The effect of steam on the catalytic fast pyrolysis of cellulose

Article abstract of DOI:10.1039/c5gc00026b

The effect of steam for the catalytic fast pyrolysis of cellulose with ZSM-5 was studied in a bubbling fluidized bed reactor. Irreversible and reversible changes due to steaming were identified. Steam caused dealumination, a loss of total acidity, an increase in the zeolite-crystal size, and agglomeration of particles. For runs both with and without steam co-feeding, these irreversible changes caused lower yields for aromatics and char/coke, and higher yields for methane and unidentified products. In addition to irreversible catalyst changes, steam co-feeding was also found to reversibly lower yields of aromatics, char/coke, and identifiable oxygenate species, increase yields of CO and methane, and not change the overall yields of CO₂ and olefins.

Full text of DOI:10.1039/c5gc00026b

Green Chemistry
View Article Online
View Journal
PAPER
The effect of steam on the catalytic fast pyrolysis
of cellulose
Cite this: DOI: 10.1039/c5gc00026b
H. Yang,a,b R. J. Coolman,b P. Karanjkar,b H. Wang,b Z. Xu,b H. Chen,a
T. J. Moutziarisc and G. W. Huber*b
The effect of steam for the catalytic fast pyrolysis of cellulose with ZSM-5 was studied in a bubbling
fluidized bed reactor. Irreversible and reversible changes due to steaming were identified. Steam caused
dealumination, a loss of total acidity, an increase in the zeolite-crystal size, and agglomeration of particles.
For runs both with and without steam co-feeding, these irreversible changes caused lower yields for aro-
matics and char/coke, and higher yields for methane and unidentified products. In addition to irreversible
catalyst changes, steam co-feeding was also found to reversibly lower yields of aromatics, char/coke, and
identifiable oxygenate species, increase yields of CO and methane, and not change the overall yields of
CO2 and olefins.
Received 5th January 2015,
Accepted 9th March 2015
DOI: 10.1039/c5gc00026b
www.rsc.org/greenchem
(450 °C) for benzofuran and char/coke.11 Carlson et al.
obtained an aromatic yield of 14% C from pine sawdust in a
Introduction
Catalytic fast pyrolysis (CFP) is a technology for the production bench-scale fluidized-bed reactor running at low biomass
of monocyclic aromatics from solid biomass.1 In a single step, weight hourly space velocities (WHSV < 0.5 h−1) and high
up to 30% of the carbon contained within solid biomass is temperature (600 °C).2 Karanjkar et al. obtained an aromatic
converted into aromatics including benzene, toluene, and yield of 39.5% C from cellulose feed at similarly low space velo-
xylenes (BTX). To perform CFP, biomass is fed into a fluidized city (0.25 h−1), but at lower temperature (500 °C), indicating
bed of spray-dried zeolite catalyst where it thermally decom- that importance of cellulose over the other components of
poses to form pyrolysis vapors. The pyrolysis vapors enter the biomass.5 This increase is consistent with earlier results from
zeolite’s pores forming aromatics and olefins along with CO, Carlson et al. which indicated that choosing biomass feed-
CO2, H2O and char/coke. The solid char/coke accumulates on stocks with high amounts of cellulose and hemicellulose will
the catalyst surface and deactivates the catalyst, so it is necess- increase aromatic yield.2
ary to send charred/coked catalyst to a furnace where the char/
Jae et al. carried out CFP of sawdust in a process develop-
coke is combusted to provide process heat. The advantage of ment unit (PDU) consisting of a bubbling fluidized-bed reactor
CFP is that pyrolysis and catalysis occur in a single reactor, with continuous catalyst addition and removal, and was
which greatly reduces process cost. Furthermore, CFP produces capable of maintaining constant product yield of aromatics
commodity chemical products (aromatics and olefins) that (15.5% C) over an extended reaction period (6 h).4 Paasikallio
already fit into existing infrastructure.1–9
et al. performed CFP of pine sawdust in a 20 kg h−1 PDU riser
Previous research has focused on optimizing CFP reactor reactor over a four-day run, and reported a pyrolysis-oil yield
conditions,2,5,10–14 optimizing reactor configurations,2,7,15,16 of 24% C.9
screening and design of improved catalysts,3,17–23 investigating
In contrast to fossil fuels, biomass feedstocks possess high
feedstocks,24–26 and determining the effects of inorganic contents of both moisture and oxygen.9,30,31 Drying and de-
contaminants.27–29 High temperatures (650 °C) are favorable hydration reactions produce large amounts of steam during CFP,
for the production of olefins and CO, intermediate tempera- much of which condenses as water in the liquid-phase pro-
tures (450–600 °C) for aromatics, and low temperatures ducts.32 Little is known about the effect of this steam on CFP
chemistry. The presence of steam during un-catalyzed biomass
pyrolysis is understood to change gas, char/coke, and liquid
aState Key Laboratory of Coal Combustion, Huazhong University of Science and
yields and promote the removal of oxygen from the liquid.33
Technology, Wuhan, 430074, China
bDepartment of Chemical and Biological Engineering, University of Wisconsin-
Williams et al. found that the presence of steam during
zeolite-catalyzed upgrading of oxygenated biomass pyrolysis
vapors led to the formation of large amounts of gaseous pro-
ducts with a yield in excess of 70% at 550 °C, proposing that
Madison, Madison, WI 53706, USA. E-mail: huber@engr.wisc.edu
cDepartment of Chemical and Biological Engineering, University of Massachusetts-
Amherst, Amherst, MA 01003, USA
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
water can act as H2 donor.34 Gilbert et al. studied the influence grams of catalyst were loaded into the reactor; a loading corres-
of water on the conversion of furan over ZSM-5 and found that ponding to roughly 15% of the reactor volume. Prior to reac-
water hydrolyzes furans to produce more propylene and CO2.35 tion, the catalyst was calcined in situ at 600 °C in air flowing at
In CFP, steam may also affect catalyst integrity. Iliopoulou 600 sccm.
et al. found that treatment of mesoporous Al-MCM-41 at
Reactor setup
550 °C under a 20% steam atmosphere caused a significant
reduction in acidity and porosity.36 Ong et al. found that steam
CFP of cellulose was performed in a fluidized-bed reactor
treatment of ZSM-5 for 5 h at 450 °C under a pure steam
atmosphere caused irreversible dealumination through stable
tetrahedrally-coordinated extra-framework aluminum neutra-
lizing aluminum/oxygen tetrahedra in the lattice, thus redu-
cing the number of Brønsted acid sites.37 Corma et al. studied
the steam catalytic cracking of naphtha over ZSM-5 at 650 °C
with a steam vapor fraction between 0% and 35.7% and found
their catalyst underwent intense dealumination, cracking rates
decreased after steam exposure times of 600 s, and these
effects are more apparent at higher temperatures.38
In addition to steam causing irreversible dealumination,
previous research has demonstrated reversible deactivation as
it can compete with reactant species (specifically n-heptane)
for adsorption on active sites, thus negatively affecting the cat-
alyst’s activity.39 Corma et al. found that post-synthesis treat-
ment with phosphorous, introduced to enhance the
hydrothermal stability, slowed down catalyst deactivation for
steam catalytic cracking.40 Lee et al. found that the phos-
phorus-treated HZSM-5 (P/ZSM-5) exhibited lower acidity,
higher hydrothermal stability and improved dimethyl ether
(DME) selectivity in methanol conversion compared to the
phosphorus-free HZSM-5.41
system, as shown in Fig. 1. The fluidized bed reactor is a 316L
stainless-steel 4.92 cm ID pipe with a freeboard height of
37 cm. Above the freeboard is a disengaging zone which
expands to a 7.79 cm ID pipe. The catalyst bed is supported by
a distributor plate made from two layers of 304 stainless-steel
cloth (200 mesh) glued to a stainless-steel screen for support.
The main reactor body and space beneath the distributor plate
(henceforth referred to as the plenum) were sealed together
using bolted flanges. The interior of the reactor and the
plenum were spray-coated with a protective layer of abrasion-
resistant ceramic to protect the reactor from corrosion.
The reactor and plenum were resistively heated using semi-
circular ceramic heaters (WATLOW). The flanges joining the
main reactor body and plenum were heated using a band
heater. Heating-zone temperatures were controlled by thermo-
couples located between the reactor body and heaters.
A typical run was carried out for 30 min time on stream.
The catalyst was fluidized using helium and/or steam at a rate
of 600 sccm (henceforth referred to as the fluidizer gas).
Heating of the plenum allowed for pre-heating of the fluidizer
gas to the reaction temperature before reaching the catalyst.
Steam was delivered to the reactor using a syringe pump and
in-line steam kettle. A stainless steel auger equipped with a
variable speed motor and rotary fitting fed cellulose into the
side of the reactor. Cellulose was supplied to this auger via a
sealed feed hopper (calibrated prior via balance and stop-
watch). To maintain an inert environment and encourage the
Zeolite catalysts used in fluidized beds are spray dried with
a binder in order to obtain larger and more attrition-resistant
particles with other desirable fluidization characteristics.42
The presence of binder can affect catalyst behavior through (1)
changes in the proton-exchange efficiency, (2) blocking of the
zeolite channels, and (3) channels in the binder trapping coke
precursors which cause a significant decrease in the amount
of coke deposited on the zeolite.43,44
The objective of this paper is to elucidate the separate
effects of (1) the long-term influences of steam’s irreversible
effects on the catalyst during CFP and how this affects product
distribution, and (2) the interaction steam itself has with chan-
ging the product distribution through potentially acting as a
homogeneous catalyst, a reactant species, or a heterogeneous
catalyst site blocker.
Experimental
Materials
Industrial grade cellulose with an average particle size of 200
µm (Lattice® NT Microcrystalline Cellulose, FMC biopolymer,
99%) was used as the feedstock for this study. In all our calcu-
lations, the empirical formula C6H10O5 was used for cellulose.
The catalyst used in these experiments was a commercial
spray-dried 40% ZSM-5 catalyst (Intercat Inc.) with a particle-
Fig. 1 Flow diagram of bubbling fluidized bed for catalytic fast pyrolysis
size average of 99 µm and standard deviation of 23 µm. Ninety (CFP).
Green Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
rapid delivery of biomass to the reactor, the hopper and augur lyst was completely dissolved in 2 ml of hydrofluoric acid
were swept with helium at a rate of 400 sccm (henceforth (50%) and several drops of boric acid (7%) were added in order
referred to as the feeding gas). Gas flows were selected to to protect against the formation of volatile SiF4. The solution
operate the reactor in the bubbling fluidized-bed regime. was heated at 80–90 °C for seven minutes to allow some of the
During the reaction, product gases exited the top of the reactor hydrofluoric acid to evaporate before the solution was allowed
and were passed through a cyclone where entrained solids to cool. About 2 ml of nitric acid (10%) and several ml of boric
were removed as cyclonics. The solid-free vapors were then acid (7%) were added to the solution, then water was added to
bubbled through four condensers each containing ∼20 ml iso- bring the solution volume to 50 ml.
propanol maintained at 0 °C using an ice bath. Here, most
The acidity of the samples was measured using Tem-
organic species were captured through dissolution. The stream perature Programmed Desorption (TPD) of isopropylamine
was then passed through four condensers maintained at (IPA-TPD) and ammonia (NH3-TPD) using a Micromeritics®
−55 °C using a dry-ice/acetone bath to condense remaining Autochem II 2920 with an inline Thermal Conductivity Detec-
organics. The non-condensable gases were then either vented, tor (TCD). Before testing, the catalyst was heated up to 600 °C
plumbed through a bubble flow meter, or sampled in Tedlar for 2 h in helium to remove adsorbed water or organic species.
gas bags (Restek) for GC analysis. At the conclusion of cellu- For IPA-TPD, once the 100 mg sample was saturated with IPA
lose feeding, the reactor was purged with 1000 sccm of helium at 50 °C for 20 min, helium was flushed at 50 sccm for 2 h in
for another 30 min to ensure a complete purge of all volatile the temperature range of 50–700 °C at a heating rate of 10 °C
organic products. The condensers were then removed and min−1. For NH3-TPD, once the 100 mg sample was saturated
rinsed with isopropanol to collect all product liquids. The with NH3 at 100 °C for 30 min, helium was flushed at 12 sccm
volume of isopropanol/product solution was recorded and ana- for 2 h in the temperature range of 150–700 °C at a heating
lyzed via GC to quantify products.
rate of 5 °C min−1. The number of Brønsted acid sites was calcu-
After feeding and purge, the reactor temperature was lated based on the TCD signal for NH3 and propylene, the pro-
increased to 600 °C, and the carrier gas was switched to air to ducts of IPA decomposition. The total number of acid sites was
combust char/coke and regenerate the catalyst. For a typical calculated based on the TCD signal for NH3. The number of
run, the catalyst was regenerated for approximately two hours Lewis acid sites was taken as the difference between the total
to ensure complete combustion of any organic species remain- number of acid sites and the number of Brønsted acid sites.
ing on the catalyst.
The zeolite crystallinity was determined by a Bruker D8
Discover diffractometer using CuKa radiation (l = 1.542 Å). The
scattering angle 2θ was varied from 5° to 60°. The crystal size
Product analysis
During catalyst regeneration, the combustion effluent contain- was calculated using the Scherrer equation.
ing CO, CO2, and water was passed through a copper catalyst
The BET surface area and pore volume of the samples were
(13 wt% CuO on alumina, Sigma Aldrich) held at 250 °C to calculated by N2 adsorption/desorption isotherms obtained at
convert CO to CO2. This stream was then passed through a −196 °C using a Micromeritics, ASAP2020 adsorption analyzer.
Dryrite trap to remove water followed by a pre-weighed Ascarite Prior to N2 adsorption, the samples were degassed at 300 °C
trap to capture CO2. Weights of the Ascarite trap before and overnight under vacuum.
after catalyst regeneration were used to determine the quantity
of char/coke generated from the reaction.
The particle size distribution and surface morphology of
catalyst was measured with a scanning electron microscope
The liquid product dissolved in isopropanol was analyzed (SEM, LEO 1550 VP).
for aromatics using a Shimadzu GC2010 system with an
Agilent HP INNOWax column (60 m, 0.32 mm, 0.5 µm) and a
flame ionization detector (FID). Column max temperature:
260 °C. Carrier gas: He. Injection mode: split ratio of 10.
Temperature Procedure: Initial temperature 70 °C, hold time
10 min, then heated up to 95 °C at 2 °C min−1, then heated up
240 °C at 15 °C and hold up for 10 min.
Non-condensable gases collected in gas bags at various
times during the reaction were analyzed using refinery gas ana-
lyzer Shimadzu GC2014 system with (1) RestekRtx (RTX) –
Alumina column and a flame-ionization (FID) detector to
analyze methane and C2–C5 olefins and (2) RTX-MS-5A column
and RTX-Q-plot column with a thermal-conductivity detector
(TCD) to analyze CO and CO2 respectively.
Definitions
The weight hourly space velocity (WHSV) of cellulose was cal-
culated by dividing the cellulose mass flow rate by the mass of
catalyst present inside the fluidized bed reactor as shown in
eqn (1). Selectivity towards a particular aromatic compound is
defined in eqn (2) by dividing the number of moles of carbon
in that aromatic product by the number of moles of carbon in
all the aromatic products. eqn (3) defines selectivity towards
olefin compounds in a similar way.
cellulose flow rate ðg hꢀ1
weight of catalyst ðgÞ
Þ
WHSVðhꢀ1Þ ¼
ð1Þ
moles of carbon in aromatic product
moles of carbon in all aromatic products
Aromatic selectivity ¼
Catalyst characterization
ꢁ 100%
A Perkin-Elmer-emission spectrometer plasma 400 was used to
do elemental analysis of the catalyst. A 100 mg sample of cata-
ð2Þ
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
moles of carbon in olefin product
moles of carbon in all olefin products
Olefin selectivity ¼
ꢁ 100%
ð3Þ
Method for estimation of an average bubble size and expanded
bed height
The assemblage model, as described by Karanjkar et al.5 was
used to determine the influence of steam on the hydro-
dynamic properties. Hydrodynamic properties such as bubble
size, bubble velocity, residence time, and density of fluidiza-
tion were calculated from inputs such as bed temperature, par-
ticle size, bed weight, output fluidizer flowrate, and input
fluidizer composition. The model was expanded to include
changes to gas density and viscosity based on its steam/
helium makeup.
Fig. 2 Carbon yield as function of input vapor fraction of steam for cat-
alytic fast pyrolysis of cellulose with fresh ZSM-5 catalyst. (Reaction con-
ditions: Cellulose feed, WHSV: 0.4 h−1, Temperature: 500 °C, time on
■
●
:
▲
◆
stream: 30 min, 90 g catalyst.) : Aromatics, : CO, : Char/coke,
□
▼
Olefins, : CO2, ★: Methane, *: Identifiable Oxygenates, : Unidentified
carbon.
furan, phenol, acetaldehyde, acetic acid, hydroxyacetone, and
5-hydroxymethyl furfural.
The aromatic yield decreased linearly from 28% C to 22% C
as input vapor fraction of steam increased from 0% to 24%,
but remained close to 22% C with further increases to steam
vapor fraction. Carbon monoxide showed a trend contrary to
aromatics, increasing with steam vapor fraction (perhaps as
H2O reacted with char/coke precursors) up to 24% steam. The
char/coke yield decreased from 15% C to 5% C as the steam
fraction increased from 0% to 40% and then it increased with
further steam vapor fraction. Olefins, methane, and identifi-
able oxygenate yields showed no obvious change with steam
vapor fraction. The yield of unidentified carbon not detected
in any of the product phases (calculated as the balance carbon
of the cellulose feed, presumed to be larger oxygenated com-
pounds) increased with increasing steam vapor fractions of
steam and reaching the maximum value at 27% C, and then
decreasing.
Results and discussion
Effect of steam vapor fraction on product yields and selectivity
To test the effect of steam’s presence on CFP of cellulose,
some fraction of the input carrier helium gas was replaced
with an equivalent molar flowrate of steam. The fraction of
steam was varied over a range of 0% to 60%. This upper
bound of 60% represents a fluidization gas completely made
up of steam (600 sccm) and a feed gas consisting of helium
(400 sccm). The hydrodynamic parameters are listed in
Table 1. Average bubble size, rise velocity and residence times
for each phase were affected more-so by outlet gas flow rate
than input composition. Regardless, all of these hydrodynamic
properties were predicted to remain relatively stable through-
out the range of vapor fraction tested.
The aromatics selectivity did not change with input vapor
fraction of steam, as shown in Fig. 3. Benzene and toluene
were the two main liquid products at these reaction con-
ditions, each with a selectivity of about 35% C. However, the
olefin selectivity did change with steam vapor fraction, as
shown in Fig. 4. Increasing the steam vapor fraction increased
the ethylene selectivity and decreased the propylene selectivity.
Fig. 2 shows the product carbon yields at different steam
vapor fractions in the range of 0–41%. A fresh catalyst was
loaded into the reactor prior to each reaction test. The main
products include aromatics, olefins, CH4, CO, CO2, char/coke,
and some light oxygenates. The aromatics consisted of
benzene, toluene, ethyl-benzene, xylenes, styrene, indene, and
naphthalene. The identifiable oxygenates consisted of benzo-
Table 1 Estimated average bubble size, bubble residence time, and expanded bed height as estimated from the Assemblage Model.5 Reaction con-
ditions: Cellulose feed, WHSV: 0.4 h−1 Temperature: 500 °C, time on stream: 30 min, 90 g catalyst. Fluidizer gas: He/Steam mix @ 600 sccm, Feed
Gas: He @ 400 sccm
Parameter
Definition
Result
Input steam fraction (%)
Outlet gas flowrate (sccm)
db.ave (cm)
Flow rate of steam/(Fluidizer gas + Feed gas)
Measured by volume flow meter
Bubble diameter, averaged over the height of the bed
Bubble rise velocity, averaged over the height of the bed
Fraction of the bed occupied by the bubble phase
Height of fluidized bed
0.00%
1608.57
0.83
33.02
11.61
8.55
5.00%
1666.67
0.85
33.46
11.98
8.59
24.00%
1715.42
0.87
33.83
12.28
8.62
40.00%
1437.71
0.80
32.04
10.79
8.47
60.00%
1479.46
0.81
32.30
11.01
8.49
ub.ave (cm s−1
δ (%)
)
Lf (cm)
Expansion (%)
τf (s)
τe (s)
% height increase from packed to fluidized bed
Average gas residence time in the fluidized bed
Gas residence time in the emulsion phase
Gas residence time in the bubble phase
41.41
1.20
19.18
0.26
42.01
1.16
22.35
0.26
42.50
1.14
26.18
0.26
40.12
1.32
26.65
0.27
40.47
1.29
25.73
0.27
τb (s)
Green Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
Influence of steam on catalyst
Brønsted : Lewis ratio via TPD. Fig. 5 shows the catalyst
acidity as a function of steam exposure time during cellulose
CFP. The Brønsted acidity undergoes a 25% decrease during
the first 30 min of steam exposure followed by a gradual
decrease. The Lewis acidity increased with steam exposure,
suggesting that Brønsted sites converted to Lewis sites as the
catalyst underwent dealumination as described by Ong et al.37
The total acidity of the catalyst decreased from 114 µmol g−1 to
42 µmol g−1. This analysis shows that the most significant
acidity changes occur during the initial 30 min of exposure to
steam.
Si : Al ratio via ICP. As shown in Fig. 6, ICP analysis showed
that the Al and Si content within the catalyst (consisting of
both zeolite and binder) did not change after steam exposure.
Other researchers observed loss of framework aluminum in
the presence of steam, though at higher temperatures. Corma
et al. found that a 35% steam atmosphere can remove some
Fig. 3 Aromatic carbon selectivity as function of input vapor fraction of
steam for catalytic fast pyrolysis of cellulose with ZSM-5 catalyst. (Reac-
tion conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction:
■
60%, Temperature: 500 °C, time on stream: 30 min, 90 g catalyst.)
:
:
▲
◆
●
▼
Benzene, : Toluene, : Xylenes + Ethylbenzene, : Naphthalenes,
Styrene, ★: Indene.
Fig. 4 Olefin carbon selectivity as function of input vapor fraction of
steam for catalytic fast pyrolysis of cellulose with ZSM-5 catalyst. (Reac-
tion conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction:
■
Fig. 5 Catalyst acid-site concentration as a function of steam exposure
time. (Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam
fraction: alternating between 0% and 60%, Temperature: 500 °C, time
60%, Temperature: 500 °C, time on stream: 30 min, 90 g catalyst.)
:
▲
◆
●
Ethylene, : Propylene, : Butenes, : Pentenes.
■
▲
◆
:
on stream: 30 min, 90 g catalyst.) : Brønsted sites, : Lewis sites,
Total acid sites.
The amount of butenes and pentenes was very low and no
clear change was shown with respect to steam vapor fraction.
Effects of alternating no-steam and steam
The data in the previous section were collected with fresh cata-
lyst for each run. These data offered little insight as to whether
these trends were driven due to steam causing irreversible
changes to the catalyst, or steam itself interfering with the
reactions of pyrolysis and/or catalysis. To elucidate this matter,
a series of cellulose CFP reactions were performed over a
single 90 g batch of catalyst wherein the fluidizer gas was alter-
nated as helium and steam (Temp. = 500 °C, WHSV = 0.4 h−1
,
time-on-stream = 30 min, fluidizer flow = 600 sccm.) At the
conclusion of these “no-steam/steam runs”, the catalyst had
been exposed to steam for a total of 330 min, though the cata-
lyst had been run through cellulose CFP for a time roughly
twice that. A small amount of catalyst (∼1.0 g) was extracted
after each steam run for analysis.
Fig. 6 ZSM-5 Si and Al concentrations as a function of steam exposure
time. (Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam
fraction: alternating between 0% and 60%, Temperature: 500 °C, time
■
▲
on stream: 30 min, 90 g catalyst.) : Si concentration, : Al concen-
◆
●
tration, : Si : Al molar Ratio, : P concentration.
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 7 XRD pattern of steamed ZSM-5 catalyst. (Reaction conditions:
Cellulose feed, WHSV: 0.4 h−1, Input steam fraction: alternating between
0% and 60%, Temperature: 500 °C, time on stream: 30 min, 90 g cata-
lyst.) A: XRD pattern; B: Crystallinity of ZSM-5 catalyst.
Fig. 8 Mesopore size distributions determined from BJH adsorption,
(Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam frac-
tion: alternating between 0% and 60%, Temperature: 500 °C, time on
■
▲
stream: 30 min, 90 g catalyst.) : fresh catalyst, : 30 min of steaming,
▼
◆
: 60 min of steaming, : 330 min of steaming.
ZSM-5 framework Al during catalytic cracking of naphtha for
production of propene and ethane at 500–700 °C when vapor
fractions of steam exceed 17.8%.38 Iliopoulou et al. also found
that steam presence decreases the acidity and Al content of
calculated from methods described by Sing et al.51 The BET
surface area, the micropore (<2 nm) area, and the mesopore
MCM-41 at 550 °C and 750 °C for vapor fractions of steam of (3–4 nm) volume increased during the first 30 min of steam-
20%.36
ing. As micro- and meso-pores are necessary for the formation
The phosphorus content was also measured via ICP and
of aromatics, this increased pore population most likely
found to be stable with steaming, remaining at a value of increases the formation of aromatic products.19 In contrast,
∼3.85 wt% as shown in Fig. 6. Both binder and phosphorus
have been reported to inhibit the volatilization of framework
aluminum via steaming.41,45
Crystallinity via XRD. The powder XRD patterns of the
steamed ZSM-5 samples are shown in Fig. 7. The intensity and
the volume of macropores (>50 nm) decreased with steaming
in the first 30 min. Since these large pores are located between
zeolite crystals and within the binder, this suggests that macro-
pores collapsed with steam exposure, a finding which is con-
sistent with the increase in zeolite crystallite size as shown
peak positions of all of the zeolite samples are in good agree- from the XRD measurements. With fewer macropores, micro-
ment with previously reported spectra, including typical diffr-
action peaks at 2θ = 7.88° (011), 8.76° (020), 23.0° (501), 23.84°
pores are more susceptible to blockage by char/coke, thus
inhibiting the formation of the hydrocarbon pool, and lower-
(033) and 24.3° (133).46,47 No change in the XRD peak posi- ing the catalyst activity.52,53 The micropore area and micropore
tions or new peaks appeared after steam exposure. However,
the zeolite crystal size increased from ∼50 nm to ∼80 nm after
volume both decreased below their initial values at 330 min
steam exposure time. The population of large macropores
a steam exposure time of 60 min and remained stable with (>50 nm) did not change substantially after its initial change
further steaming as shown in Fig. 7B. This decrease in crystal-
lite size may cause further decreases in catalyst activity by
decreasing diffusion limitations.48–50
Porosity via nitrogen adsorption. The BET surface area and
pore volume distribution based on liquid N2 isothermal
within the first 30 min, suggesting that at this time the macro-
pores had finished collapsing and the zeolite crystals began
the process of fusing. The SEM images shown in Fig. 9 quali-
tatively further confirmed this result.
Particle size distribution and physical integrity via
adsorption at −196 °C are shown in Table 2 and Fig. 8, as SEM. The particle size distribution was determined via SEM
Table 2 BET surface area and pore volume parameters of steamed catalyst. (Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam frac-
tion: alternating between 0% and 60%, Temperature: 500 °C, time on stream: 30 min, 90 g catalyst.)
BET surface area
External surface areaa
Micropore areaa
Micropore volume
Average pore size
(nm)
Sample
(m2 g−1
)
(m2 g−1
)
(m2 g−1
)
(cc g−1
)
Fresh catalyst
30 min
60 min
108.86
156.10
145.67
109.48
36.93
38.44
50.89
49.81
130.19
187.89
219.11
112.61
0.040
7.06
4.49
3.73
5.13
0.0598
0.0486
0.0314
330 min
a External surface area, micropore area, and pore volume derived based on t-plot method, Average pore size is calculated with 4 V A−1 by BET.
Determined from the N2 isothermal uptake at P/P0:0.10–0.90 at −196 °C.
Green Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
Fig. 9 Surface morphology of steamed catalyst based on SEM. (Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction: alternating
between 0% and 60%, Temperature: 500 °C, time on stream: 30 min, 90 g catalyst.) A: fresh catalyst, B: 30 min of steaming, C: 60 min of steaming,
D: 330 min of steaming.
as shown in Fig. 11. Above a steam exposure time of 60 min,
the particles appear to have finished agglomerating, as the
peak population size remains at 90 µm. Past 60 min however,
more catalyst fragments appear, suggesting that catalyst attri-
tion becomes important at longer durations of use, though
this will be true regardless of whether there is steam present
in the reactor.
Influence of steam on product yields. Fig. 12 shows the pro-
gress of product yields for the alternating no-steam/steam runs
as a function of steam-exposure time. The data have been sep-
arated into runs performed without steam co-feeding (Fig. 12A
and C) and with steam co-feeding (Fig. 12B and D). Clearly,
steaming changes the product yield. For the CFP runs without
steam co-feeding, the yield of aromatics decreased gradually
Fig. 10 Particle size distribution of steamed catalyst based on SEM.
(Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam frac-
tion: alternating between 0% and 60%, Temperature: 500 °C, time on
stream: 30 min, 90 g catalyst.)
from 29% C to 18% C over the course of steaming. The char/
coke yield decreased from 15% C to 7.8% C over the course of
steaming, while the olefin and methane yields showed a
gradual increase. The CO and CO2 both went through a
maximum yield at 60 min and 30 min of steaming respectively,
as shown in Fig. 10 and 11. Fresh catalyst particles are gradually decreasing thereafter. The identified oxygenates
almost exclusively spherical, but with a large fraction of fines increased slightly from 1.5% C to 3% C in the first 60 min but
(<20 µm). In the first 30 min of steaming, the largest differ- remained stable thereafter. The results indicate that the
ence is the loss of these fines, though the peak population product yields are greatly affected by the structural changes to
size above 20 µm does increase slightly from 40 µm to 50 the catalyst induced by steam exposure, as most changes to the
µm, suggesting that the particles are already beginning to catalyst properties do change before 90 min of steam exposure
agglomerate. With further steam exposure (60 min), the par- as shown in Fig. 5 through 11. The pore-size distribution
ticles agglomerated further and shifted the peak population doesn’t change after 30 min of steam exposure, crystal growth
size to 90 µm. More irregularly-shaped particles appear as stops occurring after 60 min steam exposure, and the particle
smaller particles become agglomerated with larger particles size distribution doesn’t change after 60 min steam exposure.
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 11 SEM images of typical catalyst particles at various stages of steaming. Inset photos show examples of various stages of particle agglomera-
tion. (Reaction conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction: alternating between 0% and 60%, Temperature: 500 °C, time on
stream: 30 min, 90 g catalyst.) A: fresh catalyst, B: 30 min of steaming, C: 60 min of steaming, D: 330 min of steaming.
Past steam exposure times of 60 min, the yields of aromatics,
coke/char, CO, and CO2 seem to correlate with decreases in
Brønsted acidity. The unidentified balance carbon increased
with steam exposure, suggesting that heavier unidentifiable
oxygenated species were produced in greater amounts as the
catalyst underwent these changes.
For CFP runs with steam co-feeding, the changes in the
product yield can be divided into two ranges: before and after
the first 90 min. With steam co-feeding, the carbon yield of
aromatics decreased from 24% C to 10% C in the first 90 min,
but remained constant after. CO and methane showed the
opposite tendency and their yields became enhanced particu-
larly in the first 90 min, and remained somewhat constant
with further steaming. With steam co-feeding, the char/coke
yield decreased gradually from 15% C to 5% C over the entire
150 min of steam exposure. The CO2 yield slightly decreased
from 4% to 3%. The olefin yield for steam co-feeding did
decrease over the span of steaming, though these results are
noisy in the range of <90 min, so it cannot be said whether
this change was abrupt (as with aromatics, CO and methane),
Fig. 12 Carbon yield as function of catalyst steam exposure time for or gradual (as with char/coke and CO2). The identifiable oxyge-
catalytic fast pyrolysis of cellulose with ZSM-5 catalyst. (Reaction con-
nate species yield was stable at about 1.5% C. As was the case
without steam co-feeding, unidentified balance carbon
increased with steam exposure, again suggesting that heavier
ditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction: alternating
between 0% and 60%, Temperature: 500 °C, time on stream: 30 min,
90 g catalyst.) A,C: Without steam co-feeding; B,D: With steam co-
unidentifiable oxygenated species were produced in greater
amounts as the catalyst underwent changes.
■
□
●
▲
◆
feeding. : Aromatics, : CO, : Char/coke, : Unidentified carbon, :
▼
Olefins, : CO2, ★: Methane, *: Identifiable Oxygenates.
Green Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
Influence of steam on aromatic and olefin selectivities. The The shift in both data sets is similar, even though
aromatic selectivity changed over the course of steam-exposure steam co-feeding consistently produces more ethylene than
time as shown in Fig. 13. For runs without steam co-feeding without steam co-feeding. In both cases, the yields of
(Fig. 13A), the aromatic selectivities did not change. For runs butenes and pentenes is quite low (<1% C), and steam
with steam co-feeding (Fig. 13B), the first 90 min of steaming co-feeding had no observable effect during the process of
caused the benzene selectivity to increase and the toluene cellulose CFP.
selectivity to decrease. The carbon selectivity of xylenes
decreased while that of styrene increased in that same period.
The naphthalenes and indene selectivities did not change.
Influence of steam on homogeneous and heterogeneous
reactions
After 90 min of steam exposure, all the selectivities were fairly Student’s t-test was performed to distinguish between (1) the
constant. These results suggest that changes to the catalyst long-term influences of steam’s irreversible effects on the cata-
(such as decreases in the amount of Brønsted acidity, shrink- lyst during CFP and how this affects product distribution, and
ing of mesopore area, and increases in zeolite crystallinity) did (2) the interaction steam itself has with changing the product
not show great influence on aromatics selectivity, but the pres- distribution through potentially acting as a homogeneous cata-
ence of steam did enhance the formation of toluene over lyst, a reactant species, or a heterogeneous catalyst site
benzene in the first 30 min.
blocker. The results are shown in Table 3. Three runs were
The ethylene selectivity increased and the propylene taken from each no-steam/steam data set. In each case, the
selectivity decreased with steam exposure time for runs data selected had a steam exposure time greater than 90 min
both with and without steam co-feeding as shown in Fig. 14. because before this point, product yields changed too rapidly
for an average to have any meaning. Although product yields
do change after 90 min, for the runs selected, an average is
ˉ
representative. The average (x) and standard deviation (s) are
shown for each group and compound in each data set. Also
shown is the t-statistic (t) and whether this value is statistically
significant to within 95% confidence (and in which direction).
Within the confines of 95% confidence, steam co-feeding was
found to decrease the yields of aromatics, char/coke, and
identifiable oxygenates, and increase the yields of methane
and CO. No measurable change in the yield of CO2 was
observed.
Steam co-feeding decreased the yield of all aromatics
species except naphthalene, though any changes to olefin yield
were not statistically significant. For oxygenates, steam co-
feeding decreased the yield of acetaldehyde, furan and hydro-
xyacetone, though no measurable influence was observed for
other oxygen-containing species detected.
Fig. 13 Aromatic carbon selectivity as function of steam co-feeding
time for catalytic fast pyrolysis of cellulose with ZSM-5 catalyst. (Reac-
tion conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction:
alternating between 0% and 60%, Temperature: 500 °C, time on stream:
30 min, 90 g catalyst.) A: Without steam co-feeding; B: With steam co-
Worth noting is the near statistical significance in the yield
changes in ethylene and propylene. While the effects at this
level of confidence are not measurable within this sample size,
these results do suggest that there may be an effect of steam
co-feeding that causes an increase in the yield of ethylene and
a decrease in the yield of propylene. These results run contrary
to the work of Gilbert et al. who showed for furan CFP, more
propylene is formed with steam co-feeding.35 While this dis-
parity might be attributable to differences in operating temp-
erature and/or vapor residence time, a stronger implication is
that cellulose does not form a furan-related intermediate
during CFP.
■
●
:
▲
◆
:
feeding.
:
Benzene,
▼
:
Toluene,
Xylenes
+
Ethylbenzene,
Naphthalenes, : Styrene, ★: Indene.
In comparison with steam co-feeding, the carbon yield of
aromatics, char/coke and oxygenates showed higher yields,
while that of CO and methane is lower for the run without
steam co-feeding for each catalyst with same steam exposure
time. We speculate that steam is competing with pyrolytic
vapors for adsorption on active sites and therefore reversibly
blocking the catalytic sites.38 Also the presence of steam might
dilute the concentration of aromatics intermediates/pyrolytic
Fig. 14 Olefin carbon selectivity as function of steam co-feeding time
for catalytic fast pyrolysis of cellulose with ZSM-5 catalyst. (Reaction
conditions: Cellulose feed, WHSV: 0.4 h−1, Input steam fraction: alter-
nating between 0% and 60%, Temperature: 500 °C, time on stream:
30 min, 90 g catalyst.) A: Without steam co-feeding, B: With steam co-
■
●
▲
◆
feeding. : Ethylene, : Propylene, : Butenes, : Pentenes.
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
Table 3 The influence of steam on cellulose CFP based on Student’s t-test (Reaction conditions: Cellulose feed, WHSV: 0.4 h-1, Input steam frac-
tion: alternating between 0% and 60%, Temperature: 500 °C, time on stream: 30 min, 90 g catalyst.) x = sample mean, s = sample standard deri-
¯
vation, t = t-statistic, 1-α = confidence level, tcrit = threshold of statistical significance according to chosen α. “↑” indicates a statistically significant
increase, “↓” indicates a statistically significant decrease, and “—” indicates no statistically significant change
No steam, n = 3
Steam, n = 3
α = 5%
ˉ
x
ˉ
x
s
s
All
Yield (%C)
Yield (%C)
t
tcrit = 2.78
Aromatics
Olefins
Id. oxygenates
CO
17.46%
9.31%
3.26%
25.16%
3.49%
7.62%
3.95%
70.24%
1.65%
0.36%
0.09%
1.25%
0.33%
0.84%
0.24%
3.92%
11.16%
10.16%
2.06%
31.89%
3.54%
5.51%
7.45%
71.77%
0.58%
1.09%
0.07%
3.94%
0.65%
0.51%
0.92%
6.76%
−6.24
↓
—
↓
↑
—
↓
↑
—
1.28
−18.22
2.82
0.12
−3.7
6.39
0.34
CO2
Char/coke
Methane
TOTAL
Aromatics
Yield (%C)
Yield (%C)
Benzene
Toluene
Xylenes
Naphthalene
Ethyl-benzene
Styrene
Indene
Total aromatics
6.19%
5.89%
1.58%
1.94%
0.42%
0.62%
0.82%
17.46%
0.41%
0.72%
0.23%
0.25%
0.02%
0.02%
0.06%
1.65%
3.94%
3.19%
0.89%
1.82%
0.36%
0.44%
0.52%
11.16%
0.23%
0.20%
0.05%
0.25%
0.03%
0.03%
0.04%
0.58%
−8.19
−6.28
−5.1
−0.59
−3.96
−8.72
−7.31
−6.24
↓
↓
↓
—
↓
↓
↓
↓
Olefins
Yield (%C)
Yield (%C)
C2
C3
C4
C5
7.33%
1.49%
0.18%
0.31%
9.31%
0.49%
0.61%
0.11%
0.14%
0.22%
9.06%
0.73%
0.14%
0.23%
10.16%
1.02%
0.12%
0.11%
0.18%
0.38%
2.64
−2.12
−0.47
−0.58
1.28
—
—
—
—
—
Total olefins
Id. oxygenates
Yield (%C)
Yield (%C)
Benzofuran
Phenol
Hydroxyacetone
HMF
Acetaldehyde
Furan
2-MF
0.22%
0.45%
0.11%
0.60%
0.85%
0.96%
0.05%
0.01%
3.26%
0.01%
0.05%
0.01%
0.04%
0.04%
0.09%
0.09%
0.00%
0.09%
0.20%
0.41%
0.06%
0.63%
0.43%
0.31%
0.02%
0.01%
2.06%
0.00%
0.02%
0.02%
0.10%
0.08%
0.08%
0.04%
0.01%
0.07%
−2.24
−1.49
−4
—
—
↓
—
↓
0.4
−8.47
−9.6
−0.53
−2.35
−18.22
↓
—
—
↓
Acetic acid
Total Id. oxygenates
vapors, thus lowering aromatics conversion and decreasing
Steam caused dealumination and
a
conversion of
aromatics and char/coke yields.39 However, the presence of Brønsted acidity to Lewis acidity. In the first 60 min of steam
steam might enhance steam-shifting and methanation of exposure, the catalyst was also found to change in pore struc-
pyrolytic vapors and char/coke to form more CO and CH4, an ture as more fine mesopores formed at the cost of larger
effect that will be more apparent after the irreversible effects of mesopores in the binder. Exposure of the catalyst to steam
steam on the catalyst are complete.34
increased the zeolite crystal size. Furthermore, the catalyst
particles agglomerated to form larger particles with steam co-
feeding.
These irreversible catalyst changes were observed to effect
product yields, both with and without steam co-feeding. In
Conclusion
Steam’s effect on catalytic fast pyrolysis of cellulose was both cases, steamed catalyst had a lower aromatics and char/
studied at 500 °C using a ZSM-5 catalyst in a bubbling flui- coke yield and a higher methane and unidentified products
dized bed reactor. Steam affects CFP through (1) irreversible yield.
changes to the catalyst structure and (2) direct interference
through changing chemistries of pyrolysis and/or catalysis.
Reversible effects of steam co-feeding were also observed.
Steam co-feeding lowers yields of aromatics, char/coke, and
Green Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Green Chemistry
Paper
identifiable oxygenate species, increases yields of CO and 16 J. Adam, E. Antonakou, A. Lappas, M. Stocker,
methane, and does not change the overall yields of CO2 and
olefins. The data suggest that steam co-feeding favors the pro-
M. H. Nilsen, A. Bouzga, J. E. Hustad and G. Oye, Micro-
porous Mesoporous Mater., 2006, 96, 93–101.
duction of ethylene over propylene, but more data are required 17 J. Li, X. Y. Li, G. Q. Zhou, W. Wang, C. W. Wang,
to confirm this finding.
S. Komarneni and Y. J. Wang, Appl. Catal., A, 2014, 470,
115–122.
18 A. Q. Zheng, Z. L. Zhao, S. Chang, Z. Huang, H. X. Wu,
X. B. Wang, F. He and H. B. Li, J. Mol. Catal. A: Chem.,
2014, 383, 23–30.
Acknowledgements
This work was primarily funded by the National Science Foun- 19 J. Jae, G. A. Tompsett, A. J. Foster, K. D. Hammond,
dation Office of Emerging Frontiers in Research and Inno-
vation (EFRI) grant number 0937895. The first author also
S. M. Auerbach, R. F. Lobo and G. W. Huber, J. Catal., 2011,
279, 257–268.
received support from Guangzhou Science and Technology 20 D. J. Mihalcik, C. A. Mullen and A. A. Boateng, J. Anal. Appl.
Program key projects (National 973 project: no. 2013CB228102) Pyrolysis, 2011, 92, 224–232.
and National Natural Science Foundation of China (no. 21 G. T. Neumann and J. C. Hicks, ACS Catal., 2012, 2, 642–
51376076). We would like to thank Dr Hyung Ju Kim for the 646.
SEM measurement and discussion about these results. Pro- 22 E. F. Iliopoulou, S. Stefanidis, K. Kalogiannis, A. C. Psarras,
fessor G.W. Huber has an ownership interest in Anellotech,
which has licensed the technology reported in this
publication.
A. Delimitis, K. S. Triantafyllidis and A. A. Lappas, Green
Chem., 2014, 16, 662–674.
23 A. Pattiya, J. O. Titiloye and A. V. Bridgwater, Fuel, 2010, 89,
244–253.
24 S. C. Du, Y. J. Sun, D. P. Gamliel, J. A. Valla and
G. M. Bollas, Bioresour. Technol., 2014, 169, 188–
197.
Notes and references
1 T. Dickerson and J. Soria, Energies, 2013, 6, 514–538.
2 T. R. Carlson, Y. T. Cheng, J. Jae and G. W. Huber, Energy
Environ. Sci., 2011, 4, 145–161.
3 Y. T. Cheng, J. Jae, J. Shi, W. Fan and G. W. Huber, Angew.
Chem., Int. Ed., 2012, 51, 1387–1390.
25 H. Y. Zhang, T. R. Carlson, R. Xiao and G. W. Huber, Green
Chem., 2012, 14, 98–110.
26 K. G. Wang and R. C. Brown, Green Chem., 2013, 15, 675–
681.
27 P. R. Patwardhan, J. A. Satrio, R. C. Brown and
B. H. Shanks, Bioresour. Technol., 2010, 101, 4646–
4655.
4 J. Jae, R. Coolman, T. J. Mountziaris and G. W. Huber,
Chem. Eng. Sci., 2014, 108, 33–46.
5 P. U. Karanjkar, R. J. Coolman, G. W. Huber, M. T. Blatnik, 28 C. A. Mullen and A. A. Boateng, Ind. Eng. Chem. Res., 2013,
S. Almalkie, S. M. de Bruyn Kops, T. J. Mountziaris and
W. C. Conner, AIChE J., 2014, 60, 1320–1335.
6 G. Yildiz, M. Pronk, M. Djokic, K. M. van Geem, F. Ronsse,
52, 17156–17161.
29 C. A. Mullen, A. A. Boateng, R. B. Dadson and
F. M. Hashem, Energy Fuels, 2014, 28, 7014–7024.
R. van Duren and W. Prins, J. Anal. Appl. Pyrolysis, 2013, 30 H. P. Yang, R. Yan, H. P. Chen, D. H. Lee and C. G. Zheng,
103, 343–351. Fuel, 2007, 86, 1781–1788.
7 H. Y. Zhang, R. Xiao, B. S. Jin, D. K. Shen, R. Chen and 31 S. L. Du, H. P. Yang, K. Z. Qian, X. H. Wang and
G. M. Xiao, Bioresour. Technol., 2013, 137, 82–87. H. P. Chen, Fuel, 2014, 117, 1281–1287.
8 X. Wu, J. Markham, X. S. Sun and D. Wang, T Asabe, 2012, 32 Q. Zhang, J. Chang, T. J. Wang and Y. Xu, Energy Convers
55, 1879–1885. Manage, 2007, 48, 87–92.
9 V. Paasikallio, C. Lindfors, E. Kuoppala, Y. Solantausta, 33 E. Kantarelis, W. H. Yang and W. Blasiak, Energy Fuels,
A. Oasmaa, J. Lehto and J. Lehtonen, Green Chem., 2014,
16, 3549–3559.
10 H. Y. Zhang, Y. T. Cheng, T. P. Vispute, R. Xiao and
G. W. Huber, Energy Environ. Sci., 2011, 4, 2297–2307.
2013, 27, 4748–4759.
34 P. T. Williams, N. Nugranad and P. A. Horne, Biomass Gasi-
fication and Pyrolysis: State of the Art and Future Prospects,
1997, 422–430.
11 Y. T. Cheng and G. W. Huber, ACS Catal., 2011, 1, 611– 35 C. Gilbert, J. Espindola, W. Conner, J. O. Trierweiler and
628. G. W. Huber, ChemCatChem, 2014, 6, 2497–2500.
12 P. A. Horne, N. Nugranad and P. T. Williams, J. Anal. Appl. 36 E. F. Iliopoulou, E. V. Antonakou, S. A. Karakoulia,
Pyrolysis, 1995, 34, 87–108.
13 N. Y. Chen, T. F. Degnan and L. R. Koenig, Chem. Technol.,
1986, 16, 506–511.
14 M. Olazar, R. Aguado, M. J. San Jose and J. Bilbao, J. Chem.
Technol. Biotechnol., 2001, 76, 469–476.
I. A. Vasalos, A. A. Lappas and K. S. Triantafyllidis, Chem.
Eng. J., 2007, 134, 51–57.
37 L. H. Ong, M. Domok, R. Olindo, A. C. van Veen and
J. A. Lercher, Microporous Mesoporous Mater., 2012, 164, 9–
20.
15 C. A. Mullen, A. A. Boateng, D. J. Mihalcik and 38 A. Corma, J. Mengual and P. J. Miguel, Appl. Catal., A, 2012,
N. M. Goldberg, Energy Fuels, 2011, 25, 5444–5451.
417, 220–235.
This journal is © The Royal Society of Chemistry 2015
Green Chem.

View Article Online
Paper
Green Chemistry
39 A. Corma, O. Marie and F. J. Ortega, J. Catal., 2004, 222, 46 C. Baerlocher, W. M. Meier and D. Olson, Atlas of zeolite
338–347. framework types, Elsevier, Amsterdam, New York, 2007.
40 A. Corma, J. Mengual and P. J. Miguel, Appl. Catal., A, 2012, 47 S. H. Jhung, J. S. Chang, J. S. Hwang and S. E. Park, Micro-
421, 121–134. porous Mesoporous Mater., 2003, 64, 33–39.
41 Y. J. Lee, J. M. Kim, J. W. Bae, C. H. Shin and K. W. Jun, 48 K. Barbera, F. Bonino, S. Bordiga, T. V. W. Janssens and
Fuel, 2009, 88, 1915–1921. P. Beato, J. Catal., 2011, 280, 196–205.
42 D. S. Shihabi, W. E. Garwood, P. Chu, J. N. Miale, 49 P. Ratnasamy, G. P. Babu, A. J. Chandwadkar and
R. M. Lago, C. T. W. Chu and C. D. Chang, J. Catal., 1985,
93, 471–474.
43 J. M. Fougerit, N. S. Gnep, M. Guisnet, P. Amigues,
S. B. Kulkarni, Zeolites, 1986, 6, 98–100.
50 S. K. Sahoo, N. Viswanadham, N. Ray, J. K. Gupta and
I. D. Singh, Appl. Catal., A, 2001, 205, 1–10.
J. L. Duplan and F. Hugues, Zeolites and Related Micro- 51 K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou,
porous Materials: State of the Art 1994, 1994, 84, 1723–
1730.
R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure
Appl. Chem., 1985, 57, 603–619.
44 E. Kantarelis, W. Yang and W. Blasiak, Fuel, 2014, 122, 52 N. Viswanadham, G. M. Dhar and T. S. R. P. Rao, J. Mol.
119–125. Catal. A: Chem., 1997, 125, L87–L90.
45 T. Blasco, A. Corma and J. Martinez-Triguero, J. Catal., 53 N. Viswanadham, R. Kamble, M. Singh, M. Kumar and
2006, 237, 267–277. G. M. Dhar, Catal. Today, 2009, 141, 182–186.
Green Chem.
This journal is © The Royal Society of Chemistry 2015