(682d) Deactivation Pathways of Raw Bio-Oil Hydrodeoxygenation over Noble Metal Supported Bifunctional Catalysts
AIChE Annual Meeting
2019
2019 AIChE Annual Meeting
Catalysis and Reaction Engineering Division
Deoxygenation and Oxidation in Biomass Upgrading
Thursday, November 14, 2019 - 1:24pm to 1:42pm
Bio-oil is the liquid product
of biomass pyrolysis, with a potential application as alternative and
sustainable source of fuels and chemicals.1 However, it requires the removal or upgrading of oxygenated
compounds that are its main constituents. Among the available processes, hydrodeoxygenation
(HDO) offers the possibility of yielding hydrocarbons for fuel applications (severe
HDO) or partially hydrodeoxygenated chemicals (mild HDO).1,2 For these goals, catalysts based on noble metals (Pt,
Pd, Ru) or transition metals (Ni, Co) are respectively used.3 However, the bottleneck of HDO is the formation of
deactivating carbonaceous species on the catalyst surface, which cause severe
operational difficulties using raw bio-oil. In this context, understanding the
deactivation pathways during the reaction plays a key role in setting the
optimum operation conditions that allows processing raw bio-oil by controlling
the formation of carbon deposits. In this work, we have studied these pathways using
catalysts based on noble metal nanoparticles (Pt-Pd) supported on mildly acid
supports (an activated carbon and an equilibrated FCC catalyst) at conditions
of accelerated deactivation. The activated carbon support
was prepared form olive stone by chemically activation with H3PO4
using an impregnation ratio of 3 gH3PO4 gprecursor-1
and a temperature of 500 ºC. An industrial FCC catalyst provided by
Petronor was also used as support in order to compare the obtained results. Pt
and Pd were incorporated as metallic function in both supports with nominal
concentrations of 1 and 0.5 wt%, respectively. The raw bio-oil was
obtained through fast pyrolysis of black poplar sawdust at 450 °C in a pilot
plant, provided with a conical spouted bed reactor. Catalysts were named with
the abbreviation of the used support: ACP and FCC. HDO runs were carried out in
a fixed bed reactor under the following conditions: 400-450 ºC;
65 bar; 0.18 gcatalyst h g gbio-oil-1;
2000 cm3H2 cm-3bio-oil and time on
stream, 0-8 h. Bio-oil feedstock and the
liquid product (with aqueous and organic fractions) obtained at each hour on
stream were characterized through several chromatographic and spectroscopic
techniques: (i) Karl-Fischer titration for measuring the water content (Metrohm
830 K F Titrino plus apparatus); (ii) GC-MS for measuring the composition of
the aqueous fraction (Shimadzu GC/MS QP2010); (iii) GC´GC-MS for measuring the composition of the organic
fraction (Agilent 7890A-5975C) and; (iv) ex-situ FTIR (Nicolet 6700
apparatus). Deactivated catalysts were characterized by means of
temperature-programmed oxidation (TPO) coupled with thermogravimetric (TG)
measurements and FTIR spectroscopy. Figure 1 shows the evolution of the HDO conversion
(XHDO) with time on stream for the ACP catalyst at 400, 425 and
450 ºC and for the FCC catalyst a 450 ºC. A significant drop of the
conversion is observed in all cases during the initial hours of reaction.
However, a more pronounced decay is exhibited during the reaction at 400 and
425 ºC, being negligible the conversion after 6 h at 400 ºC. A
massive deposition of carbonaceous species was observed at these conditions; thus,
leading to the reactor plugging. Otherwise, both catalysts reach a pseudo-steady
activity state at 450 ºC, which is attributed to the similar rates of the
formation of deactivating species and of the hydrocracking of their precursors.2 For this equilibrium of rates, the presence of an
acid function in the catalyst is crucial. This explains the higher registered
conversion with the ACP catalyst, which present stronger acid sites and a more
suitable porous structure than FCC catalyst. Monitoring the evolution of HDO
runs with ex-situ GC-MS, GC´GC-MS
and FTIR analyses, the presence of alkyl-phenols in the reaction medium from
the very first hours of reaction is observed at temperatures lower than
450 ºC for the ACP catalyst. Their concentration increases with TOS, as
well as the ones of other oxygenates (ketones and alcohols, among others).
Otherwise, the concentration of alkyl-phenols is much lower rising the
temperature up to 450 ºC but higher and constant concentrations of
alkyl-aromatics are registered from 4 h on stream, which coincides with the
pseudo-steady state of the catalyst.
Figure 1. Evolution of the HDO conversion with time
on stream Once finished the experimental
runs, used catalysts were analyzed by TPO-TG and the obtained profiles are
depicted in Figure 2. Three different species can be identified, namely
thermal lignin (TL), catalytic coke (CK) and activated carbon support (ACP).
Experiments with bare TL (thermal degradation of bio-oil in an inert solid bed)
and ACP were carried out in order to reinforce these assignations (see the
inset in Figure 2). The amount of TL is greater at the lowest HDO
temperature, whereas the one of CK increases with the temperature. According to
the thermal and catalytic origin of each species and these results, TL should
be deposited on the mesopores of the catalyst. This causes a severe
deactivation since the access of the reactant towards the active sites, mainly
located in the micropores, is hindered. On the contrary, CK should be located
within the micropores of the support near to the active sites but causing a
less severe deactivation. Interestingly, the capability of reaching the
pseudo-steady activity state (Figure 1a) is strictly related to the minimum
formation of TL and the maximum proportion of CK. Hence, this deactivation can
be partially controlled by using a HDO temperature of 450 °C, which favors the
HDO of oxygenated compounds and the partial cracking of coke precursors.
Regarding the results with the FCC catalyst, more heterogeneous carbonaceous
deposits, formed by both TL and CK, are observed. Nevertheless, higher presence
of TL is shown compared to that of ACP at the same temperature. The difference
in the porous structure and acidity of both supports could explain this
discrepancy that also explains the lower conversion value at the pseudo-steady
state displayed in Figure 1. The FTIR analyses of the used catalyst are
consistent with these results and indicate a higher contribution of the bands
at 1260, 1700 and 322-3600 cm-1 at lower temperatures and an
opposite trend for those between 2800-3000 cm-1. The first group is
attributed to oxygen-containing functionalities typical of TL, which is formed
by phenolic chains resulting from the bio-oil thermal repolymerization. The
second one is associated with aliphatic brands in aromatics structures and can
be related to the presence of CK.
Figure 2. TPO profiles of the used ACP and FCC
catalysts From the chromatographic
results of the last hour of reaction, insights into the main precursors of coke
could be provided in order to elucidate the origin of each deactivating species
identified (TL and CK). Figure 3 shows the relative concentration of
each of these precursors in the reaction medium. X axis is given in terms of DBE,
which is a standard magnitude for measuring the unsaturation degree of a
molecule,2 thus the higher the DBE value, the more untsaturated
is the molecule. Oxygenate and hydrocarbon molecules exhibit two contrasting
trends. On the one hand, the concentration of alkyl-phenols and C6+
oxygenated molecules decreases when the HDO temperature is raised. On the other
hand, alkyl-aromatics with more than 2 rings increase their concentration with
temperature, even though they remain significantly lower in contrast with
oxygenated coke precursors. This result is consistent with the favored oxygen
removal and aromatization at high temperature that leads to higher HDO
conversions (Figure 1), lower amounts of O-containing TL and higher of
polyaromatic structures of CK (Figure 2).
Figure 3. Distribution
of representative products in the reaction medium using ACP and FCC catalysts
at 400-450 ºC Then, two deactivation pathways
can be differentiated during the HDO of raw bio-oil: (i) a thermal pathway from
alkyl-phenols and oxygenated compounds of bio-oil that forms the TL and causes
a fast catalyst deactivation; (ii) a catalytic pathway from alkyl-aromatics
that leads to the development of CK near to the active sites of the catalyst. HDO
temperature, along with the catalyst acidity, plays a key role in the
composition of the reaction medium and allows controlling these two pathways.
Thus, the formation of TL can be minimized and a pseudo-steady state of
conversion can be obtained at temperatures as high as 450 ºC using a noble
metal supported catalyst on a mildly acid activated carbon. (1) Gollakota, A.
R. K.; Reddy, M.; Subramanyam, M. D.; Kishore, N. A Review on the Upgradation
Techniques of Pyrolysis Oil. Renew. Sustain. Energy Rev. 2016, 58,
1543–1568. (2) Cordero-lanzac,
T.; Palos, R.; Hita, I.; Arandes, J. M.; Rodríguez-Mirasol, J.; Cordero, T.;
Bilbao, J.; Castaño, P. Revealing the Pathways of Catalyst Deactivation by Coke
during the Hydrodeoxygenation of Raw Bio-Oil. Appl. Catal. B Environ. 2018,
239, 513–524. (3) Patel, M.;
Kumar, A. Production of Renewable Diesel through the Hydroprocessing of
Lignocellulosic Biomass-Derived Bio-Oil: A Review. Renew. Sustain. Energy Rev. 2016, 58, 1293–1307.
of biomass pyrolysis, with a potential application as alternative and
sustainable source of fuels and chemicals.1 However, it requires the removal or upgrading of oxygenated
compounds that are its main constituents. Among the available processes, hydrodeoxygenation
(HDO) offers the possibility of yielding hydrocarbons for fuel applications (severe
HDO) or partially hydrodeoxygenated chemicals (mild HDO).1,2 For these goals, catalysts based on noble metals (Pt,
Pd, Ru) or transition metals (Ni, Co) are respectively used.3 However, the bottleneck of HDO is the formation of
deactivating carbonaceous species on the catalyst surface, which cause severe
operational difficulties using raw bio-oil. In this context, understanding the
deactivation pathways during the reaction plays a key role in setting the
optimum operation conditions that allows processing raw bio-oil by controlling
the formation of carbon deposits. In this work, we have studied these pathways using
catalysts based on noble metal nanoparticles (Pt-Pd) supported on mildly acid
supports (an activated carbon and an equilibrated FCC catalyst) at conditions
of accelerated deactivation. The activated carbon support
was prepared form olive stone by chemically activation with H3PO4
using an impregnation ratio of 3 gH3PO4 gprecursor-1
and a temperature of 500 ºC. An industrial FCC catalyst provided by
Petronor was also used as support in order to compare the obtained results. Pt
and Pd were incorporated as metallic function in both supports with nominal
concentrations of 1 and 0.5 wt%, respectively. The raw bio-oil was
obtained through fast pyrolysis of black poplar sawdust at 450 °C in a pilot
plant, provided with a conical spouted bed reactor. Catalysts were named with
the abbreviation of the used support: ACP and FCC. HDO runs were carried out in
a fixed bed reactor under the following conditions: 400-450 ºC;
65 bar; 0.18 gcatalyst h g gbio-oil-1;
2000 cm3H2 cm-3bio-oil and time on
stream, 0-8 h. Bio-oil feedstock and the
liquid product (with aqueous and organic fractions) obtained at each hour on
stream were characterized through several chromatographic and spectroscopic
techniques: (i) Karl-Fischer titration for measuring the water content (Metrohm
830 K F Titrino plus apparatus); (ii) GC-MS for measuring the composition of
the aqueous fraction (Shimadzu GC/MS QP2010); (iii) GC´GC-MS for measuring the composition of the organic
fraction (Agilent 7890A-5975C) and; (iv) ex-situ FTIR (Nicolet 6700
apparatus). Deactivated catalysts were characterized by means of
temperature-programmed oxidation (TPO) coupled with thermogravimetric (TG)
measurements and FTIR spectroscopy. Figure 1 shows the evolution of the HDO conversion
(XHDO) with time on stream for the ACP catalyst at 400, 425 and
450 ºC and for the FCC catalyst a 450 ºC. A significant drop of the
conversion is observed in all cases during the initial hours of reaction.
However, a more pronounced decay is exhibited during the reaction at 400 and
425 ºC, being negligible the conversion after 6 h at 400 ºC. A
massive deposition of carbonaceous species was observed at these conditions; thus,
leading to the reactor plugging. Otherwise, both catalysts reach a pseudo-steady
activity state at 450 ºC, which is attributed to the similar rates of the
formation of deactivating species and of the hydrocracking of their precursors.2 For this equilibrium of rates, the presence of an
acid function in the catalyst is crucial. This explains the higher registered
conversion with the ACP catalyst, which present stronger acid sites and a more
suitable porous structure than FCC catalyst. Monitoring the evolution of HDO
runs with ex-situ GC-MS, GC´GC-MS
and FTIR analyses, the presence of alkyl-phenols in the reaction medium from
the very first hours of reaction is observed at temperatures lower than
450 ºC for the ACP catalyst. Their concentration increases with TOS, as
well as the ones of other oxygenates (ketones and alcohols, among others).
Otherwise, the concentration of alkyl-phenols is much lower rising the
temperature up to 450 ºC but higher and constant concentrations of
alkyl-aromatics are registered from 4 h on stream, which coincides with the
pseudo-steady state of the catalyst.
Figure 1. Evolution of the HDO conversion with timeon stream Once finished the experimental
runs, used catalysts were analyzed by TPO-TG and the obtained profiles are
depicted in Figure 2. Three different species can be identified, namely
thermal lignin (TL), catalytic coke (CK) and activated carbon support (ACP).
Experiments with bare TL (thermal degradation of bio-oil in an inert solid bed)
and ACP were carried out in order to reinforce these assignations (see the
inset in Figure 2). The amount of TL is greater at the lowest HDO
temperature, whereas the one of CK increases with the temperature. According to
the thermal and catalytic origin of each species and these results, TL should
be deposited on the mesopores of the catalyst. This causes a severe
deactivation since the access of the reactant towards the active sites, mainly
located in the micropores, is hindered. On the contrary, CK should be located
within the micropores of the support near to the active sites but causing a
less severe deactivation. Interestingly, the capability of reaching the
pseudo-steady activity state (Figure 1a) is strictly related to the minimum
formation of TL and the maximum proportion of CK. Hence, this deactivation can
be partially controlled by using a HDO temperature of 450 °C, which favors the
HDO of oxygenated compounds and the partial cracking of coke precursors.
Regarding the results with the FCC catalyst, more heterogeneous carbonaceous
deposits, formed by both TL and CK, are observed. Nevertheless, higher presence
of TL is shown compared to that of ACP at the same temperature. The difference
in the porous structure and acidity of both supports could explain this
discrepancy that also explains the lower conversion value at the pseudo-steady
state displayed in Figure 1. The FTIR analyses of the used catalyst are
consistent with these results and indicate a higher contribution of the bands
at 1260, 1700 and 322-3600 cm-1 at lower temperatures and an
opposite trend for those between 2800-3000 cm-1. The first group is
attributed to oxygen-containing functionalities typical of TL, which is formed
by phenolic chains resulting from the bio-oil thermal repolymerization. The
second one is associated with aliphatic brands in aromatics structures and can
be related to the presence of CK.
Figure 2. TPO profiles of the used ACP and FCCcatalysts From the chromatographic
results of the last hour of reaction, insights into the main precursors of coke
could be provided in order to elucidate the origin of each deactivating species
identified (TL and CK). Figure 3 shows the relative concentration of
each of these precursors in the reaction medium. X axis is given in terms of DBE,
which is a standard magnitude for measuring the unsaturation degree of a
molecule,2 thus the higher the DBE value, the more untsaturated
is the molecule. Oxygenate and hydrocarbon molecules exhibit two contrasting
trends. On the one hand, the concentration of alkyl-phenols and C6+
oxygenated molecules decreases when the HDO temperature is raised. On the other
hand, alkyl-aromatics with more than 2 rings increase their concentration with
temperature, even though they remain significantly lower in contrast with
oxygenated coke precursors. This result is consistent with the favored oxygen
removal and aromatization at high temperature that leads to higher HDO
conversions (Figure 1), lower amounts of O-containing TL and higher of
polyaromatic structures of CK (Figure 2).
Figure 3. Distributionof representative products in the reaction medium using ACP and FCC catalysts
at 400-450 ºC Then, two deactivation pathways
can be differentiated during the HDO of raw bio-oil: (i) a thermal pathway from
alkyl-phenols and oxygenated compounds of bio-oil that forms the TL and causes
a fast catalyst deactivation; (ii) a catalytic pathway from alkyl-aromatics
that leads to the development of CK near to the active sites of the catalyst. HDO
temperature, along with the catalyst acidity, plays a key role in the
composition of the reaction medium and allows controlling these two pathways.
Thus, the formation of TL can be minimized and a pseudo-steady state of
conversion can be obtained at temperatures as high as 450 ºC using a noble
metal supported catalyst on a mildly acid activated carbon. (1) Gollakota, A.
R. K.; Reddy, M.; Subramanyam, M. D.; Kishore, N. A Review on the Upgradation
Techniques of Pyrolysis Oil. Renew. Sustain. Energy Rev. 2016, 58,
1543–1568. (2) Cordero-lanzac,
T.; Palos, R.; Hita, I.; Arandes, J. M.; Rodríguez-Mirasol, J.; Cordero, T.;
Bilbao, J.; Castaño, P. Revealing the Pathways of Catalyst Deactivation by Coke
during the Hydrodeoxygenation of Raw Bio-Oil. Appl. Catal. B Environ. 2018,
239, 513–524. (3) Patel, M.;
Kumar, A. Production of Renewable Diesel through the Hydroprocessing of
Lignocellulosic Biomass-Derived Bio-Oil: A Review. Renew. Sustain. Energy Rev. 2016, 58, 1293–1307.