Paper
Catalysis Science & Technology
−
1
Argon was also co-fed at a flow rate of 100 ml min at STP.
The triol evaporated and mixed with the argon inside the
preheating tube and the resulting vapor stream was then
flowed into a packed-bed flow reactor (12 mm OD steel tube,
10 wt% solutions of THP-2-M and 1,6-hexanediol in water
were used as reagents following the same flow reactor proce-
dure outlined above. The reactions were performed at 573 K,
using the same molar flow of reagent as was used in the
1,2,6-hexanetriol runs. 133 mg of H-ZSM5 (80/1 Si/Al ratio)
was used as the catalyst.
1
5 cm long bed). The bed was packed with a measured quan-
tity of catalyst diluted in 1 mm glass beads. Temperatures
were monitored at the end of the preheating tube as well as
the entrance and exit of the packed bed. All three tempera- Results
tures were maintained within 2 K of the set point temperature.
Catalyst properties
The effluent stream from the reactor passed into a condenser,
where the products condensed on the inner walls and the con-
densate then dripped into a collection vial. The residence time
of the condenser was held at 1 min, as shorter residence times
led to insufficient condensation. The argon in the effluent
stream exited the condenser through an exit line at the top of
the condenser and the liquid products were collected in a vial
at the bottom of the condenser.
The product samples were highly viscous and required a
relatively long time to exit the condenser, so sample blending
was minimized by operating the reactor continuously. In a
given experiment, 5 product samples were taken for each
catalyst with the samples prepared for analysis by diluting to
The catalyst slate was selected to provide sufficient variance
in terms of acid strength and pore size as to examine the
importance of these factors. The measured properties of the
catalysts are given in Table 1. The acid strengths for the fresh
catalysts, as measured by ammonia TPD, spanned the typical
15–17
range that has been commonly reported in the literature.
Per previous reports, weak acid sites were attributed to the
temperature range from 423 to 523 K, and the strong acid
sites to peaks ranging from 573 to 773 K. In terms of shape
selectivity, the average pore diameters for the zeolites and
non-zeolites varied by an order of magnitude.
Reaction testing
1
0 wt% in water (0.9000 grams water, 0.1000 grams sample)
and adding 10 μL of methanol as an internal standard. Once
prepared, the samples were analyzed in an Agilent 7890 gas
chromatogram equipped with a flame ionization detector
and mass spectrometer. Product molecules not readily identi-
fied, i.e., those with less than 85% statistical match in the
NIST mass spectral libraries, were interpreted manually and
the interpretation is given in the ESI.†
All experiments were replicated at least once to measure
experimental consistency. Only replicated experiments in which
the overall mass balance was greater than 90% were included
in the published results. The reason for mass loss was related
to catalyst coking during the reaction (~1% loss) or to losses in
the reagent inlet line (3–5% loss). Those discarded runes in
which the mass balance was below 90% could be attributed to
vapor losses in the condenser, which made the selectivity
results unreliable due to losses of volatile compounds.
Deactivation of the catalysts due to coking was commonly
observed during the study, with much of the deactivation
occurring during the beginning of an experimental run. To
account for this, data for the first sample (the deactivation
sample) of a given experiment was treated separately from
the remaining four samples (steady-state samples) in an
experiment, which were averaged to give the reported values.
Each of the catalysts were tested according to the above
flow reaction procedure at 573 K. The weight of catalyst
added was based on a constant number of high-strength acid
sites, which was determined by multiplying catalyst mass by
acid site density (Table 1) to get the number of acid sites for
that sample. 50 mg of H-ZSM5, 30/1 Si/Al ratio (0.024 mmol
of high-strength acid sites) was used as the basis for deter-
mining the required mass loading for the other catalysts, as
this catalyst had the highest acid site density amongst the
catalysts tested.
Catalyst activity was measured by quantifying the amount
of 1,2,6-hexanetriol that was consumed and appeared as
GC-identifiable products. Losses attributed to deposition of
the triol in transfer lines were excluded from the overall reac-
tion rate. GC analysis of the non-condensable gases found
that cracking of the triol to smaller molecules was negligible
and losses due to catalyst coking accounted for less than 1%
of the reaction products.
The turnover frequency of each catalyst is reported in
Fig. 1 as a function of TPD peak temperature. All of the cata-
lysts deactivated significantly from the start of the experi-
ment until a relatively stable operation was achieved. The
decrease in activity was as much as an order of magnitude.
Overall, the acid strength and activity were only weakly corre-
lated after the first sample point (still in the deactivation
region) and the correlation remained weak for the remaining
samples. TPD analysis of the used catalysts showed that both
the strong and weak acid sites were significantly diminished
and in most cases the strong acid sites appeared to have been
completely suppressed for the catalysts. As an example, the
Fig. 1 Activity of acid catalysts, as measured by turnover (moles
converted per acid site per second). Activity was measured both for
the first sample taken (■), and for the four following samples (bars).
The bars denote the complete range of observed activities.
2262 | Catal. Sci. Technol., 2014, 4, 2260–2266
This journal is © The Royal Society of Chemistry 2014
Nolan, Michael R.
