Dwi at Moko 2016

G Model
CATTOD-9757; No. of Pages 7 ARTICLE IN PRESS

Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Hydrodeoxygenation of lignin-derived monomers and lignocellulose

pyrolysis oil on the carbon-supported Ru catalysts
Adid Adep Dwiatmoko a,b , Lipeng Zhou a,c , Inho Kim a,e , Jae-Wook Choi a , Dong Jin Suh a,b,d,
Jeong-Myeong Ha a,d,e,∗
a
Clean Energy Research Center, Korea Institute of Science and Technology, Seoul 02792, Republic of Korea
b
Department of Green Process and System Engineering, Korea University of Science and Technology, Daejeon 34113, Republic of Korea
c
College of Chemistry and Molecular Engineering, Zhengzhou University, 100 Kexue Road, Zhengzhou 450001, Henan, PR China
d
Green School (Graduate School of Energy and Environment), Korea University, Seoul 02841, Republic of Korea
e
Department of Clean Energy and Chemical Engineering, Korea University of Science and Technology, Daejeon 34113, Republic of Korea
a r t i c l e i n f o a b s t r a c t
Article history: Five carbon materials, including multi-walled carbon nanotubes (MWCNT), carbon aerogel (CARF), carbon
Received 30 June 2015 black (Vulcan carbon), activated carbon (AC), and graphite, were used as supports of the carbon-supported
Received in revised form 15 August 2015 Ru catalysts, and the hydrodeoxygenation of lignin-derived monomers and lignocellulose pyrolysis oil
Accepted 18 August 2015
was performed. Ru/MWCNT exhibited the highest deoxygenation activity, and the origin of the improved
Available online xxx
catalytic activity was studied. The metal dispersion, the acidity as measured by means of temperature-
programmed desorption, the pore structure, and the surface area were investigated in an effort to
Keywords:
understand the catalysis results. We observed that the quantity of accessible Ru nanoparticles on the
Guaiacol
Hydrodeoxygenation
mesopores determined the hydrodeoxygenation activity.
Ruthenium © 2015 Elsevier B.V. All rights reserved.
Carbon
Pyrolysis oil
1. Introduction removing oxygen atoms as water, alcohols, and other oxygenates

[7], producing stable, petroleum-grade fuels [8,9]. In order to
Lignocellulosic biomass is a potential alternative source for the develop reliable catalytic HDO processes and highly active cat-
production of biofuels as a replacement for petroleum [1]. This type alysts, efforts have been made using bio-oil model compounds
of biomass can be converted to the liquefied intermediates through including phenol [10], guaiacol [11,12], eugenol [13], and vanillin
biological [2] or thermochemical [3] degradation. Fast pyrolysis [4] [14], as bio-oil is a complex mixture whose contents and prop-
is one of the most promising thermochemical methods for val- erties are often difficult to understand. It is well known that
orizing lignocellulosic biomass to obtain its liquefied product or reactant–catalyst interactions play a role in the HDO of lignin-
a bio-oil. Although bio-oil is obtained from a sustainable source of derived compounds [15]. When supported catalysts can catalyze
biomass and is predicted to reduce the total amount of CO2 dur- the HDO process, the supports may also improve the catalytic activ-
ing its life cycle, its high oxygen content limits its direct use in ity. Thus, extensive studies of HDO have been performed for the
machines which use petroleum. In addition, the instability [5] and noble metals supported on solid acids, including silica [16], alumina
difficult storage [6] of oxygenate-rich bio-oil represent the major [10,17], silica–alumina [12,18], titania [19], zeolites [20], and other
challenges preventing its wider use. Therefore, the development of metal oxides. These metal oxides, however, frequently cause coke
processes which stabilize the oil and remove the oxygen content is formation because of their acidity. In addition, the large amount of
required to upgrade bio-oils to obtain petroleum-like hydrocarbon aqueous acids in the bio-oil leaches the metal oxide supports, deac-
fuels. tivating the catalyst. Considering the challenges associated with
Hydrodeoxygenation (HDO) is a process which can be used pyrolysis oil, stable supports with better resistance to water or
to upgrade bio-oil to deoxygenated hydrocarbons by catalytically acidic or basic media are required for the HDO of actual bio-oils.
Carbon is a commonly used support in catalytic applications.
It can anchor metal particles on a surface that exhibits weak
∗ Corresponding author at: Clean Energy Research Center, Korea Institute of Sci-
acid–base properties or catalytic functions [21]. In addition, its
ence and Technology, Seoul 02792, Republic of Korea.
structure is stable at high temperatures in an oxygen-free envi-
E-mail address: jmha@kist.re.kr (J.-M. Ha). ronment. Its amphoteric characteristic improves the adsorption of
http://dx.doi.org/10.1016/j.cattod.2015.08.027
0920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: A.A. Dwiatmoko, et al., Hydrodeoxygenation of lignin-derived monomers and lignocellulose pyrolysis
oil on the carbon-supported Ru catalysts, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.08.027
G Model
2 A.A. Dwiatmoko et al. / Catalysis Today xxx (2015) xxx–xxx
metal nanoparticles as well. It is also less expensive compared to 2020 instrument. The total surface area and pore volume were
alumina and silica supports, and active noble metals can be recov- determined using the BET (Brunauer–Emmett–Teller) equation.
ered by eliminating the support by burning the carbon [22]. The pore size distribution (PSD) curves were calculated by BJH
In this study, we observed the HDO activity of ruthenium cat- (Barrett–Joyner–Halenda) method. Temperature-programmed
alysts supported on different carbons, in this case multi-walled desorption analyses coupled with mass-spectroscopy (TPD-MS)
carbon nanotubes (MWCNTs), activated carbon (AC), carbon aero- were performed using a BELCAT-B catalyst analyzer (BEL Japan,
gel (CARF), Vulcan carbon, and graphite. Possible descriptors of the Inc.) with the evolution of H2 O (m/z = 18), CO (m/z = 28), and CO2
HDO catalysis process, including acidity, metal particle size, and (m/z = 44) being monitored by mass spectrometry using a BELMass
pore structure, were studied by analyzing the reaction results. Gas Analyzer (Osaka, Japan) instrument. The quantities of these
compounds were also measured using a thermal conductivity
2. Experimental detector (TCD). Prior to the analysis, the sample (ca. 50 mg) was
placed in a fixed-bed U-shaped quartz tubular micro-reactor and
2.1. Materials dried at 150 ◦ C under flowing He for 1 h. The temperature was
then increased at a rate of 10 ◦ C/min to 900 ◦ C under a flow of
All chemicals were used without further purification unless helium (50 mL/min). CO-chemisorption was performed using a
otherwise indicated. Guaiacol (2-methoxyphenol, 98.0%), phenol BELCAT-B catalyst analyzer. The catalyst powder (30 mg), after
(99%), cyclohexanone (99.8%), cyclohexanol (99%), (±)-trans- being placed in a U-shaped quartz reactor, was heated under a He
1,2-cyclohexanediol (96%), resorcinol (99%), AC (Darco® G60, flow (50 mL/min) to 350 ◦ C at a rate of 10 ◦ C/min. This was followed
−100 mesh, powder), and the MWCNTs (carbon >95%, O.D. × L by a H2 treatment at 350 ◦ C for 1 h. The powder was then cooled to
6–9 nm × 5 ␮m) were purchased from Aldrich (Milwaukee, WI, 50 ◦ C under a He flow (50 mL/min) and further treated at 50 ◦ C for
USA). 2-Methoxycyclohexanol (97%), carbon graphite (powder, 30 min in order to clean the metal surface of the residual adsorbed
−325 mesh), and ruthenium chloride (anhydrous, Ru 47.7% min.) hydrogen. The powder was subsequently exposed to pulses of 10%
were purchased from Alfa (Ward Hill, MA, USA). The formalde- (v/v) CO/He at 50 ◦ C under a He flow (30 mL/min). The number of
hyde solution (37% in water) used here was purchased from Junsei active-surface metal atoms was calculated from the measured CO
Chemical Co. Ltd. (Tokyo, Japan). Vulcan XC-72 was obtained from concentrations of the effluent gas.
Cabot Corp. (Alpharetta, Georgia, USA). Sodium carbonate (99%,
anhydrous) was obtained from Daejung Chemicals and Metals Co.
Ltd. (Siheung, Korea). Oxygen diluted in nitrogen (1%, v/v), helium,
hydrogen diluted in argon (5%, v/v), and nitrogen were purchased 2.4. Catalysis
from Shinyang Sanso (Seoul, Korea).
The hydrodeoxygenation of guaiacol was performed in a stain-
2.2. Catalyst preparations less steel autoclave reactor (∼100 mL). In a typical reaction
procedure, guaiacol (8.96 mmol), water (29 mL), and a solid cat-
CARF was prepared by the sol–gel polymerization of resorcinol alyst (20 mg) were added to the reactor. After flushing with
and formaldehyde [23]. Resorcinol (15 g) was dissolved in DI water N2 gas, the reactor was pressurized to 40 bar with H2 at room
(35.43 mL). Sodium carbonate (0.0481 g) was then added to accel- temperature. The reaction was performed at 270–300 ◦ C for 1 h
erate the dehydrogenation of resorcinol. After stirring for 15 min, a at an agitation rate of 800 rpm. After the reaction, the reac-
formaldehyde solution (22.115 g) was added slowly to the solution tor was cooled to room temperature and the mixture was
to form a sol. The resulting sol was cured in a vial at 80 ◦ C for 24 h extracted with ethyl acetate. The catalyst powder was removed
to produce a resorcinol–formaldehyde (RF) gel. Solvent exchange and the liquid product was collected. The extracted products
was performed with acetone at 50 ◦ C for 2 days. The residual sol- were identified using a gas chromatograph–mass spectrometer
vent was replaced with fresh acetone every 3 h to remove water combination (GC/MS, Agilent 7890A with 5975C inert MS XLD)
thoroughly from the pores of the wet gel. Ambient drying was then with a HP-5MS capillary column (60 m × 0.25 mm × 250 ␮m) and
performed at 50 ◦ C for 1 day. Carbon aerogel was finally obtained by was further quantified using a gas chromatograph with a HP-
carbonization of the RF gel at 800 ◦ C for 2 h. 5 wt% Ru on a carbon 5 capillary column (60 m × 0.25 mm × 250 ␮m) equipped with a
support was prepared by an incipient wet impregnation method flame ionized detector (GC/FID, HP Chemstation Series II 5890)
using an aqueous solution of ruthenium chloride. Typically, 0.108 g using ethyl guaiacol as an internal standard. As observed in our
of ruthenium chloride was dissolved in 20 mL of DI water. After previous study [12,24], guaiacol was hydrogenated to produce
stirring for a few minutes, 1 g of carbon support was mixed into 2-methoxycyclohexanol, (including both cis- and trans-isomers)
the solution. The mixture was then stirred well at ambient temper- and further deoxygenated to cyclohexanol, cyclohexanone, and
atures for 12 h. Excess water was removed using a rotavaporator cyclohexane. Phenol and catechol were also obtained via the
at 70 ◦ C in a vacuum. The solid residue obtained was subsequently hydrogenolysis of the methyl–oxygen bond of the methoxy group
dried in an oven at 105 ◦ C for 12 h, crushed and calcined in air at and the further elimination of the hydroxyl group. Ring-opening
400 ◦ C for 2 h, and further reduced in hydrogen diluted in argon reaction products, including cyclopentane and methylcyclopen-
(5%, v/v) at 400 ◦ C for 4 h. The obtained catalyst was passivated tane, also formed. All of these compounds were then classified into
under oxygen diluted in nitrogen (1%, v/v) at room temperature for three groups, i.e., compounds containing no oxygen atom (0-Os),
30 min. those containing one oxygen atom (1-Os), and those containing
two oxygen atoms (2-Os). The conversion of guaiacol (XGUA ), the
2.3. Characterizations product yields (Yi ), and the oxygen removal (percent of oxygen
eliminated during the reaction depicting the efficiency of HDO
Transmission electron micrographs (TEM images) were reaction) were calculated according to a modified method [25], as
obtained using a Tecnai G2 F20 device at 200 kV. The powder X-ray follows:
diffraction (powder XRD) results were obtained using a Rigaku
X-ray diffractometer with a CuK␣1 ( = 0.15406 nm) radiation f
n0GUA − nGUA
source operated at 40 kV and 30 mA. N2 adsorption–desorption Conversion (XGUA , %) = × 100
isotherms were measured at −196 ◦ C using a Micromeritics ASAP n0GUA
G Model
A.A. Dwiatmoko et al. / Catalysis Today xxx (2015) xxx–xxx 3
of the ruthenium nanoparticles were confined in the cylindrical

carbon nanotubes (Fig. S1).
3.2.2. Power X-ray diffraction

Powder X-ray diffraction (XRD) exhibited the formation of
metallic Ru particles for all of the supported Ru catalysts (Fig. 3).
Along with the sharp crystalline peak of graphite at 2 = 26.2◦
[26] and the broad scattering peak of amorphous carbons at
2 = 20◦ –30◦ [26–28], the diffraction peaks of metallic Ru parti-
cles were observed at 2 = 38.4, 42.2, and 43.9◦ , corresponding to
the (1 0 0), (0 0 2), and (1 0 1) planes of the hexagonal Ru structure.
Ru particle size calculated using Ru (1 0 1) peaks at 2 = 43.9◦ was
5–15 nm (Table 1). Crystalline Ru oxide was not observed by XRD,
although its formation cannot be excluded because of exposure of
the Ru particles to the ambient air.
3.2.3. N2 -physisorption
The pore structures of the supports, which may adjust the
Fig. 1. Hydrodeoxygenation of guaiacol using CARF-supported Ru, Ni, Pt, and Pd.
A mixture of guaiacol (GUA, 8.96 mmol), catalyst (20 mg), water (29 mL), and H2 structures of the deposited metal nanoparticles and their catalytic
(40 bar at room temperature) was stirred (800 rpm) at 250 ◦ C for 1 h. activity, were observed with N2 -physisorption (Fig. S2 and Table 1).
While AC, MWCNT, Vulcan carbon, and CARF exhibited type IV
behavior with hysteresis generated by the capillary condensation
ni of N2 , graphite exhibited type II behavior, indicating its poor poros-
Yield (Yi , %) = mi × 100
0 ity. The BET surface area (SBET ) of the carbons decreased in the order
mGUA nGUA
of AC > CARF > MWCNT > Vulcan carbon > graphite. The BET surface
area of graphite was negligible. The total pore volumes (Vt ) of the
Y
i i
· ai
Oxygen removal (%) = XGUA × 1− supports decreased in the order of MWCNT > CARF > AC > Vulcan
XGUA · 2 carbon ≈ graphite. From the BJH desorption pore size distributions
(Fig. S2), CARF exhibited distinct pores of 4–10 nm while the Vul-
f
where n0GUA is the initial mol of guaiacol, nGUA is the final mol of gua- can carbon and graphite had negligible pore sizes. AC exhibited
iacol, ni is the mol of the i-product in the liquid phase, mi and mGUA mostly micropores. MWCNT exhibited mesopores 10–100 nm in
are the stoichiometric coefficients during the guaiacol conversion size because of the interparticle spaces between the nanotubes.
reaction into the i-product, ai is the number of oxygen atoms in the
molecule of the i-product, and i-product excludes guaiacol. 3.2.4. Temperature-programmed desorption for oxygen
functionalities
3. Results and discussion The quantity of oxygen functionalities on the carbon supports
was measured by means of temperature-programmed desorption
3.1. Hydrodeoxygenation of guaiacol on several supported metal coupled with mass spectrometer (TPD-MS), which are responsi-
catalysts ble for the surface acids and bases (Table 1, Fig. 4 and Fig. S3)
[29]. Oxygen functionalities including carboxylic acids, carboxylic
The hydrodeoxygenation (HDO) of guaiacol was performed anhydrides, phenols, ethers, lactones, and quinones are predicted
using Ru, Pd, Pt, and Ni supported on CARF (Table S1) to find to be decomposed into H2 O and CO2 at lower temperatures and
the most active metals for the reaction (Fig. 1 and Table S2). The CO at higher temperatures [30,31]. The desorbed water observed at
TON, turnover number = (mol of converted guaiacol)/(metal sur- 50–600 ◦ C can be attributed to the physisorbed water on the carbon
face area), of guaiacol and its oxygen removal depend on the surface and the dehydrated water formed by the condensation of
deposited metals. Ru/CARF exhibited the highest TON (3.1 mol/m2 ) two neighboring carboxylic acids to produce anhydride and water.
and oxygen removal (46.4%), while the others exhibited the activity The desorption of CO2 occurred at 150–750 ◦ C, while that of CO was
of Ru Ni > Pt ≈ Pd. Upon further observations of the HDO prod- observed up to 950 ◦ C. For all carbons in this study, the amount of
ucts on Ru/CARF, 2-methoxycyclohexanol and cyclohexanol were CO was greater than that of CO2 . The total quantity of oxygen func-
found to be the major products, whereas phenol was less preferred. tionalities decreased in the order of AC > CARF > MWCNT > Vulcan
Ni/CARF, however, produced more phenol compared to Ru/CARF, carbon graphite, which is nearly identical to the order of the BET
indicating that the hydrogenation of phenyl of guaiacol was less surface area (Fig. 5). The similarity between the total quantity of
preferred on Ni/CARF. Based on these observations, Ru was selected oxygen functionalities and the BET surface area indicates that the
as the metal to be deposited on the supports in this study. number of oxygen functionalities per surface area does not differ
significantly.
3.2. Characterizations of carbon-supported Ru catalysts
3.2.5. CO-chemisorption
3.2.1. High-resolution transmission electron microscopy The CO-chemisorption results exhibited that ruthenium
Based on the results depicted at Fig. 1, ruthenium catalysts nanoparticles were well dispersed on AC, CARF, Vulcan car-
supported on carbons, including multi-walled carbon nanotubes bon, and MWCNT with [CO]/[Ru] = 0.073–0.189 (mol/mol),
(MWCNTs), activated carbon (AC), carbon aerogel (CARF), Vulcan whereas they were not well dispersed on graphite with
carbon, and graphite, were prepared, and their formation was [CO]/[Ru] = 0.017 (mol/mol) (Table 1). The low dispersion of
observed using high-resolution TEM (HR-TEM) (Fig. 2). Uniformly Ru/graphite, which is not predicted from the small particles
small Ru particles of 5–10 nm were observed for all of the carbon observed in Fig. 1, can be attributed to the agglomeration of Ru
supports prepared in this study. Particularly for Ru/MWCNT, TEM particles on the carbon surface, which is not displayed in the TEM
images with tilt angles adjusted at 32◦ demonstrated that some images. The supports with larger BET surface areas exhibited wider
G Model
Fig. 2. Representative HR-TEM images of ruthenium supported on (a) AC, (b) vulcan carbon, (c) MWCNT, (d) CARF, and (e) graphite.
dispersions (Fig. 5). Both the dispersion of metal nanoparticles and and Ru/CARF exhibited lower oxygen removals compared to that
the number of oxygen functionalities appeared to increase with an of Ru/MWCNT, although they exhibited a larger Ru dispersion
increase in the BET surface area, which indicates that the surface ([CO]/[Ru]) with more oxygen functional groups on the carbon sup-
chemistry of these carbon supports do not differ significantly from ports (Table 1). The conversions of guaiacol and oxygen removals
one another. were 96.1 and 65.6% for Ru/AC and 81.7 and 48.0% for Ru/CARF,
respectively. In addition, Ru/MWCNT produced significantly more
3.3. Hydrodeoxygenation of guaiacol using carbon-supported Ru fully deoxygenated cyclohexane compared to the other carbon-
supported Ru catalysts (Table S3). The yields of cyclohexane were
The catalytic activity, depicted by the oxygen removal, 34.9, 24.7, and 11.7% for Ru/MWCNT, Ru/AC, and Ru/Vulcan carbon,
for the HDO of guaiacol was measured using carbon- respectively.
supported Ru catalysts prepared in this study (Fig. 6). The origin of the higher HDO activity of Ru/MWCNT was inves-
The activities of supported Ru decreased in the order of tigated. The hydrogenation activity of metal catalysts is usually
Ru/MWCNT > Ru/AC > Ru/CARF > Ru/Vulcan carbon > Ru/graphite. proportional to the number of accessible metal atoms [32]. When
Ru/MWCNT exhibited the highest guaiacol conversion of 98.1% the dependence of oxygen removal on the dispersion of Ru was
and the highest oxygen removal of 81.6%. Interestingly, Ru/AC investigated, the larger accessible Ru surface was not responsible
G Model
Table 1
Pore structures and surface chemistry of the carbon-supported Ru catalysts.
Catalyst Ru/AC Ru/Vulcan Ru/MWCNT Ru/CARF Ru/graphite
SBET (m2 /g) 733.6 206.3 236 553.8 8.1

At (m2 /g) 599.1 115.7 18.2 215.5 7.9
N2 -physisorptiona Am /At 0.2 0.8 12 1.6 0.03
Dp (nm) <2 Negligible 10–100 4–10 Negligible
Vt (cm3 /g) 0.29 0.06 1.02 0.71 0.03
[CO]/[Ru] (%, mol/mol) 18.9 7.3 9.9 7.6 1.7

CO-chemisorption
Ru diameter (nm) 4.8 12.3 9.1 17.5 52.9
XRD Ru diameterc (nm) (FWHM (◦ )) 9.7 (0.9053) 7.0 (1.238) 9.7 (0.9053) 15.3 (0.5940) 6.7 (1.2911)
H2 O (mmol/g) 0.09 0.01 0.01 0.04 0.002

Amount of oxygen atoms on the CO2 (mmol/g) 0.13 0.02 0.01 0.15 0.003
carbon surfaceb CO (mmol/g) 1.36 0.11 0.45 0.81 0.04
Totald (mmol/g) 1.71 0.16 0.48 1.15 0.05
a
Carbon support only. SBET : BET surface area; At : micropore surface area; Am : mesopore surface area; Dp : BJH desorption peak pore diameter; Vt : total pore volume.
b
Measured using TPD using TCD and MS detectors.
c
Calculated by the Scherrer equation using Ru (1 0 1) peaks at 2 = 43.9◦ .
d
Total amount of oxygen groups = 2 × [CO2 ] + [H2 O] + [CO].
for the deoxygenation activity; Ru/MWCNT exhibited a greater oxy-

gen removal than the most dispersed Ru/AC (Fig. 7). Because the
acid properties of carbon surfaces are not significantly different,
the pore structures of these catalysts may determine the catalytic
activity. From the pore structures of the carbons, 82% of the BET
surface area of AC was attributed to micropores, but only 8% of the
BET surface of MWCNT was attributed to micropores. The ratio of
mesopore to micropore surface areas (Am /At ) of MWCNT was much
higher than that of AC (Table 1). With the uniformly distributed
particles through the micro- and mesopores, the poor accessibility
of the reactants to Ru particles deposited on the micropores may
suppress the catalytic activity, leading to lower activity of Ru/AC
compared to that of Ru/MWCNT, even with its wider Ru disper-
sion. Based on these observations, it is clear that the deoxygenation
activity was strongly dependent on the accessible Ru surface area.
The dependence of HDO activity on the acidity, as measured by
desorbed CO2 and CO in TPD, was also considered, but no clear
relationships were found.
Fig. 4. TPD results of the carbon supports.
3.4. HDO of lignin-derived monomers and dimers on Ru/MWCNT
The catalytic activity of Ru/MWCNT was observed for the HDO Eugenol (EUG), cerulignol (CER), and vanillin (VAN) were fully con-
of other possible phenolic compounds in order to confirm its use verted into their corresponding cycloalkanes and cyclic alcohols.
for the HDO of the lignin-derived phenolic mixture (Table 2). The compositions of product mixtures, including cyclic alcohols,
cycloketones, and cycloalkanes in this case, depend on the starting
reactants (Table 2 and Table S4). High selectivities to fully deoxy-
genated compounds were obtained for propylcyclohexane (28.3%)
Fig. 3. X-ray diffraction pattern of ruthenium supported on (a) AC, (b) Vulcan carbon, Fig. 5. Relationship between the BET surface area of the supports, the amount of
(c) MWCNT (d) CARF, and (e) graphite. Inset: graphite with strong diffraction peaks. oxygen on the surfaces of the supports, and the dispersion of the supported Ru.
G Model
Fig. 6. Effects of carbon supports on the hydrodeoxygenation of guaiacol. A mixture

of guaiacol (GUA, 8.96 mmol), a catalyst (5 wt.% Ru/carbon, 10 mg), water (29 mL),
Fig. 7. Oxygen removal depending on the dispersion of Ru.
and H2 (40 bar at room temperature) was stirred (800 rpm) at 270 ◦ C for 1 h.
from EUG and for propylcyclohexane (39.5%) from CER. Further-

more, 1-(cyclohexylmethyl)-3-methylcyclohexane (12.7%), which a fluidized bed (Table S6). The HDO reaction was performed at
may form because of the condensation of two phenolic compounds, 300 ◦ C by mixing the bio-oil (1.276 g), water (40 mL), and the cat-
and cyclohexane (10.0%) were obtained from VAN. Compared to alyst (0.6 g). A color change was observed when the brown bio-oil
the other lignin phenolic monomers, however, the conversions of changed to a colorless liquid (Fig. S4). The compositions of the
1,3,5-trimethoxybenzene (3MB) were as low as 9.7% (in water) raw bio-oil and the liquid product were determined by GC–MS
and 18.8% (in n-decane) probably because of its bulky molecular (Fig. S5 and Table S6), and organic compounds, including acids,
structure. The main products from 3MB were methoxycyclohexane alcohols, aldehydes, and alkanes, were identified (Table 3). The
(25.8%) and 1,3-dimethoxybenzene (30.7%). These results suggest composition of organic acids, including acetic acid and propionic
that the catalyst can be applied to the hydrodeoxygenation of differ- acid, decreased from 13.8 to 0.6% after the reaction. Similarly, alde-
ent types of phenolic monomers when upgrading bio-oil phenolic hydes, phenols, and sugars disappeared, while alcohols (23.4%),
components. alkanes/alkenes (13.4%), and cycloalkanes (22.2%) formed during
the reaction. Based on the peak areas, the cycloalkane products cor-
3.5. HDO of water-soluble bio-oil on Ru/MWCNT respond to substituted cyclohexane and cyclopentane (Table S6).
These results indicate that the Ru/MWCNT was highly active for the
The HDO of bio-oil was performed using Ru/MWCNT. The bio- upgrading of bio-oil toward alkanes/cycloalkanes in the presence
oil reactant was obtained by the pyrolysis of oak chips using of water.
Table 2
Hydrodeoxygenation of lignin monomers on Ru/MWCNT.a
Phenolic lignin monomer Conversion (%) Selectivity (%)b
100 Others
EUG 28.3 14.1 15.3 22.1 20.2
99 Others
CER 39.5 20.7 28.2 11.7
100 Others
VAN 10.0 32.7 27.9 12.7c 16.7
9.7 Others
3MB 3.4 9.7 24.2 26.9 35.8
18.8 Others
3MB (in decane) 7.0 8.2 25.8 30.7 28.3

a
Reaction conditions: phenolic monomers (8.96 mmol) or vanillin (1.8 mmol), catalyst (20 mg), water (29 mL), initial H2 pressure at room temperature = 40 bar, 270 ◦ C,
1 h, stirred at 800 rpm.
b
Selectivity was calculated based on peak area.
c
Three geometric isomers exist.
G Model
Table 3 Appendix A. Supplementary data

Compositions of (a) the raw bio-oil and (b) the upgraded liquid product.a
Group Area % Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.cattod.2015.08.
Raw bio-oil Upgraded liquid product
027.
Acids 13.8 0.6
Alcohols 1.3 23.4
Aldehydes 11.3 References
Alkanes/alkenes 0.6 13.4
Cycloalkanes 22.2 [1] S. Nanda, P. Mohanty, K.K. Pant, S. Naik, J.A. Kozinski, A.K. Dalai, BioEnergy
Cycloketones 3.8 Res. 6 (2013) 663–677.
Esters/ethers 6.0 3.7 [2] Y. Xia, H.H.P. Fang, T. Zhang, RSC Adv. 3 (2013) 15528–15542.
Furans/pyrans 6.0 9.3 [3] H.J. Huang, X.Z. Yuan, H.N. Zhu, H. Li, Y. Liu, X.L. Wang, G.M. Zeng, Energy 56
(2013) 52–60.
Ketones 4.2 6.2
[4] A.V. Bridgwater, Biomass Bioenergy 38 (2012) 68–94.
Phenols 14.2
[5] X.X. Jiang, N. Ellis, Energy Fuels 24 (2010) 2699–2706.
Sugars 7.8 [6] A.K. Hossain, P.A. Davies, Renew. Sustain. Energy Rev. 21 (2013) 165–189.
Others (<0.5%) 31.0 21.3 [7] A.V. Bridgwater, Environ. Prog. Sustain. Energy 31 (2012) 261–268.
a
Reaction conditions: A mixture of oak oil (1.276 g), water (40 mL), catalyst [8] H.M. Wang, J. Male, Y. Wang, ACS Catal. 3 (2013) 1047–1070.
[9] A.R. Ardiyanti, S.A. Khromova, R.H. Venderbosch, V.A. Yakovlev, H.J. Heeres,
(5 wt.% Ru/MWCNT, 0.6 g), and H2 (50 bar at room temperature) was stirred
Appl. Catal. B: Environ. 117 (2012) 105–117.
(800 rpm) at 300 ◦ C for 2 h. After reaction, the mixture was extracted with ethyl-
[10] A. Popov, E. Kondratieva, L. Mariey, J.M. Goupil, J. El Fallah, J.P. Gilson, A.
acetate (25 mL).
Travert, F. Mauge, J. Catal. 297 (2013) 176–186.
[11] M.V. Bykova, D.Y. Ermakov, V.V. Kaichev, O.A. Bulavchenko, A.A. Saraev, M.Y.
Lebedev, V.A. Yakovlev, Appl. Catal. B 113 (2012) 296–307.
4. Conclusion [12] C.R. Lee, J.S. Yoon, Y.-W. Suh, J.-W. Choi, J.-M. Ha, D.J. Suh, Y.-K. Park, Catal.
Commun. 17 (2012) 54–58.
The hydrodeoxygenation of phenolic monomers and bio-oil [13] J. Horacek, G. St’avova, V. Kelbichova, D. Kubicka, Catal. Today 204 (2013)
38–45.
was performed using ruthenium catalysts supported on different
[14] M.T. Jimare, F. Cazana, A. Ramirez, C. Royo, E. Romeo, J. Faria, D.E. Resasco, A.
carbons, in this case multi-walled carbon nanotubes (MWCNTs), Monzon, Catal. Today 210 (2013) 89–97.
activated carbon (AC), carbon aerogel (CARF), Vulcan carbon, and [15] X.H. Zhang, T.J. Wang, L.L. Ma, Q. Zhang, Y.X. Yu, Q.Y. Liu, Catal. Commun. 33
graphite. The pore structure, acidity, and the sizes of Ru metal (2013) 15–19.
[16] R.N. Olcese, M. Bettahar, D. Petitjean, B. Malaman, F. Giovanella, A. Dufour,
particles were investigated to find a descriptor with which to ascer- Appl. Catal. B 115–116 (2012) 63–73.
tain the HDO of guaiacol, while the metal particle size with the [17] J. Payormhorm, K. Kangvansaichol, P. Reubroycharoen, P. Kuchonthara, N.
pore structure determined the HDO activity of carbon-supported Hinchiranan, Bioresour. Technol. 139 (2013) 128–135.
[18] J.S. Yoon, Y. Lee, J. Ryu, Y.-A. Kim, E.D. Park, J.-W. Choi, J.-M. Ha, D.J. Suh, H.
Ru catalysts. The acidity, previously reported as being the most Lee, Appl. Catal. B 142–143 (2013) 668–676.
important factor [12], did not significantly adjust the HDO activ- [19] S.L. Wan, T. Pham, S. Zhang, L. Lobban, D. Resasco, R. Mallinson, AIChE J. 59
ity and the accessible Ru particles were more important probably (2013) 2275–2285.
[20] I. Graça, J.M. Lopes, M.F. Ribeiro, F. Ramôa Ribeiro, H.S. Cerqueira, M.B.B. de
because the same carbon materials were used as supports for Almeida, Appl. Catal. B 101 (2011) 613–621.
the same Ru nanoparticles. Ru/MWCNT was the most active cat- [21] B. Yan, S.Y. Huang, S.P. Wang, X.B. Ma, ChemCatChem 6 (2014) 2671–2679.
alyst for the HDO of guaiacol, with the result achieved with the [22] E. Lam, J.H.T. Luong, ACS Catal. 4 (2014) 3393–3410.
[23] H.W. Park, U.G. Hong, Y.J. Lee, I.K. Song, Appl. Catal. A 409 (2011) 167–173.
combination of the high Ru surface area and the high external [24] A.A. Dwiatmoko, S. Lee, H.C. Ham, J.-W. Choi, D.J. Suh, J.-M. Ha, ACS Catal. 5
surface area of the MWCNTs. The successful hydrodeoxygena- (2015) 433–437.
tion of bio-oil and other phenolic compounds was observed using [25] M.V. Bykova, D.Y. Ermakov, S.A. Khromova, A.A. Smirnov, M.Y. Lebedev, V.A.
Yakovlev, Catal. Today 220 (2014) 21–31.
Ru/MWCNT.
[26] Z. Kowalczyk, S. Jodzis, W. Raróg, J. Zieliński, J. Pielaszek, A. Presz, Appl. Catal.
A 184 (1999) 95–102.
Acknowledgments [27] E. Antolini, F. Cardellini, J. Alloys Compd. 315 (2001) 118–122.
[28] X. Zheng, S. Zhang, J. Xu, K. Wei, Carbon 40 (2002) 2597–2603.
[29] H.P. Boehm, Carbon 40 (2002) 145–149.
This work was supported by the National Research Foundation [30] P. Brender, R. Gadiou, J.C. Rietsch, P. Fioux, J. Dentzer, A. Ponche, C. Vix-Guterl,
of Korea (NRF) grant funded by the Korean Government (MSIP) Anal. Chem. 84 (2012) 2147–2153.
(NRF-2009-C1AAA001-0093293). [31] D. Rosenthal, M. Ruta, R. Schlogl, L. Kiwi-Minsker, Carbon 48 (2010)
1835–1843.
This work was also supported by the Brain Pool Program of the [32] J. Chupin, N.S. Gnep, S. Lacombe, M. Guisnet, Appl. Catal. A 206 (2001) 43–56.
MSIP (Ministry of Science, ICT & future Planning).

Dwi at Moko 2016

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dwi at Moko 2016

Uploaded by

Copyright:

Available Formats

G Model

CATTOD-9757; No. of Pages 7 ARTICLE IN PRESS

Contents lists available at ScienceDirect

Hydrodeoxygenation of lignin-derived monomers and lignocellulose

1. Introduction removing oxygen atoms as water, alcohols, and other oxygenates

of the ruthenium nanoparticles were conﬁned in the cylindrical

3.2.2. Power X-ray diffraction

Catalyst Ru/AC Ru/Vulcan Ru/MWCNT Ru/CARF Ru/graphite

SBET (m2 /g) 733.6 206.3 236 553.8 8.1

[CO]/[Ru] (%, mol/mol) 18.9 7.3 9.9 7.6 1.7

H2 O (mmol/g) 0.09 0.01 0.01 0.04 0.002

for the deoxygenation activity; Ru/MWCNT exhibited a greater oxy-

3.4. HDO of lignin-derived monomers and dimers on Ru/MWCNT

Fig. 6. Effects of carbon supports on the hydrodeoxygenation of guaiacol. A mixture

from EUG and for propylcyclohexane (39.5%) from CER. Further-

Phenolic lignin monomer Conversion (%) Selectivity (%)b

EUG 28.3 14.1 15.3 22.1 20.2

CER 39.5 20.7 28.2 11.7

VAN 10.0 32.7 27.9 12.7c 16.7

3MB 3.4 9.7 24.2 26.9 35.8

3MB (in decane) 7.0 8.2 25.8 30.7 28.3

Table 3 Appendix A. Supplementary data

You might also like