Bioresource Technology 361 (2022) 127727

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Highly selective hydropyrolysis of lignin waste to benzene, toluene and

xylene in presence of zirconia supported iron catalyst
William Lonchay a, Giuseppe Bagnato a, b, Aimaro Sanna a, *
a
Advanced Biofuels Lab, School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh EH14 4AS, UK
b
Department of Engineering, Lancaster University, Lancaster LA1 4YW, UK

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Hydrolysis lignin was hydropyrolysed to

BTX using metal doped zirconia
catalysts.
• Fe/ZrO2 was able to produce 67% BTX
from lignin and cyclopentadiene from
cellulose.
• Strong deoxygenation of the HYPy oil
(5 wt% O) was achieved.
• Catalytic performance linked to Fe0
oxophilicity, good dispersion and sup­
port acidity.
• 86% Hydrocarbons had C distribution of
C5-C10.

A R T I C L E I N F O A B S T R A C T

Keywords: The use of lignin to produce Benzene, Toluene and Xylene (BTX) is a promising pathway to strength the economic
Lignin case, over the production of advanced bio-fuels alone. In this work, Ce, Na, Pd and Fe supported on zirconium
Hydropyrolysis oxide were evaluated for the ex-situ hydropyrolysis (HyPy)/hydrodeoxygenation (HDO) of Etek lignin under mild
Bio-oil
conditions (600 ◦ C, 1 atmosphere) towards the production of BTX. Fe/ZrO2 was able to selectively produce BTX
Hydrodeoxygenation
Iron catalyst
(67 area%) and cycloalkenes (13.5 area%) and strongly deoxygenate the HyPy oil to about 5 wt% oxygen
Aromatics content, resulting in an oil with a carbon distribution of 85.5 % in C5-C10 hydrocarbons. The high selectivity of
Fe/ZrO2 was related to the iron oxophilicity, the strong reduction potential of zero-valent iron, the good
dispersion of Fe nanoparticles on the support and the presence of mesopores and acid sites, which enhanced the
interactions between the reacting species and the catalyst surface.

1. Introduction below 1.5 ◦ C (Welsby et al., 2021). One of the main products from the
petroleum industry are chemical feedstocks, which represents around
In recent years, there has been a concerted effort to decrease the 10 % of the global petroleum industry in terms of production volume,
reliance of modern consumer products on the fossil fuel industry and the but a significantly higher financial value due to the increased price of
need for a sharp decline in their use to keep the temperature increase petrochemicals compared to fuels. In particular, BTX are used to

* Corresponding author.
E-mail address: [email protected] (A. Sanna).

https://doi.org/10.1016/j.biortech.2022.127727
Received 23 June 2022; Received in revised form 28 July 2022; Accepted 29 July 2022
Available online 6 August 2022
0960-8524/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

manufacture common plastics like poly ethylene (PE), polystyrene (PS), the HDO catalyst should selectively produce BTX molecules by removing
polyurethanes (PUL), and nylon (NY) among other chemicals (“Biogreen the oxygen content (Jan et al., 2015). Some studies show HDO catalysts
Energy Syngas,” 2020). In consequence, there is growing interest in further cracking the bio-oil phenols into small olefins or alkane gas
replacing fossil derived with bio derived feedstocks, although it has been molecules, which have lower value than aromatic BTX compounds
suggested that despite biomass is an “indispensable” resource for the (Marker et al., 2013). Additional reduction of BTX molecules can pro­
circular economy, combusting it to electricity only is inefficient and the mote char build-up and catalyst deactivation, which is undesirable.
path to chemicals should be pursued (Hamer, 2020). Accordingly, lignin There are two main catalyst types used in the hydropyrolysis of
as renewable source of aromatics is an important element in achieving lignocellulosic biomass, the pyrolysis catalyst and the HDO catalyst,
the ‘net-zero emission by mid-century’ target outlined by The Paris where the most well-known conversion processes include both of them,
Agreement (Griffin et al., 2018). Lignin possesses a phenolic polymer such as the integrated hydropyrolysis and hydroconversion (IH2®)) and
structure comprised of p-hydroxy-phenyl, guaiacol, and syringyl groups, a two-step biofuel process (H2Bioil) (Venkatesan et al., 2020). The py­
and hence has good potential to provide renewable aromatic compounds rolysis catalyst is primarily used to increase the hydrogenation of the
for future use as feedstocks to the chemical manufacturing industry. biomass molecules and promote the breaking of the weakest C–C bonds
Aromatic production could be an economical way of kick-starting large- in the structure which in turn, breaks down the lignin and cellulose
scale lignocellulosic pyrolysis as a means to reduce dependency on oil polymer chains into oligomers, dimers and monomers (Sirous-Rezaei
derived chemicals and help promote renewable biomass as a staple in­ and Park, 2020). The most efficient catalysts for this are either powered
dustry in the global economy. metal oxide catalysts on a fluid bed reactor or acid site zeolite-based
Fast pyrolysis, which is the rapid heating of biomass in absence of structures. Zeolites are relatively expensive to prepare and have
oxygen for a short residence time followed by rapid quenching of the several well-documented shortcomings when used in biomass pyrolysis,
produced vapours can be used to produce either fuels or aromatics, with namely char build-up, fast deactivation and high regeneration energy,
appropriate variations in process conditions, which includes the use of suggesting they may not be suitable for large scale processes, although
catalysts (Liu et al., 2021a; Farooq et al., 2022; Hendry et al., 2020). On they promote aromatics production (Stummann et al., 2018; Zhu et al.,
an economic level, biomass pyrolysis has been assessed for fuel pro­ 2022; Jindal et al., 2022).
duction. A report published in 2015 by the US Department of Energy In the literature the main success in HDO catalysts for BTX produc­
shows an economic projection of how advanced bio-fuels can be made tion have been in metal oxides on a porous, surface-active supports
cost competitive vs fossil fuel derived transportation fuels (Dutta et al., under high H2 pressure (Resende, 2016; Stummann et al., 2021; Ven­
2015). The conclusion of this report was that biomass pyrolysis can katesan et al., 2020; Wang et al., 2013). Metal oxide reducing powders
compete with fossil fuel derived fuels based on a total yield of combined appear to be a better catalyst due to their higher reactivity, which does
gasoline and diesel equivalent fuels of 78 gallons / US ton dry of biomass not depend on molecule mobility rates through zeolite pores, as well as
and a fuel price of around $3.50 /gallon final product. The report out­ regeneration being easier and faster than that of zeolites (Jan et al.,
lines that the total product yield should be minimum 25 wt% of dry 2015). Palladium is a precious metal in the same chemical group as
feedstock for the whole process, including ex-situ HDO upgrading of Nickel and Platinum and is used widely in industry as a hydrogenation
pyrolysis vapours. catalyst, usually for the reduction of C– – C or C–– O bonds in organic
Recent studies show that pyrolysis under hydrogen atmosphere chemistry. Jan et al published a study in 2015 that showed very high
operating at a rapid heating rate namely hydropyrolysis (HyPy) is a yields for BTX from lignin samples using in-situ palladium doped HZSM-
promising technology for converting biomass into liquid fuels (e.g., bio- 5 zeolite (Jan et al., 2015). The reported yield was 40 wt% aromatics at
oil and C4 + hydrocarbons), since the addition of H2 enhances the H/C 600 ◦ C and 17 bar of H2. However, the catalyst was used with a catalyst
ratio of the bio-oil, reduces the O/C ratio, facilitates hydrotreating re­ to biomass ratio of 10:1 or 20:1, with the 1:1 resulting in only 4 wt%
actions that involve C–C coupling, hydrocracking, alkylation, decar­ aromatics produced, with C6-C10 aromatic and polyaromatic molecules
boxylation, decarbonylation, hydrogenation, HDO, and recombination counting in that total. BTX consisting of C6 and C7 aromatics made up
(Oh et al., 2021; Stummann et al., 2021). In particular, HDO of approximately 18 wt% with 20:1 catalyst to biomass ratio. The high
oxygenated aromatics (e.g., phenolic compounds) occurs via hydroge­ catalyst to biomass ratio would cause problems when scaling up these
nation of the aromatic ring followed by deoxygenation (Liu et al., experiments as the assumed reason for such high catalyst ratio is the
2021b). The second advantage is that the H• radicals ‘stabilise’ the high tendency for palladium and zeolites to quickly form char that leads
highly reactive intermediates species (e.g. ketone and aldehyde) and to rapid deactivation. Stummann et al. (2018) studied the hydro­
prevent them from condensing into coke, which reduces the carbon re­ pyrolysis of beech wood at 26 bar and 450 ◦ C in presence of ‘bog-iron’ as
covery and lead to higher operational costs (Resende, 2016). Presence of a hydropyrolysis catalyst and nickel-molybdenum alloy supported on
hydrogen also enhances demethylation of methoxy groups in phenolic aluminium oxide as the HDO catalyst. This study showed 24.7 wt% of
products and favours dehydration vs C loss pathways (–CO, –CO2). For condensed organic and C4 + molecules, which can be used as gasoline
example, non-methoxy phenolics reached 19.68C% at 30 bar and 500 additive (Stummann et al., 2020). This is important as one of the reac­
◦
C, where monocyclic and polycyclic aromatic hydrocarbons as well as tion paths to aromatics from hydropyrolysis is through C4 + interme­
condensable aliphatic hydrocarbons were observed from lignin hydro­ diate species (Norinaga et al., 2014).
pyrolysis (Wang et al., 2022). Temperature plays an important role in Sirous-Rezaei and Park (2020) studied the HyPy/HDO of kraft lignin
hydropyrolysis, where HDO takes place at increasing temperatures to using HY as in-situ catalyst and Fe/HBeta, FeReOx/MCM-41, Fe/ZrO2
with efficient elimination of furanic and phenolic oxygen-containing and FeReOx/ZrO2 as ex-situ HDO catalyst, under a continuous flow of H2
compounds at T > 500 ◦ C (Wang et al., 2013). However, Tian et al at 1 atmosphere pressure (Sirous-Rezaei and Park, 2020). While FeR­
(2021) showed that when hydropyrolysis temperature is increased from eOx/ZrO2 resulted the most efficient HDO catalyst (7.1 wt% aromatic
700 to 800 ◦ C and higher, the hydropyrolysis oil (from pine sawdust) hydrocarbons of which 4.8 wt% BTX), all FeReOx/MCM-41, Fe/ZrO2
decreased from 35 wt% to 25 wt% at expenses of gas phase (from 50 to and FeReOx/ZrO2 led to significantly lower yield of coke compared to a
70 wt%) (Tian et al., 2021). zeolite-supported catalyst like Fe/HBeta. Similar aromatics (7.6 wt.
The majority of studies show that hydrogen atmosphere (either at %/5.4 wt% BTX) were obtained using palladium rhenium oxide (Sirous-
pressure or ambient) with no catalyst can be sufficient to provide the Rezaei and Park, 2020). The lower cost ferrous rhenium catalyst mate­
desired bio-oil composition and vaporisation before the HDO step rial and low reaction pressure is significant because it represents a
(Zheng et al., 2017). This pathway demands a good understanding of the relevant advantage if this process is scaled up. The products from the
HyPy products so that a suitable HDO catalyst can be selected. For BTX iron-based catalyst also gave 11.4 wt% oxygenates of which 8.7 wt%
production, the pyrolysis vapour should consist of simple phenols and was phenol. These compounds are usually the primary targets for HDO

2
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

catalysts suggesting phenols deoxygenation can be improved. The same 2.3. Hydropyrolysis/HDO Set-up and method
research group showed the efficiency of ferrous rhenium oxide on zir­
conium oxide support in the upgrading and deoxygenating of mono- A -micro-pyrolysis reactor coupled with a gas chromatography/mass
phenols into BTX (Starting material – BTX wt%: Guaiacol – 21.6 wt%, spectrometer (Py-GC/MS) was used for the experimental campaign. The
m-Cresol – 61.7 wt%, Anisol – 48.3 wt%) (Sirous-Rezaei et al., 2018). system consisted of a CDS 5250 (Chemical Data Systems, USA) pyrop­
However, the starting materials for these studies were pure lignin/ robe hyphenated to a GC–MS Trace DSQ II (Thermo Scientific, USA).
phenols without any additional species that could interfere with the Fig. 1 schematises the reactor setup, consisting in an autosampler quartz
catalyst action (e.g. cellulose fraction). An early research show that Etek tube (1.9 mm diameter) where one mg of the Etek lignin was sand­
(Etanol Teknik) lignocellulosic filtration residue from acid straw hy­ wiched between 2 layers of 5 mg of selected catalyst and the two ma­
drolysis resulted in 34 % gas, 61 % liquid and 5 % coke (down from 19.6 terials were kept in place and separated by layers of quartz wool.
% in N2) when underwent hydropyrolysis at 800 ◦ C with both lignin and The reaction tests were performed at 600 ◦ C (heating rate of 10 ◦ C/
cellulose derived compounds accounted in the liquid product (Windt ms) in presence of hydrogen (100 mL/min) with final temperature
et al., 2009). maintained for 30 s. Although the true temperature in the microprobe is
Overall, the literature shows that integrated Hy-Py/HDO is the most likely 75-100◦ C lower than the set point (Thangalazhy-Gopakumar
effective pathway for deoxygenating pyrolysis bio-oil and also that et al., 2011). The temperature of 600 ◦ C was selected since literature
bifunctional zirconia supported metal catalysts such as FeZrO2 can turn indicates that this temperature was required to maximise recovery of
lignin into valuable aromatics under mild hydropyrolysis conditions. phenolics from Etek lignin pyrolysis (Hendry et al., 2020). The volatiles
However, selectivity on BTX is a unresolve challenge. Therefore, this generated in the pyroprobe were collected in a tar trap for separating the
study wants to investigate in-house synthetised ZrO2 supported Ce, Na non-condensable gases and further analyse them by GC–MS. The tar trap
Fe, PdFe catalysts in the integrated Hy-Py/HDO of Etek lignin waste consisted in a 1/8′′ Tenax kept at 40 ◦ C. The condensed species were
under ambient H2 pressure at 600 ◦ C using a pyroprobe reactor coupled then desorbed from the tar trap at 300 ◦ C for 3 min and sent to a GC–MS
with a gas chromatograph/mass spectrometer (Py-GC/MS), with the aim for analysis by a transfer line heated at 350 ◦ C. The bio-oil was injected
to (i) maximise BTX selectivity on oil product, (ii) deoxygenation of the in the GC at 280 ◦ C, using 25 mL/min of helium, as a carrier, and the
Hy-Py bio-oil and (iii) compare the effect of hydrogen and the selected split ratio of 80:1. The column used was an Agilent: HP-5MS, 19091S-
catalysts on the products distribution with those obtained in presence of 433; length, 30 m; internal diameter, 250 μm; film thickness, 0.25 μm
nitrogen. The GC oven temperature started at 40 ◦ C for 2 min, heated up to 320 ◦ C
at 12 ◦ C/min and kept at temperature for 15 min. The raw area per­
2. Materials and methods centages were recalculated by excluding the GC–MS peaks after 20 min
(see supplementary material) and normalising to 100 %. This was done
2.1. Materials because those peaks belonged to contaminants leached out from the GC
column at high temperature. The peaks were finally identified by the
Etek (Etanolteknik AB, SE) lignin is an industrial filtration residue NIST library. The Total Sum Normalization (TSN) method for normal­
remaining after a two stages weak acid hydrolysis of soft wood and ising GC–MS data, which consists in dividing the area of each peak in
consists of 41 % holocellulose and 59 % lignin (Nowakowski et al., each chromatogram by the total sum of all peaks within that chro­
2010). The Etek lignin contained 51 wt% C, 5.7 wt% H, 1.6 wt% N and matogram, was used in this work. Sum normalised data were multiplied
37.7 wt% O (by difference), while the proximate analysis gave 76.6 wt% by 100 and expressed in terms of their percent contribution (area%) to
volatiles, 19.4 wt% fixed carbon and 4 wt% ash. Na2CO3 (99.6 % purity), the total area. To evaluate the catalysts’ ability on steering the HyPy/
CeO2 and ZrO2 (99.0 % purity), ZrOCl2⋅8H2O (99.5 % purity) and PdCl2 HDO reactions to produce BTX in the oil product from the starting lignin
and FeN3O9x9H2O used for the synthesis of the catalysts were all pur­ fraction in Etek lignin, the term selectivity was used. The product was
chased from Sigma-Aldrich. studied in term of functional groups using Excel. Non-aromatic mole­
cules were classified in the following order: cycloalkene, cycloalkane,
epoxide ring, esters, ketones + aldehydes and olefins as summarised in
2.2. Catalysts preparation Table 1 including a brief description of the main uses. Alcohol groups
was a very common feature so were ignored in terms of classifications.
Na/ZrO2 (1.5_1), CeNa/ZrO2 (1_1_1), CeNa/ZrO2 (2_1_1) were syn­ Hence there are molecules in the cycloalkene class that are also ketones
thetized mixing by pestle & mortar Na2CO3, CeO2 and ZrO2 and then and have ester groups on branches off the cycloalkene ring. 2-Dodecenal
calcining the mixture at 900 ◦ C for 4 h in air. ZrO2 support for Fe/ZrO2 is an example of an olefin chain with an aldehyde group but was clas­
and PdFe/ZrO2 was instead prepared by dropwise addition of sified in the ketone + aldehyde group and not the olefin group as per the
ZrOCl2⋅8H2O to water (100 mL). After addition, the suspension was aged order above. The BTX content based on starting Etek lignin weight was
20 h at 90 ◦ C, dried in an oven at 110 ◦ C for 15 h and subsequently estimated considering (i) the bio-oil wt% resulting from Etek lignin
calcined in flowing air (30 mL/min) at 500 ◦ C for 3 h. 5 %Fe and 1 % Pd hydropyrolysis in absence of catalyst (61.2 wt% based on dry lignin)
on ZrO2 were then prepared by incipient wet impregnation method with under similar conditions (ambient pressure H2, 800 ◦ C) (Windt et al.,
PdCl2 and FeN3O9x9H2O as precursors. The dried resulting salts were 2009); (ii) a fraction of bio-oil detectable by GC–MS (~16.4–40 %)
then calcined in air at 500 ◦ C for 3 h to achieve good dispersion (Hendry (Windt et al., 2009); (iii) the fraction of HDO oil (23 wt%) obtained from
et al., 2020). lignocellulosic material under similar conditions (450 ◦ C, H2 at ambient

Fig. 1. Experimental set-up for catalytic tests.

3
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

Table 1 3. Results and discussion

Common products from the catalytic HDO tests.
Notable Species Potential uses 3.1. HyPy-GC–MS results
Epoxide Ring Oxirane Chemical
intermediate Table 2 shows the HyPy oils’ chemical composition from the GC–MS
Esters Isopropenyl Acetate Chemical feedstocks analysis together to the biomass component each group is originated
Ketones þ Aldehydes Acetone, Butanone, Solvents, chemical from. The HyPy oil product using CeNa/ZrO2 (1_1_1) (see supplemen­
Acetylpropionyl feedstocks tary material) contained 107 distinct compounds. This shows the low
Acids Acetic acid Solvent, chemical
feedstock
selectivity of the catalyst, with Ketones + Aldehydes (11.04 %), and
Olefins – Various, potential for Guaiacols (22.25 %) being the most abundant functionalities. The effect
fuels of hydrogen vs nitrogen during the Etek lignin pyrolysis was compared
Cycloalkenes Cyclopentadiene, Various, significant (see supplementary material) using the data produced in a previous
Cyclohexadiene uses
work (Hendry et al., 2020). The high HyPy temperature used (600 ◦ C)
Cycloalkanes – Potential for fuels,
solvents was necessary to maximise lignin break-down into monomers, which
Alkyl Phenols Cresol, Creosol, Xylenol HDO to BTX would then be converted into hydrocarbon via hydrogenation, decar­
Guaiacols Methyl-, Ethyl-, propyl-, HDO to BTX bonylation, deoxygenation pathway generating fuel grade hydrocarbons
Vinyl-Guaiacols (Hendry et al., 2020; Jindal et al., 2022). Clearly, the distribution of the
Other Aromatics Catechol, Phenyl carbamate, HDO to BTX
Bi-Phenyls
bio-oil compounds shifted to lower molecular weight compounds, such
BTX Benzene, Toluene, Xylene Chemical feedstocks as butanone and cyclopentadiene, although the most abundant com­
Polyaromatic Biphenyl, Fluorene, Used in dye industry pounds in N2 atmosphere (e.g. phenol; phenol-2-ethoxy; 3-creosol; lev­
Hydrocarbons (PAH) Phenanthrene oglucosan) remained qualitatively unchanged when H2 was used. Very
Oxygenated Sugar Levoglucosan, Stevioside, d- Limited uses
limited benzene was detected suggesting that both Ce and Na are not
Derivative Mannose
Other – Limited uses good in hydrogenating and deoxygenating the substituted aromatic
Furans Furfural, 3-Furaldehyde, 2,5- Speciality chemicals, rings. The GC–MS analyser from the hydropyrolysis reactions in pres­
dimethyl-Furan limited use ence of the same catalyst but with two moles of Ce instead of one [CeNa/
ZrO2 (2_1_1)] identified 46 distinct compounds in the oil product. This
shows an increased peak area of the catalyst over CeNa/ZrO2 (1_1_1),
pressure and 5 %Pd catalyst + NiMo/Al2O3) (Gholizadeh et al., 2016)
with the 20 most abundant molecules making up 87.73 % of the total
and (iv) the fraction of BTX from the GC–MS analysis (63 %) assuming
bio-oil, most of which presented a lower retention time compared to
this is proportional to the actual weight content of BTX.
CeNa/ZrO2 (1_1_1) and larger content in anhydrosugars derived from
cellulose, suggesting that Ce is more effective in breaking down cellulose
2.4. Lignin and catalysts characterisation
into small oligomers/monomers including some cycloalkenes and
cycloalkanes. Presence of Ce/ZrO2 (1_1) resulted in 81 distinct com­
Proximate and ultimate analysis of the Etek lignin were obtained by
pounds, mostly Oxygenated sugars (34.7 %), Guaiacols (23.5 %) and
using an Exeter CE-440 Elemental Analyser and a Mettler Toledo TGA2
Esters (7.66 %), indicating its propension in converting cellulose-
thermogravimetric analysis, respectively. The catalysts characterisation
derived components than CeNa/ZrO2 (2_1_1). The GC–MS analysis in
included XPS spectra that were acquired using a PHI Quantera II Scan­
presence of Na/ZrO (1.5_1) identified 60 distinct compounds with Ke­
ning XPS Microprobe -instrument. Samples were sputter-cleaned using
tones + Aldehydes (23 %) and Guaiacols (42.3 %) being the most
argon ions prior to analysis. Calibration of spectra was done using the C
abundant. Basic sites of the catalyst can favour reduction of acids and
1 s peak (284.8 eV) for adventitious carbon. SEM-EDX analysis was done
deoxygenation via ketonization and aldol condensation reactions (Ste­
using a Zeiss Leo 1530 microscope equipped with a FEG and operated at
fanidis et al., 2016). In comparison, when Na/ZrO3 (1.5_1) was pyro­
a voltage of 2.5 kV and equipped with an Oxford Instruments X-Max
lysed in presence of N2, the GC–MS resulted in 47 distinct compounds
silicon drift detector (SDD) on non-coated samples. TEM analyses were
mainly Guaiacols (62.4 %) (Hendry et al., 2020). This suggests that the
performed using a Titan Themis 200 scanning/transmission electron
catalyst had good reactivity to both the cellulose and lignin components
microscope (S/TEM) equipped with an X-FEG Schottky field emission
of the biomass in presence of H2. Olefin intermediates quickly reform
gun and a Super-X high sensitivity windowless EDX detector com­
into larger molecules, eventually leading to polyaromatic hydrocarbons
plemented by a Gatan Enfinium EELS Detector. The catalyst samples for
and char (Norinaga et al., 2014). The presence of H2 gas stabilises the
TEM were ultrasonically dispersed in ethanol and then deposited on
olefin intermediates and stops further reactions that produce coke on the
carbon-coated copper grids using capillary and dried in air for 30 min.
catalysts’ surface. This also explains why the tests in presence of H2
XRD was instead run using a Bruker Nonius X8-Apex2 CCD equipped
showed much higher cellulose derived compounds.
with an Oxford Cryosystems Cryostream (typically operating at 100 K),
The GC–MS analysis with 1 wt% Pd and 5 wt% Fe on ZrO2 identified
and an X-ray source with a Cu anode working at 40 kV and 40 mA, and
62 distinct compounds, which are much closer to the desired products (i.
an energy-dispersive one-dimensional detector. The Fe amount on the
e. deoxygenated hydrocarbons) than Na/ZrO2 and the other two Na/Ce
Fe/ZrO2 catalyst was determined by atomic absorption spectroscopy
catalysts. Ketones + Ketone Alcohols, Cycloalkenes, Guaiacols, and BTX
(AAS) using a PerkinElmer Analyst 100, while surface analysis was
were the main functionalities with total area from each being 12.06 %,
carried out using a Tristal II Plus Micromeritics analyzer. H2-TPR was
18.67 %, 7.88 % and 16.31 %, respectively. The presence of poly­
carried out using a lab-made instrument (Bagnato et al., 2020). The
aromatic hydrocarbons (5.74 % - highest in any of the catalytic tests)
catalyst was first dried at 110 ◦ C for 18 h. and then heated at 10 ◦ C/min
shows that olefins created from cellulose reduction are contributing to
from 25 ◦ C to 900 ◦ C in a 5 % H2/He flow (40 mL/min). The effluent
the end products and indicate that species are over-reduced and forming
gases were analysed by a TCD detector. Surface acidity was analysed by
char deposits on the catalytic acid sites, limiting further catalytic ac­
NH3-TPD using an AutoChem II system (Micromeritic, USA).
tivity (Stefanidis et al., 2016). The higher area% of cycloalkenes and
lower yield of Ketones + Aldehyde shows that PdFe/ZrO2 is more active
and more reducing than Na/ZrO2. Likewise, the 16.3 area% of BTX and
low relative area% in Alkylphenols and Guaiacols shows the reducing
factor is equally applied to the lignin component as well as the cellulose
component.

4
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

Table 2
GC–MS results from the catalytic hydropyrolysis tests (Area%).
Functionality CeNa/ZrO2 CeNa/ZrO2 Ce/ZrO2 Na/ZrO2 PdFe/ZrO2 Fe/ZrO2 Origin
(1_1_1) (2_1_1)

Esters 1.27 0.24 7.66 0.55 0.48 – Cellulose

Ketones þ Aldehydes 11.04 19.53 5.28 23.04 12.06 15.30 Cellulose/ Lignin
Acids 0.78 0.00 2.62 0.63 7.92 – Lignin
Olefins 0.58 1.24 2.14 0.77 1.44 – Cellulose + Lignin
Cycloalkenes 5.93 7.05 4.59 10.19 18.67 13.46 Cellulose + Lignin
(Cyclopentadiene þ Cyclohexadiene) (0.38) (0.00) (0.00) (0.63) (0.90) (9.59)
Cycloalkanes 2.64 4.66 1.38 8.79 2.58 – Lignin
Alkyl Phenols – – 2.07 3.38 2.26 – Lignin
Guaiacols 22.25 40.05 23.47 42.32 7.88 – Lignin
Other Aromatics 0.86 1.42 2.33 – 0.91 – Lignin
BTX 0.26 0.48 0.49 0.42 16.31 66.82 Lignin
PAH (anthracene, indenes) 0.09 – – 0.36 5.74 2.94 Cellulose + Lignin
Oxygenated Sugar Derivative 3.72 16.62 34.70 2.51 15.44 – Cellulose
Other 50.30 8.75 11.31 6.98 7.69 1.50 Cellulose + Lignin
Furans – – 1.08 – 0.64 – Cellulose

The HyPy bio-oil obtained using Fe/ZrO2 showed the best charac­ cyclopentadiene (R10), acetone (R11) and 3-cyclopentene 1,2-diol
teristics among all the employed catalysts. The GC–MS analyser identi­ (R12) were considered as listed below:
fied only 11 distinct compounds in the product flow from the
R1: C6H6O + H2 = C6H6 + H2O (1)
hydropyrolysis- test, which was duplicated for consistency. Ketones
(acetone), Cycloalkenes and BTX were the most abundant functionalities R2: C6H6O + 2H2 = C6H8 + H2O (2)
with total area from each being respectively: 15.3 %, 13.5 %, and 66.8 %
(35.4 % benzene, 25.6 % toluene and 5.8 % xylene). This shows a very R3: C6H6O + 3H2 = C6H12O (3)
high selectivity of the catalyst compared with the other catalysts. FeZrO2
R4: C6H6O + 4H2 = C6H12 + H2O (4)
also resulted in a carbon distribution of 85.5 % C5-C10 (10.5 % C5, 41.1
% C6, 27.6 % C7 and 6.3 % C8, 3.3 % C10), which was similar to the 84 % R5: C6H6O + 10H2 = 6CH4 + H2O (5)
aromatic hydrocarbons (C6-C11) obtained by the Hy-Py/HDO of pine
sawdust at 500 ◦ C and 20 bar over a hydrocracking catalyst (20:1 cat: R6: C7H8O + H2 = C7H8 + H2O (6)
biomass wt ratio) bed maintained at 500 ◦ C (Venkatesan et al., 2020). R7: C7H8O + 4H2 = C7H14 + H2O (7)
They also showed that at 1 bar, ~63 % aromatics (34 % BTX of which
12.8 % B., 13.8 % T. and 5.8 % X.) and significant amount of Poly Ar­ R8: C7H8O + 3H2 = C7H14O (8)
omatic Hydrocarbons (PAH) (~16 %) were produced at 500 ◦ C, which
R9: C7H8O + 11H2 = 7CH4 + H2O (9)
confirm the higher selectivity of FeZrO2 on BTX under the studied
conditions. Similarly, Agarwal et al. (2017) produced alkylphenolics R10: C6H12O6 + 3.6H2 = 1.2C5H6 + 6H2O (10)
(17 wt% on lignin) and aromatics (4 wt% on lignin) for a total of 34 wt%
lignin oil, from the hydrotreating of Kraft lignin at 450 ◦ C, 100 bar H2, 4 R11: C6H12O6 + 4H2 = 2C3H6O + 4H2O (11)
hrs using Fe rich limonite catalyst (Agarwal et al., 2017). To compare the
R12: C6H12O6 + 4.8H2 = 1.2C5H8O2 + 3.6H2O (12)
BTX yield with those in literature, the weight % of the BTX (based on
starting Etek lignin) was estimated (1.5–3.6 wt%) resulting in between ΔG, ΔH, ΔS and the equilibrium constant (Kc) were calculated as
those obtained using FeZrO2 (1.3 wt%) and PdReOx/ZrO2 (5.4 wt%) function of temperature and the resulting log Kc for the above reactions
under lower temperature (350 ◦ C), where oxygenated phenolics were are shown in Fig. 2, where values of Kc larger than zero indicate spon­
the predominant oil components (Sirous-Rezaei and Park, 2020). taneity and larger the number greater the thermodynamic viability. As
However, proper BTX quantification would be required.to confirm this. can be seen for the two phenolic species, in general lower the temper­
Moreover, the deoxygenation power resulted the best with only about ature more thermodynamically favourable are the reactions R1 to R10
5.5 wt% O in the HyPy oil compared to 15.5 wt% for FePd/ZrO2, based (due to exothermicity of all these reactions) and only the methanation
on GC–MS HyPy oil composition and larger for the other catalysts, which and HDO to benzene and toluene are thermodynamically favoured at
confirms the selective recovery of C ad H in the HyPy oil. temperature between 500 and 600 ◦ C as used in this work, with the first
being more favourable. This is in agreement with the experimental
findings with the exception of cyclohexadiene (R2), where only benzene
3.2. Reaction pathways and toluene were detected. Typically, the phenolic ring is hydrogenated
to form cyclohexanol or cyclohexanone at low temperature (as sup­
This study proposes a reaction pathway leading to BTX considering ported by the Kc in Fig. 2) and this explains the absence of these products
the already proposed mechanisms for the cracking and reforming of in this work. The HDO reactions of glucose (R10-12) are instead endo­
biomass. To evaluate if the main products distribution from the HyPy of thermic and therefore favoured at higher temperature as shown in Fig. 2,
Etek lignin was dictated by the thermodynamical favourability, the main where it can be seen that they are thermodynamically favourable under
reactions for three representative model compounds (phenol and 2- the studied temperature. If acetone (R11) is common to both Fe/ZrO2
methoxy-phenol for lignin and glucose for cellulose) were simulated and PdFe/ZrO2, the HDO to 1,3-cyclopentadiene occurs only in presence
using HSC Chemistry 5.11, Outokumpu between 300 and 800 ◦ C, at of Fe/ZrO2 suggesting that Fe is more prone to deoxygenating cellulose.
ambient pressure. For phenol and 2-methoxy-phenol the thermody­ Due to the reducing nature of the H2 atmosphere, it is expected that
namic viability of hydrodeoxygenation to benzene (R1), cyclohexane cellulose is reduced to non-condensable gases as shown by V.K. Ven­
(R4), toluene (R6) and methyl-cyclohexane (R7), the hydrogenation to katakrishnan et al (2014), with the remaining contribution from cellu­
cyclohexadiene (R2), cyclohexanol (R3), methyl-cyclohexanol (R8) and lose being reduced to simple chain molecules like Ketones, Aldehydes,
the methanation reactions (R5 and R9) were evaluated. For the cellulose Olefins as well Cyclopentenes (Venkatakrishnan et al., 2014). Recently,
fraction of Etek lignin instead, the HDO of glucose to 1,3-

5
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

3.3. FeZrO2 catalyst characterisation

The GC–MS analysis clearly indicates that FeZrO2 can selectively

produce BTX from the HyPy/HDO of Etek lignin under the studied
conditions. Therefore, the properties of this catalyst were examined by
different techniques to elucidate the reasons beyond it and correlate
them to the proposed reactions mechanismss. AAS confirmed that the Fe
content in the FeZrO2 catalyst was 4.98 wt%, while the surface analyses
resulted in a BET surface of 65 m2/g, a pore volume of 0.26 cm3/g and a
BJH average pore size of 20–30 nm (see supplementary material). The
relatively large volume of mesopores represent an important factor for
the promotion of hydrotreating reactions, since it favours the diffusion
of large molecules (e.g. guaiacols) into the pores where they can then
react in the active sites. Guaiacol HDO in presence of Pd/meso-ZSM-5
catalyst exhibited superior guaiacol conversion and product distribution
when compared with Pd supported on conventional microporous ZSM-5,
due to the improved diffusion and accessibility of active sites inside
meso-ZSM-5 (Wang et al., 2020). X-ray powder diffraction pattern of Fe-
doped ZrO2 catalyst (see supplementary material) appears as a very well
crystalline mixture of different zirconium oxide phases. ZrO2 monoclinic
patters typical at ambient temperature (2-θ of 28.2◦ , 31.5◦ , 38.5◦ , 50.1◦
and 59.8◦ ) are visible, although additional XRD peaks (30.4◦ , 31.4◦ and
35.4◦ ) suggest the presence of other ZrO2 phases. ZrO2 has been shown
to be a good support for hydropyrolysis reactions due to its thermal
stability and surface acidity (Li, Su, et al., 2022), XPS spectra of the
FeZrO2 catalyst were taken (see supplementary material). The Fe 2p
binding energy (BE) bands indicate presence of Fe3O4 (711 and 715 eV)
and Fe2O3 (725 eV) (Bagnato et al., 2020), while the Zr3d bands indicate
presence of ZrO2 (182.5 eV and 183.9 eV) and ZrOx suboxide (180.1,
182.5 eV). Finally, the O1s spectra show the oxygen corresponding to
ZrO2 (531 and 532.8 eV), ZrOx (530.6 eV) and Fe species (528.9 eV).
SEM-EDX (see supplementary material), which was used to establish the
overall size of the particles and the content in iron, suggests particles in
the 10–50 nm size with presence of some larger agglomerates and
confirm that the weight content of iron close to 5 wt% as per synthesis.
TEM (see supplementary material) confirms that the FeZrO2 particles
are of 20–30 nm size, but do not allow for a clear distinction between Zr
and Fe due to poor contrast difference, although the darker area of ~ 5
Fig. 2. Equilibrium constants (Kc) as a function of temperature for hydro­
nm can be assigned to Fe. It is well known that a good dispersion of
deoxygenation (HDO), hydrogenation, hydrogenation/HDO and methanation
reactions (R1-R11).
metals on the support surface is related to good catalytic activity
(Rhodes et al., 2005). Therefore, to evaluate the dispersion of Fe on the
ZrO2 support, TEM-EDX (see supplementary material) was analysed.
Li et al (2022) showed that cellulose hydropyrolysis under 25 bar
The TEM-EDX show a good dispersion of the Fe on the ZrO2 support and
hydrogen and 500 ◦ C mainly produced C5-C7 ketones with carbon yield
confirmed that the Fe nanoparticles are well dispersed in the rage of 5
of 27.2 % (Li, Miao, et al., 2022). In particular, C5-C7 chain ketones and
nm, which can be linked to the catalyst performance. In this context, the
cyclic ketones were generated by furfural through HDO and hydro­
low selectivity of PdFe vs Fe could be linked to the poor dispersion of Pd
cracking reactions. The olefin-radical pathway for cellulose to BTX
as shown by TEM-EDX (see supplementary material). The acidity, ba­
conversion is not prevalent in H2 atmosphere because the presence of H2
sicity and H2 reducibility of FeZrO2 were also analysed. The H2-TPR (see
will protonate and stabilise the radicals and prohibit further aromati­
supplementary material) showed an H2 uptake of 280 μmol/g with
sation reactions from cellulose derivatives. However, it is assumed that
reduction to Fe0 occurring between 450 and 600 ◦ C with peak at 550 ◦ C,
cyclopentadiene is formed via this route.
which confirm that reported in a previous work (Bagnato et al., 2020;
Various reaction pathways have been previously proposed to explain
Hendry et al., 2020; Spreitzer & Schenk, 2019). NH3-TPD resulted (see
the mechanism for lignin hydropyrolysis and bio-oil vapour HDO up­
supplementary material) in an acidity of about 260 µmol NH3/g, which
grade into BTX (Jan et al., 2015; Liu et al., 2019; Sirous-Rezaei et al.,
consisted of weak and mild acid sites denoted by desorption of NH3 from
2018; Sirous-Rezaei and Park, 2020). It involves the hydrogenation of
100 to 400 ◦ C, while FePdZrO2 had considerably lower acidity as shown
the lignin structure during pyrolysis into phenolic monomers, including
in previous work (Bagnato et al., 2020). These findings support the fact
guaiacols, which then are reduced to BTX. At high temperature as used
that under the studied conditions PdFeZrO2 performance was not good
in this work, phenol deoxygenation occurs either via decarbonylation or
as the mono metallic FeZrO2 catalyst and suggest that the selectivity and
dehydration of the alcohol group (RC), resulting in mono-aromatic hy­
deoxygenation capability of FeZrO2 is due to the synergic activity of
drocarbons (MAHs). Over acid sites, these MAHs may undergo alkyl­
weak and mild acid sites in bonding the carbonyl group and the capacity
ation, cyclization, minor rates of hydrogenation to form cycloalkanes
of Fe of adsorb H2 at high temperature and hydrogenating the aromatic
and polymerization (RD) to form naphthalenes and poly-aromatic hy­
ring. It was recently shown that aromatic rings are repulsed by oxyphilic
drocarbons (Gamliel et al., 2018). Based on previous literature, it is
Fe/ZrO2, which instead adsorb and hydrogenate the carbonyl group
expected that coking reactions took place at 600 ◦ C (Gholizadeh et al.,
followed by deoxygenation in the ZrO2 acid sites (Yung et al., 2019; Li,
2016).
Miao, et al., 2022)). Moreover, Kumar et al. (2021) studied the influence
of the support ZrO2, for Co based catalyst at different reaction

6
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

temperature, 300–600 ◦ C, at 1 bar for the hydropyrolysis of prot lignin References

(Kumar et al., 2021). ZrO2 resulted in 29.8 % H-type phenolics at 600 ◦ C
due to enhanced demethoxylation. In addition, the Fe performance can Agarwal, S., Chowdari, R.K., Hita, I., Heeres, H.J., 2017. Experimental studies on the
hydrotreatment of kraft lignin to aromatics and alkylphenolics using economically
be ascribed to a better distribution of active Fe inside the support mes­ viable FE-based catalysts. ACS Sustain Chem. Eng. 5 (3), 2668–2678.
opores and the ability of monoaromatics to enter the mesopores to reach Bagnato, G., Signoretto, M., Pizzolitto, C., Menegazzo, F., Xi, X., ten Brink, G.H., Kooi, B.
the closely located Fe and ZrO2 active sites. A similar effect was observed J., Heeres, H.J., Sanna, A., 2020. Hydrogenation of biobased aldehydes to
monoalcohols using bimetallic catalysts. ACS Sustainable Chem. Eng. 8 (32),
in the hydrotreating of guaiacol with Ni-ZrO2 on mesosilica, where 11994–12004.
mesopores facilitated confinement of Ni and ZrO2 nanoparticles, which Biogreen Energy Syngas. (2020). Retrieved from http://www.biogreen-energy.com/syn
proximity was considered necessary for achieving a high selectivity into gas/.
Dutta, A., Sahir, A., Tan, E., Humbird, D., Snowden-Swan, L., Meyer, P., Ross, J.,
deoxygenated products HDO (Lopez et al, 2020). Sexton, D., Yap, R., 2015. Process Design and Economics for the Conversion of
Therefore, the remarkable catalytic activity and selectivity of Fe/ Lignocellulosic Biomass to Hydrocarbon Fuels—Thermochemical Research Pathways
ZrO2 was linked to the iron oxophilicity, the strong reduction potential with in situ and ex situ Upgrading of Fast Pyrolysis Vapors. NREL/TP-5100-62455.
National Renewable Energy Laboratory, Golden, CO, pp. NNL–23823.
of zero-valent iron, the good dispersion of Fe nanoparticles on the sup­
Farooq, A., Lam, S.S., Rhee, G.H., Lee, J., Khan, M.A., Jeon, B.-H., Park, Y.-K., 2022.
port and the presence of mesopores and Lewis acid sites on the zirco­ Technical benefits of using methane as a pyrolysis medium for catalytic pyrolysis of
nium oxide support, which increased the interactions between lignin Kraft lignin. Bioresour. Technol. 353, 127131.
derived phenolic molecules and the catalyst surface. Fe/ZrO2 showed Gamliel, D.P., Bollas, G.M., Valla, J.A., 2018. Two-stage catalytic fast hydropyrolysis of
biomass for the production of drop-in biofuel. Fuel 216, 160–170.
also good potential to also reduce cellulose-derived molecules to acetone Gholizadeh, M., Gunawan, R., Hu, X., Mercader, F.M., Westerhof, R., Chaitwat, W.,
and cyclopentadiene. The above suggest that Fe/ZrO2 can be used for Hasan, M.M., Mourant, D., Li, C.-Z., 2016. Effects of temperature on the
hydrodeoxygenation of biomass-derived oxygenates under the tested hydrotreatment behaviour of pyrolysis bio-oil and coke formation in a continuous
hydrotreatment reactor. Fuel Process. Technol. 148, 175–183.
conditions, which is crucial to achieve a cost-effective scale-up of the Griffin, P.W., Hammond, G.P., Norman, J.B., 2018. Industrial energy use and carbon
process. emissions reduction in the chemicals sector: A UK perspective. Appl. Energy 227,
587–602.
Hamer, M. (2020). Biomass in the balance. A sustainability framework for high-value use of
4. Conclusions bio-based raw materials. Retrieved from: https://www.ser.nl/-/media/ser/downloads
/engels/2020/biomass-in-the-balance.pdf.
The hydropyrolysis and hydrodeoxygenation of Etek lignin was Hendry, A., Åhlén, M., Fernandes, T., Cheung, O., Sanna, A., 2020. Catalytic cracking of
Etek lignin with zirconia supported metal-oxides for alkyl and alkoxy phenols
studied at 600 ◦ C in presence of zirconia supported metal catalysts under recovery. Bioresour. Technol. 317, 124008.
ambient pressure. Fe/ZrO2 led to considerably enhanced HDO of lignin- Jan, O., Marchand, R., Anjos, L.C., Seufitelli, G.V., Nikolla, E., Resende, F.L., 2015.
derived phenolics to aromatic hydrocarbons in comparison with the Hydropyrolysis of lignin using Pd/HZSM-5. Energy Fuels 29 (3), 1793–1800.
Jindal, M., Kumar, A., Kaur, R., Palla, V.C.S., Bhaskar, T., 2022. Flash hydropyrolysis of
other catalysts. The Fe/ZrO2 mild-strong reduction sites provided- good
cotton stalks: Role of temperature, metal loading, pressure for enhancement of
reducing activity, selectively hydro-deoxygenating carbonyl groups in aromatics. Bioresour. Technol. 351, 127047.
phenolic and guaiacol compounds to benzene (35.4 %), toluene (25.6 %) Kumar, A., Biswas, B., Saini, K., Kumar, A., Kumar, J., Krishna, B.B., Bhaskar, T., 2021.
and xylene (5.8 %), as the main products with a carbon distribution of Py-GC/MS study of prot lignin with cobalt impregnated titania, ceria and zirconia
catalysts. Renewable Energy 172, 121–129.
85.5 % in C5-C10 hydrocarbons. Li, T., Miao, K., Zhao, Z., Li, Y., Wang, H., Watanabe, A., Wang, K., 2022a. Understanding
cellulose pyrolysis under hydrogen atmosphere. Energy Convers. Manage. 254,
CRediT authorship contribution statement 115195.
Li, T., Su, J., Wang, H., Wang, C., Xie, W., Wang, K., 2022b. Catalytic hydropyrolysis of
lignin using NiMo-doped catalysts: Catalyst evaluation and mechanism analysis.
William Lonchay: Investigation, Formal analysis, Visualization, Appl. Energy 316, 119115.
Validation, Writing – original draft. Giuseppe Bagnato: Investigation, Liu, Y., Leahy, J.J., Grams, J., Kwapinski, W., 2019. Hydro-pyrolysis and catalytic
upgrading of biomass and its hydroxy residue fast pyrolysis vapors. Energies 12 (18),
Validation, Writing – review & editing. Aimaro Sanna: Conceptuali­ 3474.
zation, Methodology, Supervision, Validation, Writing – review & Liu, Y., Wu, S., Zhang, H., Xiao, R., 2021a. Fast pyrolysis of torrefied holocellulose for
editing. producing long-chain ether precursors in a fluidized bed. Bioresour. Technol. 341,
125770.
Liu, Q., Bai, Y., Chen, H., Chen, M., Sang, Y., Wu, K., Li, Y., 2021b. Catalytic conversion
Declaration of Competing Interest of enzymatic hydrolysis lignin into cycloalkanes over a gamma-alumina supported
nickel molybdenum alloy catalyst. Bioresour. Technol. 323, 124634.
López, M., Palacio, R., Mamede, A.S., Fernández, J.J., Royer, S., 2020.
The authors declare that they have no known competing financial
Hydrodeoxygenation of guaiacol into cyclohexane over mesoporous silica supported
interests or personal relationships that could have appeared to influence Ni–ZrO2 catalys. Microporous Mesoporous Mater. 309, 110452.
the work reported in this paper. Marker, T., Roberts, M., Linck, M., Felix, L., Ortiz-Toral, P., Wangerow, J., Maleche, E.,
2013. Biomass to Gasoline and Diesel Using Integrated Hydropyrolysis and
Hydroconversion. Retrieved from United States. https://www.osti.gov/bibli
Data availability o/1059031.
Norinaga, K., Yang, H., Tanaka, R., Appari, S., Iwanaga, K., Takashima, Y., Kudo, S.,
Data will be made available on request. Shoji, T., Hayashi, J.-I., 2014. A mechanistic study on the reaction pathways leading
to benzene and naphthalene in cellulose vapor phase cracking. Biomass Bioenergy
69, 144–154.
Acknowledgements Nowakowski, D.J., Bridgwater, A.V., Elliott, D.C., Meier, D., de Wild, P., 2010. Lignin fast
pyrolysis: Results from an international collaboration. J. Anal. Appl. Pyrol. 88 (1),
53–72.
This work was supported by Innovate UK/KTN (KTP project no. Oh, S., Lee, J., Lam, S.S., Kwon, E.E., Ha, J.-M., Tsang, D.C., Park, Y.-K., 2021. Fast
10013135). We kindly acknowledge Joe Perkins-Hall and Karen Sam, hydropyrolysis of biomass Conversion: A comparative review. Bioresour. Technol.
CDS-Analytical for Py-GC support and Aaron Naden, Department of 342, 126067.
Resende, F.L., 2016. Recent advances on fast hydropyrolysis of biomass. Catal. Today
Chemistry, University of St. Andrews for TEM-EDX analysis. 269, 148–155.
Rhodes, M.D., Pokrovski, K.A., Bell, A.T., 2005. The effects of zirconia morphology on
Appendix A. Supplementary data methanol synthesis from CO and H2 over Cu/ZrO2 catalysts: Part II. Transient-
response infrared studies. J. Catal. 233 (1), 210–220.
Sirous-Rezaei, P., Park, Y.-K., 2020. Catalytic hydropyrolysis of lignin: Suppression of
Supplementary data to this article can be found online at https://doi. coke formation in mild hydrodeoxygenation of lignin-derived phenolics. Chem. Eng.
org/10.1016/j.biortech.2022.127727. J. 386, 121348.
Sirous-Rezaei, P., Jae, J., Ha, J.-M., Ko, C.H., Kim, J.M., Jeon, J.-K., Park, Y.-K., 2018.
Mild hydrodeoxygenation of phenolic lignin model compounds over a FeReO x/ZrO
2 catalyst: zirconia and rhenium oxide as efficient dehydration promoters. Green
Chem. 20 (7), 1472–1483.

7
W. Lonchay et al. Bioresource Technology 361 (2022) 127727

Spreitzer, D., Schenk, J., 2019. Reduction of iron oxides with hydrogen—a review. Steel hydrodeoxygenation of cellulose: production of liquid fuel from biomass. Green
Res. Int. 90 (10), 1900108. Chem. 16 (2), 792–802.
Stefanidis, S.D., Karakoulia, S.A., Kalogiannis, K.G., Iliopoulou, E.F., Delimitis, A., Venkatesan, K., Prashanth, F., Kaushik, V., Choudhari, H., Mehta, D., Vinu, R., 2020.
Yiannoulakis, H., Zampetakis, T., Lappas, A.A., Triantafyllidis, K.S., 2016. Natural Evaluation of pressure and temperature effects on hydropyrolysis of pine sawdust:
magnesium oxide (MgO) catalysts: a cost-effective sustainable alternative to acid pyrolysate composition and kinetics studies. React. Chem. Eng. 5 (8), 1484–1500.
zeolites for the in situ upgrading of biomass fast pyrolysis oil. Appl. Catal. B 196, Wang, H., Li, T., Su, J., Miao, K., Wang, K., 2022. Noncatalytic hydropyrolysis of lignin in
155–173. a high pressure micro-pyrolyzer. Fuel Process. Technol. 233, 107289.
Stummann, M.Z., Høj, M., Gabrielsen, J., Clausen, L.R., Jensen, P.A., Jensen, A.D., 2021. Wang, Y., Huang, H., Baxter, N.C., Liao, Y., Zhao, Y., Wang, S., 2020. Guaiacol
A perspective on catalytic hydropyrolysis of biomass. Renew. Sustain. Energy Rev. hydrodeoxygenation over Pd catalyst with mesoporous ZSM-5 support synthesized
143, 110960. by solid-state crystallization. Catal. Today 358, 60–67.
Stummann, M.Z., Høj, M., Schandel, C.B., Hansen, A.B., Wiwel, P., Gabrielsen, J., Wang, H., Male, J., Wang, Y., 2013. Recent advances in hydrotreating of pyrolysis bio-oil
Jensen, P.A., Jensen, A.D., 2018. Hydrogen assisted catalytic biomass pyrolysis. and its oxygen-containing model compounds. ACS Catal. 3 (5), 1047–1070.
Effect of temperature and pressure. Biomass Bioenergy 115, 97–107. Welsby, D., Price, J., Pye, S., Ekins, P., 2021. Unextractable fossil fuels in a 1.5 C world.
Stummann, M.Z., Høj, M., Davidsen, B., Hansen, A.B., Hansen, L.P., Wiwel, P., Nature 597 (7875), 230–234.
Schandel, C.B., Gabrielsen, J., Jensen, P.A., Jensen, A.D., 2020. Effect of the catalyst Windt, M., Meier, D., Marsman, J.H., Heeres, H.J., de Koning, S., 2009. Micro-pyrolysis
in fluid bed catalytic hydropyrolysis. Catal. Today 355, 96–109. of technical lignins in a new modular rig and product analysis by GC–MS/FID and GC
Thangalazhy-Gopakumar, S., Adhikari, S., Gupta, R.B., Tu, M., Taylor, S., 2011. GC–TOFMS/FID. J. Anal. Appl. Pyrolysis 85 (1-2), 38–46.
Production of hydrocarbon fuels from biomass using catalytic pyrolysis under helium Yung, M.M., Starace, A.K., Griffin, M.B., Wells, J.D., Patalano, R.E., Smith, K.R.,
and hydrogen environments. Bioresour. Technol. 102 (12), 6742–6749. Schaidle, J.A., 2019. Restoring ZSM-5 performance for catalytic fast pyrolysis of
Tian, Y., Li, J., Wei, W., Zong, P., Zhang, D., Zhu, Y., Qiao, Y., 2021. Parametric effect of biomass: Effect of regeneration temperature. Catal. Today 323, 76–85.
biomass partial hydropyrolysis process in a downer reactor to co-produce high- Zheng, N., Zhang, J., Wang, J., 2017. Parametric study of two-stage hydropyrolysis of
quality tar and syngas. Bioresour. Technol. 320, 124401. lignocellulosic biomass for production of gaseous and light aromatic hydrocarbons.
Venkatakrishnan, V.K., Degenstein, J.C., Smeltz, A.D., Delgass, W.N., Agrawal, R., Bioresour. Technol. 244, 142–150.
Ribeiro, F.H., 2014. High-pressure fast-pyrolysis, fast-hydropyrolysis and catalytic Zhu, Y., Song, W., Yao, R., Zhao, Y., Xu, G., 2022. In-situ catalytic hydropyrolysis of
lignin for the production of aromatic rich bio-oil. J. Energy Inst. 101, 187–193.