n-alcohols, ethanol, and methanol, a decrease in saturation capacity
with higher Si/Al was observed. These experimental results of Milestone and Bibby
were confirmed by GC-MC simulations carried out by Xiong et al.; however no
interpretation was given [155]. Nguyen et al. studied the adsorption complexes of
C 1 –C 4 alcohols in H-ZSM-5 using DFT, with the alcohols and Bronsted acidic sites
positioned in the intersection between the zigzag and straight channels
[157]. Depending on whether the alcohol was positioned in the straight or zigzag
channel and the location of the acidic site (at different T-sites in the channel
intersection), the contribution of the dispersive Van der Waals interactions increased
from 30% for methanol to 50% for n-butanol [157]. Thus, for smaller alcohols,
hydrogen bonding of the alcohol group with the Bronsted acid site appears to be
dominant, whereas for larger alcohols, Van der Waals interactions of the aliphatic
tail with the framework increase in importance.
A difference exists between the single solute adsorption isotherms of the butanol
isomers containing branched methyl groups (isobutanol and t-butanol) compared to
n-butanol and 2-butanol on silicalite-1: it can be observed that the adsorbed amount
for the branched alcohol is always lower than that of the linear alcohol (Fig. 4)
[104]. To assess the possible difference in affinity for linear and branched alcohols,
Thamm et al. determined the differential heat of adsorption (via calorimetry) of
isobutanol and n-butanol. They observed the heat of adsorption of isobutanol to be
10–20 kJ/mol lower than that of n-butanol, indicating a weaker interaction of
isobutanol with the framework [142]. These results were confirmed by Xiong et al.
via molecular simulations using the GC-MC method; however the authors do not
discuss the distribution of these branched chain alcohols in detail [155]. Nguyen
et al. studied the physisorption of different butanol isomers using DFT calculations,
looking at the interaction of one single butanol molecule with the MFI framework
[158]. The authors report a decrease in the physisorption strength from n-butanol to
t-butanol: the interaction energy decreased with 15 kJ/mol for the straight channel
and with 63 kJ/mol for the zigzag channel for t-butanol compared to n-butanol. Due
to the smaller size of the zigzag channels (Fig. 3), repulsion effects are more
pronounced in this part of the structure. Similar results have been reported for the
adsorption of linear and branched alkanes on ZSM-5, where the vapor phase Henry
constants for linear alkanes are typically higher than the corresponding branched
isomers [159]. These observations are important, since they indicate that MFI
Efficient Downstream Processing of Renewable Alcohols Using Zeolite Adsorbents
93
different Si/Al ratio. Milestone and Bibby studied the effect of the Si/Al ratio on the
n-butanol adsorption capacity, measuring single solute isotherms of alcohols in
water, using H-ZSM-5 with Si/Al ratios of 24, 46, and 99, and also studied the
effect of cations on the adsorption equilibria of linear C1–C4 alcohols [104]. They
observed that an increase in cation size leads to a decrease in saturation capacity for
n-butanol, linked to the decrease in accessible pore volume. The variation of Si/Al
ratio led to an interesting effect on the amount adsorbed of alcohols at low concentration. For n-alcohols containing a long aliphatic tail (i.e., propanol and n-butanol),
the amount adsorbed was observed to increase with increasing Si/Al ratio. However,
for the shorter n-alcohols, ethanol, and methanol, a decrease in saturation capacity
with higher Si/Al was observed. These experimental results of Milestone and Bibby
were confirmed by GC-MC simulations carried out by Xiong et al.; however no
interpretation was given [155]. Nguyen et al. studied the adsorption complexes of
C 1 –C 4 alcohols in H-ZSM-5 using DFT, with the alcohols and Bronsted acidic sites
positioned in the intersection between the zigzag and straight channels
[157]. Depending on whether the alcohol was positioned in the straight or zigzag
channel and the location of the acidic site (at different T-sites in the channel
intersection), the contribution of the dispersive Van der Waals interactions increased
from 30% for methanol to 50% for n-butanol [157]. Thus, for smaller alcohols,
hydrogen bonding of the alcohol group with the Bronsted acid site appears to be
dominant, whereas for larger alcohols, Van der Waals interactions of the aliphatic
tail with the framework increase in importance.
A difference exists between the single solute adsorption isotherms of the butanol
isomers containing branched methyl groups (isobutanol and t-butanol) compared to
n-butanol and 2-butanol on silicalite-1: it can be observed that the adsorbed amount
for the branched alcohol is always lower than that of the linear alcohol (Fig. 4)
[104]. To assess the possible difference in affinity for linear and branched alcohols,
Thamm et al. determined the differential heat of adsorption (via calorimetry) of
isobutanol and n-butanol. They observed the heat of adsorption of isobutanol to be
10–20 kJ/mol lower than that of n-butanol, indicating a weaker interaction of
isobutanol with the framework [142]. These results were confirmed by Xiong et al.
via molecular simulations using the GC-MC method; however the authors do not
discuss the distribution of these branched chain alcohols in detail [155]. Nguyen
et al. studied the physisorption of different butanol isomers using DFT calculations,
looking at the interaction of one single butanol molecule with the MFI framework
[158]. The authors report a decrease in the physisorption strength from n-butanol to
t-butanol: the interaction energy decreased with 15 kJ/mol for the straight channel
and with 63 kJ/mol for the zigzag channel for t-butanol compared to n-butanol. Due
to the smaller size of the zigzag channels (Fig. 3), repulsion effects are more
pronounced in this part of the structure. Similar results have been reported for the
adsorption of linear and branched alkanes on ZSM-5, where the vapor phase Henry
constants for linear alkanes are typically higher than the corresponding branched
isomers [159]. These observations are important, since they indicate that MFI
Efficient Downstream Processing of Renewable Alcohols Using Zeolite Adsorbents
93