S.D. Blass et al. / Catalysis Communications 42 (2013) 54–57
55
(Si/Al = 11.5), CBV 3024E (Si/Al = 15), CBV 5524G (Si/Al = 25), and
3. Results and discussion
CBV 8014 (Si/Al = 40) zeolite samples.
3.1. Reactivity of butanone over HZSM-5
2.2. Reactor configuration and analysis
Butanone was initially unreactive over HZSM-5 without Pt/γ-Al2O3
at temperatures of up to 250 °C with conversions less than 1%. Similarly,
Chang et al. reacted acetone over HZSM-5 and reported a conversion of
3.9% at 250 °C and a LHSV of 8.0 h−1 [15]. While acetone and butanone
have been shown to be reactive over HZSM-5 to form C2–5+ products at
temperatures exceeding 370 °C through aldol condensation and crack-
ing [16], the temperatures in our reactor may be too low for this to
occur. The extent of bimolecular reactions is a function of partial pres-
sure, which relates to surface coverage and temperature [4,17]. Higher
temperatures are needed at lower surface coverages to initiate an
aldol condensation reaction. Aldol condensations of acetone are typical-
ly done in the vapor phase from 250 to 400 °C [3]. The partial pressure of
butanone used in our experimental conditions (0.83 kPa) may have
been too low to achieve the minimum surface coverage needed to initi-
ate significant bimolecular reactions such as aldol condensations,
explaining the low observed reactivity.
A 10 mm ID quartz tube was fabricated and connected to a vaporizer
consisting of a 1/4″ stainless steel tube heated to 200 °C with heating
tape as shown in Fig. 1. The vaporizer inlet was sealed to a stainless
steel nebulizer consisting of concentric 1/8″ and 1/16″ tubes. Butanone
was fed through the nebulizer at 0.042 ml min−1 and N2 and H2 were
fed through the nebulizer each at 0.70 standard liters per minute
(slpm) for a H2 to butanone molar ratio of 60.
Composition analysis was conducted with an online HP 7890A gas
chromatograph (GC) equipped with a thermal conductivity detector
and flame ionization detector in series. The GC contained a HP PLOT/Q
column (30 × 0.32 × 20). N2 was used as an internal standard. Peak
identification was verified with a HP 5890 Series II gas chromatograph
containing a HP-1 column (50 × 0.32 × 1.05) connected in series
with a HP MSD 5970 mass selective detector. Carbon balance closed to
within 15%.
3.2. Evidence for series reaction over Pt/γ-Al2O3 and HZSM-5
2.3. Experimental procedure
While butanone was inactive over HZSM-5 alone, the addition of
Pt/γ-Al2O3 to the zeolite led to yields of butane up to 67% with less
than 2% selectivity to butene and 2-butanol. The catalyst loading was
varied at each temperature to change space velocity and conversion.
Only trace amounts of ethylene and propylene were observed, indicat-
ing that cracking reactions were negligible. Less than 1% selectivity to
products larger than C4 was observed, possibly because the partial pres-
sure of butanone was low enough to limit bimolecular reactions. When
product selectivities were plotted as a function of conversion, as shown
in Fig. 2, butane selectivity decreased as conversion decreased while bu-
tene increased. Butanol selectivity decreased from 15% to 10% at 100 °C
as conversion increased from 2 to 60%. Increasing selectivity of butanol
and butene at low conversions suggests that these species are interme-
diates in a series reaction that ultimately converts butanone to butane.
In a series reaction that can be explained with Fig. 2, we propose that
butanone was reduced over a Pt site to form 2-butanol. The butanol
likely diffused to a zeolite acid site and subsequently dehydrated to an
olefin. The olefin likely diffused back to a Pt site where the C_C bond
was hydrogenated to form a paraffin compound. No butanol was ob-
served at temperatures greater than 160 °C, possibly because it was
consumed rapidly by the dehydration reaction. When compared across
different temperatures at similar conversions, butane selectivity in-
creased with temperature while butanol selectivity decreased, likely be-
cause zeolite-catalyzed dehydration rates increased with temperature
[14], limiting the concentration of butanol. Deactivation was also ob-
served as shown by a decrease in butanone conversion by 5% at
100 °C and 20% at 160 °C at the same time-on-stream. It is possible
that coke formation may occur more readily at 160 °C causing greater
changes in conversion than at 100 °C. Selectivities did not change
with time-on-stream by more than 5% at any loading.
The catalyst was pretreated at 500 °C for 20 min in dry air at
0.3 ml min−1 and 20 min in H2 at 0.70 slpm. A constant N2 flow rate
was maintained at 0.70 slpm during pretreatment and throughout the
course of the experiments. After pretreatment, the reactor was cooled
to the desired temperature while maintaining constant flow rates.
Three injections were analyzed at a time-on-stream of 7, 22, and
37 min. The catalyst temperature was varied from 100 to 250 °C. The
catalyst was regenerated under the pretreatment conditions between
each temperature. Space velocity was changed by varying the total cat-
alyst loading from 12 to 400 mg at a constant reactant flow rate. Data at
all loadings are plotted without consideration of bypass, which occurred
at loadings below 30 mg, because the conversion did not exceed 10%.
The WHSV was varied from 27 to 850 h−1 or a residence time of 3–
33 ms calculated on a gas hourly basis. Besides the bifunctional catalyst,
monofunctional catalysts consisting of solely HZSM-5 or Pt/γ-Al2O3
were synthesized using the methods described above.
3.3. Reactivity of butanone over Pt/γ-Al2O3
When butanone was reacted over Pt/γ-Al2O3, the butane selectivity
increased with temperature while butanol selectivity decreased, as
shown in Fig. 3. As the temperature was increased, dehydration cata-
lyzed by Lewis acid sites on γ-Al2O3 became important. γ-Al2O3 con-
taining Lewis acid sites has been shown to be an active dehydration
catalyst for isobutanol from 160 to 300 °C [18]. Lewis acid-catalyzed
dehydration likely explains the decrease in butanol selectivity and sub-
sequent increase in butane selectivity. When Pt/γ-Al2O3 was combined
with HZSM-5, both Brønsted and Lewis acid sites likely carried out
dehydration.
Fig. 1. Reactor schematic.