Hydrodeoxygenation of Sorbitol into Bio-Alkanes and -Alcohols Over Phosphated Ruthenium Molybdenum Catalysts

Article abstract of DOI:10.1002/cctc.201801214

Biofuels such as renewable alkanes and higher alcohols have drawn considerable interests for the use in internal combustion engines. Especially, higher alcohols could be used as a blending agent for diesel fuels. Herein, carbon supported phosphated ruthenium-molybdenum (RuMoP) catalysts were employed in continuous trickle-bed reactor for converting sorbitol into renewable alkanes and higher alcohols. The results showed that RuMoP on an active carbon (AC) support presented a complete sorbitol conversion and high yields of alkanes and alcohols in gasoline and diesel range. Subsequently, carbon nanotube (CNT) supported RuMoP was prepared and studied in detail for comparison. RuMoP/CNT presented a low C?C bond cracking property in sorbitol conversion and high selectivity of C6 products in gas-phase (C6 alkane, 74.7 %) and oil-phase (C6 alkane and alcohols, 87.8 %). Finally, detailed characterizations (N₂-adsorption, XRD, HRTEM, XPS, NH₃-TPD, Py-IR spectrums, etc.) were performed over relevant catalysts (RuMoP/C and RuMoP/CNT) for correlating their catalytic and physicochemical properties.

Full text of DOI:10.1002/cctc.201801214

Accepted Article
Title: Hydrodeoxygenation of sorbitol into bio-alkanes and -alcohols
over phosphated ruthenium molybdenum catalysts
Authors: Yujing Weng, Tiejun Wang, Chenguang Wang, Qiying Liu,
Yulong Zhang, Peigao Duan, Longlong Wang, Hongxing Yin,
Shijun Liu, and Longlong Ma
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201801214
Link to VoR: http://dx.doi.org/10.1002/cctc.201801214
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
Hydrodeoxygenation of sorbitol into bio-alkanes and -alcohols
over phosphated ruthenium molybdenum catalysts
[
a, b]
[b]
[b]
[b]
[a]
[a]
Yujing Weng, *
Tiejun Wang, Chenguang Wang, Qiying Liu, Yulong Zhang, Peigao Duan,
[a]
[a]
[b]
[b]
Longlong Wang, Hongxing Yin, Shijun Liu, and Longlong Ma*
Abstract: Biofuels such as renewable alkanes and higher alcohols
have drawn considerable interests for the use in internal combustion
engines. Especially, higher alcohols could be used as a blending
agent for diesel fuels. Herein, carbon supported phosphated
ruthenium-molybdenum (RuMoP) catalysts were employed in
continuous trickle-bed reactor for converting sorbitol into renewable
alkanes and higher alcohols. The results showed that RuMoP on an
active carbon (AC) support presented a complete sorbitol conversion
and high yields of alkanes and alcohols in gasoline and diesel range.
Subsequently, carbon nanotube (CNT) supported RuMoP was
prepared and studied in detail for comparison. RuMoP/CNT presented
a low C-C bond cracking property in sorbitol conversion and high
selectivity of C6 products in gas-phase (C6 alkane, 74.7%) and oil-
phase (C6 alkane and alcohols, 87.8%). Finally, detailed
alcohols and conventional fuels. The results indicated that higher
alcohol blends can be attractive fuel additives and facilitate the
storage and transportation due to the advantages of lower O
content, vapor enthalpy, water-solubility, higher energy density,
cetane number, flash point and high heat value (HHV).[1a, 1b, 4a]
Even though many researchers reported the production of butanol
via fermentation with some success, significantly less work has
been conducted on C5 and C6 higher alcohols.[1a, 1b] Meanwhile,
C4–C6 alcohols are not produced from fermentation at
a
commercial scale since an energy intensive distillation operation
is required to obtain water-free alcohols. [ Recently, production
of C4-C6 alcohols from carbohydrates and their derivatives via
aqueous-phase hydrodeoxygenation (HDO) has been carried out,
whereby the total or partial removal of oxygen from water-soluble
starting materials through bifunctional catalysts with reactions
comprises C-O bond degradation, C-C bond cleavage, and C = C
or C = O hydrogenation.[1b, 1c, 5] Therefore, bifunctional catalysts
with metal and acid sites are significantly needed for the HDO
reactions. Among various bifunctional catalysts, solid acid-
supported noble metal catalysts have been studied as effective
HDO catalysts for biomass into mono higher alcohols. Huber et.al.,
reported the bimetallic catalysts based on Pt (e.g., Pt/Zr-P, Pd/Zr-
1b]
2 3
characterizations (N -adsorption, XRD, HRTEM, XPS, NH -TPD, Py-
IR spectrums, etc.) were performed over relevant catalysts (RuMoP/C
and RuMoP/CNT) for correlating their catalytic and physicochemical
properties.
Introduction
P, and Pt-ReO
with the best yield of 29% of C4-C6 mono-alcohols on Pt-ReO
P catalysts.[1b, 5-6] Tomishige et.al., reported biphasic (water + n-
decane) catalytic system with catalysts of Ir-ReO /SiO and
HZSM-5 or H SO for microcrystalline cellulose and xylan into
higher alcohols with the best yield of 52%. Additionally, Wang
et.al., reported the bifunctional catalytic system of Pd/NbOPO
x
/Zr-P) for aqueous phase HDO of sugar and polyol
Increasing issues on environmental pollution and rapid
consumption of fossil resources promote interests to explore
sustainable production of chemicals and fuels from renewable
resources. Compared to fossil resources (petroleum, coal, natural
gas, etc.), the feature of biomass is the higher oxygen content, so
the promising targets for biomass-derived chemicals are oxygen
containing ones.[1] Recently, bio-alcohols have drawn
considerable interests as oxygenate blending for diesel fuels.[2]
The most widely used alternative to gasoline is bioethanol,
derived from the fermentation of biomass.[3] However, bioethanol
is by no means the ideal gasoline replacement. It not only
presented a lower energy density than gasoline, but also exhibited
the corrosive and storage problems due to its hygroscopic
nature.[3] By contrast, C4-C6 alcohols, as opposed to the low
chain alcohols, have a promising future as fuel additives in
combustion engines and as petrochemical building blocks.[1a-c, 3-4]
Table S1 listed the comparison of fuel characteristics between
x
/Zr-
x
2
2
4
[7]
4
and TfOH for converting biomass-derived furfural-acetone into 1-
octanol with 62% carbon yield.[8] Although some pioneering work
has been reported in the area, the R&D is still in its infancy. Higher
reaction temperature or longer reaction time was required to
ensure the high conversion and selectivity. Carbohydrates are
instable at elevated temperature (>473 K) and favor to coke
formation due to polymerization of intermediates. Additionally, the
reported works are mostly focused on expensive noble metals
and silica- and alumina-based supports which are easy to be
hydrolyzed to break Si-O-Si or Al-O-Al bonds in hydrothermal
environment. Therefore, further studies are needed to develop
inexpensive, highly active, selective, hydrothermal-stable
catalytic systems.
Our group attempted to directly convert biomass derivates into
renewable biofuels in a trickle-bed reactor over carbon supported
bimetallic catalysts as the cheap and hydrothermal stable
properties of active carbon. As mentioned in our previous work,
ruthenium molybdenum bimetallic catalysts exhibited high yield
[
[
a]
b]
Dr. Y. Weng, Prof. Y. Zhang, Prof. P. Duan, L. Wang, H. Yin
Henan Key Laboratory of Coal Green Conversion, Henan
Polytechnic University, Jiaozuo, Henan, 454003, PR China
E-mail: wyj162@mail.ustc.edu.cn
Dr. Y. Weng, Prof. T. Wang, Prof. C. Wang, Prof. Q. Liu, S. Liu,
Prof. L. Ma
[9]
Guangdong Provincial Key Laboratory of New and Renewable
Energy Research and Development, Guangzhou Institute of Energy
Conversion, Chinese Academy of Sciences
Nengyuan Road No.2, Guangzhou,510640 (China)
E-mail: mall@ms.giec.ac.cn
and selectivity towards to C5/C6 alkanes in sorbitol conversion
due to its higher cleavage rate of C-O vs. C-C bond.[9a, 9b, 10] Herein,
carbon supported phosphated ruthenium-molybdenum (RuMoP)
catalysts were employed in a continuous trickle-bed reactor for
biomass sorbitol into renewable alkanes and higher alcohols.
Sorbitol was chosen for its higher thermal stability and natural
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
hydrophility due to its high oxygean content than carbohydrates
and readily avalable from degradation/hydrogenation of glucose
or cellulose. The result showed that AC supported RuMoP
presented complete sorbitol conversion and high yields of
gasoline- and diesel-range alkanes and alcohols. Subsequently,
RuMoP/CNT catalyst was prepared and studied in detail for
comparison. RuMoP/CNT displayed a less C-C cracking activity
with much lower content of C1-C4 products and higher content of
C6 species both in gas- and oil-phases. The surface morphology,
surface area, compositions and acidity of the catalysts were
yield of gas-phase products significant decreased but the yield of
aqueous-phase and oil-phase products increased gradually. The
maximum yield of oil-phase products was 29.8 % at the LHSV of
3.0 h-1 with 53.2 % selectivity to higher alcohols. It is noteworthy
that the comparison here was the volumetric space velocity.
Actually, much less RuMoP/CNT catalyst was used versus
RuMoP/C catalyst at the same LHSV due to their varied density.
Thus, Table 1 indicated that RuMoP/CNT displayed less cracking
performance than that of RuMoP/C at all LHSVs.
Table 1. Products distribution of HDO reactions.
3
studied by XRD, XPS, BET, TEM, NH -TPD and Py-IR.
Products
LHSV/h-1
Con./%
RuMoP/C
1.5
RuMoP/CNT
1.5
0.75
99>
/
3.0
0.75
99>
3.0
Results and Discussion
93.1
6.3
90.0
21.5
47.2
21.3
100
99>
91.0
45.6
15.2
29.8
99.6
Hydrodeoxygenation performance
Y
Y
Y
aqueous/%
11.7
70.8
11.1
93.6
30.1
We firstly studied the effects of different additives to the Ru/C
catalyst on the conversion of sorbitol in a trickle-bed reactor. As
shown in Table S2, the monometallic Ru/C catalyst presented a
gas/%
99.1
/
75.7
8.1
42.5
high C-C cracking activity with high selectivity to short alkanes (C
1
oil/%
26.0
69.0%) in gas-phase. On the contrary, bimetallic RuMo/C catalyst
CB/% [a]
99.1
97.0
98.6
showed low selectivity to short alkanes (C
1
11.8.0%) but higher
35.9%) and
32.0%). This could be attributed to the bimetallic Ru-
structure on catalyst which suppressed the C-C cracking
selectivity to long alkanes such as pentane (C
hexane (C
MoO
5
aqueous selectivity/%
6
Iso. [b]
/
/
/
/
9.5
35.3
11.2
16.2
37.3
51.1
24.6
11.5
12.8
68.4
15.6
8.3
77.8
10.2
6.4
x
and enhanced the hydrodeoxygenation performance. Additionally,
trace amount of oil products (liquefied alkanes, mono-alcohol,
etc.) were also obtained. For the RuMoP/C catalyst, in addition to
gaseous products rich in C5 and C6 alkanes (C5 37.9.0%, C6
Methanol
Ethanol
Others
25.1
34.1
31.3
7.7
3.6
36.5.0%), it is also obtained a small amount of oil products with
high content of C4-C6 alcohols. This indicated that the RuMoP/C
catalyst exhibited lower C-C cracking property than others.
To optimize the yield and selectivity of higher alcohols, the effect
of sorbitol liquid space velocity (LHSV) and carbon supports were
determined and presented in Table 1. For RuMoP/C catalyst, a
complete sorbitol conversion with 99.1% gas-phase products was
Gas selectivity/%
C
C
C
C
C
C
1
2
3
4
5
6
9.4
3.0
9.8
3.6
8.3
5.1
1.1
1.2
4.1
2.5
5.0
4.4
4.0
3.6
5.3
3.7
3.9
5.8
-
1
obtained at lower LHSV of 0.75 h . The products in gas-phase
were mainly C4-C6 alkanes (83.6%). While, the products
distribution especially in aqueous- and oil-phase slightly changed
after the increase of sorbitol LHSV. The major aqueous products
were residue sorbitol, isosorbide and polar mono-oxygenated
intermediates such as methanol, ethanol and propanol.
Additionally, major species of the oil-phase products were mono-
alcohols (1-butanol, 2-pentanol, 3-pentanol, 1-pentanol, 3-
hexanol, 2-hexanol, 2-methyl-pentanol, 1-hexanol, etc.), mono-
ketones (2-hexanone, 3-hexanone, hexanal, etc.) and long chain
alkanes (>C5). These products were spontaneously separated
into the gaseous-, aqueous- and oil-phases according to its
boiling point and polarity. The maximum oil yield was 21.3% with
8.7
8.8
12.7
36.0
32.6
5.6
6.0
8.7
32.6
42.3
37.9
36.5
13.7
74.7
29.4
54.1
33.9
42.2
Oil selectivity/%
Hexane
/
/
/
/
/
/
/
/
/
1.0
1.1
1.2
2.1
15.4
6.2
3.6
2.9
4.6
1.2
1.3
29.8
/
19.9
0.1
0.1
0.2
1.9
1.4
3.0
9.1
13.0
12.3
2.0
1
2
3
1
-butanol
-pentanol
-pentanol
-pentanol
2.1
/
1.2
1.9
/
1.1
-
1
7
6.1% selectivity to higher alcohols at the LHSV of 3.0 h .
10.1
4.3
1.1
1.2
1.4
7.1
9.8
2.0
The RuMoP/CNT catalyst were also investigated in detail at
-
1
different sorbitol LHSV. For the reaction at LHSV of 0.75 h ,
0.8 % gas-phase products and 11.1% oil-phase products were
obtained. Interestingly, the selectivity of C6 species in gas-phase
C6 alkane, 74.7%) and oil-phase (C6 alkane and alcohols,
7.8%) was much higher than former catalysts, indicating a lower
3-hexanone
4.6
7
2
-hexanone
-hexanol
5.7
5.4
(
3
15.7
18.8
14.7
22.2
8
-
1
2-hexanol
C-C cracking property. For the higher LHSV of 1.5 and 3.0 h , the
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
around 40 MJ/Kg, which was more than three times of sorbitol
1
-hexanol
/
/
/
/
29.1
6.4
26.2
2.7
3.1
6.9
25.8
3.1
28.8
1.5
22.8
0.2
(
16.7 MJ/Kg). Especially, the maximum HHV of oil samples pf
Undecane
Dodecane
Others
-1
RuMoP/CNT catalyst at the LHSV of 0.75 h reached up to 44.8
MJ/Kg, close to the HHV value of gasoline and diesel oil.[2b, 2c]
9.9
8.3
3.3
1.5
Table 2. Elemental analysis and HHV of oil samples.
16.5
12.4
23.7
32.8
Elemental analyses
HHV/MJ. Kg-1
[
4
a] carbon balance. [b] Isosorbide. Reaction conditions: temperature 573 K,
Sample
-1
MPa H
2
pressure, GHSV = 2250 h , 20 wt. % sorbitol solutions, 4 mL
C/%
H/%
O/%
H/C
V
Calculated
v
Measured
catalyst (RuMoP/C≈2.0 g, RuMoP/CNT≈1.0 g), “/” trace amount.
RuMoP/C[a]
71.5
69.3
82.1
80.1
76.1
14.5
14.1
14.0
12.9
14.1
14.0
16.6
3.9
2.4
2.5
1.9
2.0
2.2
40.6
39.1
44.8
42.4
42.2
37.0
35.9
42.4
41.9
42.4
GC-MS was applied to gain insight into the molecular
compositions of the oil samples (Figure S1). For the oil sample of
RuMoP/C catalyst, major components were C4-C6 mono-
alcohols and ketones. While, for the sample of RuMoP/CNT, it
presented more C6 species indicating the low C-C cracking
property. To figure out the oil sample in detail, comprehensive
RuMoP/C[b]
RuMoP/CNT[c]
RuMoP/CNT[d]
RuMoP/CNT[e]
7.0
two-dimensional
gas
chromatography/time-of-flight
mass
spectrometry (GC×GC-TOFMS) equipped with Rxi-5ms column
and Rxi-17 column was applied to analysis of the oil samples.[11]
As shown in Figure 1, different spots were the representative of
different compounds. Compared with alkanes and olefins,
oxygenates were located at the upper part of the 2D plane due to
their higher polarity. Generally, the shape and the color depth of
a spot in chromatogram represents the relative content of
compounds in the tested sample. Isomers were grouped into the
same band of the chromatogram and can be recognized owing to
the roof-tile effect. Thus, the result indicated that the major group
here was the oxygenate compounds group, which reached up to
the selectivity of 92% according to the area normalization method.
Importantly, the higher alcohols accounted for about 80% in the
group of oxygenate compounds.
9.8
-
1
-1
[
0
a] sorbitol LHSV of 1.5 h . [b] sorbitol LHSV of 3.0 h . [c] sorbitol LHSV of
.75 h . [d] sorbitol LHSV of 1.5 h . [e] sorbitol LHSV of 3.0 h .
-1
-1
-1
Catalyst stability would be challenging for the aqueous HDO of
biomass feedstock at present. Here, the long-term performance
over RuMoP/CNT catalyst was carried out to test its stability in the
sorbitol conversion. As shown in Figure S2, no obvious decline in
the catalytic activity and oil yield was observed after operating for
7 days (168 h) time on stream, indicating the stability of the
catalyst under hydrothermal environment. Meanwhile, the spent
catalyst characterization of XRD and TEM also suggested that no
evident change in the structure of catalyst during the reaction.
Certainly, the result here was not the best in the recently work
(
maximum 95% alkanes[7b] and 52% alcohols[7a]), but we
demonstrated the conversion in continuous reactor with high
catalytic activity and hydrothermal stablility [
1b, 1c]
.
Catalyst charcterizations
To
understand
the
catalytic
mechanism,
detailed
characterizations were performed over relevant catalysts
(
RuMoP/C and RuMoP/CNT). Figure S3 presented the N
2
adsorption-desorption isotherms and pore size distribution.[
10]
RuMoP/C exhibited a heterogeneous porous structure with a
round knee at low relative pressures (P/P < 0.05) suggesting a
0
wide microporous structure and a hysteresis loops at medium-
high relative pressures (0.4 < P/P0 < 0.8) indicating a mesoscale
porosity.[13] However, RuMoP/CNT exhibited hysteresis with
much narrower pore size distribution of mesoscale porosity.
Disappearance of the micropore structure could be explained by
the intact carbon-nanotube without inner micropore from end
damage during phosphating process and aqueous phase
reaction.[ Moreover, Table 3 listed the data of physicochemical
properties. AC samples showed higher surface areas (ABET) than
CNT samples due to the heterogeneous structure. While, CNT
samples showed lower surface areas, but a larger pore volume,
which can be explained by the presence of abundant mesopore
structure from the intertwining of nanotubes.
Figure 1. GC-MS×MS result of RuMoP/C catalyst at LHSV of 3.0 h-1.
Elemental analysis of oil samples was listed in Table 2. Obviously,
oil samples of the RuMoP/C catalyst had higher O content (14.0 %
and 16.6 %) and H/C molecular ratio (2.4 and 2.5) than that of the
RuMoP/CNT catalysts. It could be explained by the fact that the
RuMoP/C catalyst produced oil products with more C4 and C5
alcohols, whose O contents were higher than that of C6 alcohols
and hexane.[12] For the oil samples of RuMoP/CNT catalysts, the
O content and H/C molecular ratio were decreased to 3.9 % and
14]
-
1
-1
1
.9 at LHSV of 0.75 h , 7.0 % and 1.9 at LHSV of 1.5 h , and
-
1
9.8 % and 2.2 at LHSV of 3.0 h , respectively. It indicated that the
elemental compositions of oil samples were influenced by the
liquid space velocity (LHSV) of feedstock solution. Moreover, the
higher heat values (HHV) of all samples calculated by elemental
compositions and measured with bomb calorimeter were at
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
As Ru3d spectrum overlaps with the C1s spectrum, Ru3p
spectrum (Figure 3d) was selected for further analysis. The small
Table 3. Structural characteristic of the catalyst samples.
A
m /g)
BET
2
V
p
A
t
V
mic
3
peak at around 462.3 eV observed on catalysts could be assigned
to the metallic Ru (Ru0, Ru (3p3/2)) according to relevant
reports.[18]
Sample
[a]
[b]
[c]
[c]
3
2
(
(cm /g)
(m /g)
(cm /g)
RuMoP/C
1110
574
211
125
0.68
253
0.36
a
b
RuMoP/CNT spent
RuMoP/CNT spent
RuMoP/CNT
RuMoP/C spent
RuMoP/CNT
RuMoP/CNT spent
0.37
1.38
1.00
208
211
125
0.16
RuMoP/CNT
/
/
5
40 539 538 537 536 535 534 533 532 531 530 529 528
Binding energy (eV)
139
138
137
136
135
134
133
132
131
Binding energy (eV)
c
d
RuMoP/C spent
RuMoP/C spent
[
a] ABET, calculated using the Brunauer–Emmett–Teller equation with proper
correlation coefficient r (r> 0.9999) and C constant (C>0). [b]. Vp, estimated as
the liquid N at a relative pressure of 0.99 with fitting errors (±0.5%). [c] At, the
RuMoP/CNT spent
RuMoP/CNT spent
2
external surface area (i.e., mesoporous plus macroporous) determined by the t-
plot method. Vmic, determined by the t-plot method with fitting errors (±1%). /,
trace amount.
2
38 237 236 235 234 233 232 231 230 229 228 227 226
Binding energy (eV)
490
485
480
475
470
465
460
Binding energy (eV)
Figure 3. XPS spectra of O1s (a), P2p (b), Mo 3d (c) and Ru3p (d).
XRD patterns were shown in Figure 2. The weak diffraction peaks
at 44.1°and 26.0°on RuMoP/C catalysts could be assigned to
SEM and TEM experiments were also conducted. As shown in
Figure S4, the RuMoP/C catalyst presented a heterogenous
structure with high BET surface area, leading to a good dispersion
of metal particles with an average size about 2.0 nm. Moreover,
the Ru (101) and MoO
2
(011) planes. Besides diffraction peaks of
C, Ru and MoO , peaks at 34.5°, 38.0°, 39.6°, 52.3°and
2
6
1.8°on RuMoP/CNT catalysts were assigned to the (100), (002), metal Ru and RuO were found on HRTEM images according to
2
(
(
101), (102) and (110) planes of molybdenum carbide (Mo
2
C)
the analysis of various lattice fringes and characteristic acute
angles.[19] As to the case of the RuMoP/CNT spent sample, TEM
Reference pattern: 03-065-8766).[15] Generally, transition metal
carbides were formed in situ via carbothermal reduction at high
temperature (> 873 K). [16] Interestingly, a facile method was
developed by in situ carburization of Ru-doped molybdenum
oxide on CNT via temperature-programmed reaction (TPR) with
and HRTEM images in Figure 4 also showed fine nanoparticles
and good dispersion. It is noteworthy that the images in Figure 4a,
4b and 4d presented an intact carbon-nanotube structure without
any damage, consistent with the analysis of N adsorption-
2
final temperature of 673 K in this work.
desorption. HRTEM images with insets of fast Fourier transform
(FFT) patterns were presented to identify to nanostructure of
nanoparticles. As shown in Figure 4c, lattice fringes with
interplanar spacing of 0.205 and 0.232 nm corresponded to (101)
C
b
C
Ru
Mo C
a
SiO
2
Ru
2
MoO
2
MoO
2
RuMoP/C spent
RuMoP/C
RuMoP/CNT spent
RuMoP/CNT
CNT
0
[20]
and (100) facets of metal Ru particles, respectively.
Another
nanoparticle in vicinity was depicted with clearly lattice fringes of
0.228 nm corresponding to the (101) facets of hexagonal close
AC
packed (HCP) Mo
with low crystallization were found among the Ru and Mo
nanoparticles, which could be attributed to MoO species or
phosphated species. Unfortunately, no obvious MoO
nanoparticles were found on the support, probably due to the fine
particles of amorphous MoO
species.[9a, 10, 13] Additionally, the
2
C.[15a, 15c] Moreover, some amorphous species
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
θ( °)
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
2
2θ( °)
2
C
Figure 2. XRD spectra of RuMoP/C and RuMoP/CNT catalysts (a and b).
XPS was applied to identify the possible chemical states on
catalysts. The O1s spectra in Figure 3a indicated the presence of
several different species: (1) 529.2 ~530.6 eV (metal oxides, Mo-
x
x
x
O), (2) 531.0~531.9 eV (C=O, O-C=O*, P-O), (3) 532.0~534.5 eV
O).[9b,
10]
corresponding EDX elemental mappings of catalysts revealed
that the P, Ru, and Mo elements were distributed homogeneously
and overlapped on the outer surface of support.
(
C-OH, C-O-C, *O-C=O), (4) 534.5~536.1 eV (H
2
Moreover, a broad band at around 135.5 eV was observed in P2p
spectra (Figure 3b), which could be corresponded to the high
3 4 4 2 5 4 2 7
valance of P species (e.g. H PO , H P O , H P O
case of Mo3d, an obvious difference was observed between
RuMoP/C and RuMoP/CNT in Figure 3c. A new binding energy at
around 228.2 eV was emerged on the RuMoP/CNT sample, which
). [10, 17] In the
Figure 5 presented NH
at around 450 K could be attributed to the desorption of weakly
adsorbed NH and peaks at around 620 K could be ascribed to
the desorption of moderately adsorbed NH
.[17b, 21] RuMoP/C
exhibited a big NH desorption peak of weak acid and a small
broad NH desorption peak of moderate acid, but RuMoP/CNT
only presented a large NH desorption peak at the temperature
range of 400-600 K. The quantified total acid sites of RuMoP/CNT
0.17 mmol/g) was slight lower than that of RuMoP/C (0.22
mmol/g).The possible reason for the decrease of acidity could be
3
-TPD results of relevant catalysts. Peaks
3
3
C.32, 34, 35 Curve fitting of
3
could be assigned to the formation of Mo
2
3
the Mo3d region (3d3/2+3d5/2) by least squares method using the
Gaussian-Lorentzian peak shape and Shirley line as background
were carried out. Table S3 presented semi-quantitative results of
relative contents of Mo species by the XPS spectrum. 16.1%
3
(
2
Mo C species were detected on the RuMoP/CNT spent catalyst.
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
attributed to the decrease of MoO
x
species which were
suppress the cracking property of metal Ru. Additionally, the
adsorbed phosphate groups would be Bronsted acid sites and
would work in concert with metal sites in HDO reaction according
to literature.[17a, 25] Among the catalysts, RuMoP/CNT exhibited
excellent HDO performance with high selectivity to C6 products in
gas-phase (C6 alkane, 74.7%) and oil-phase (C6 alkane and
alcohols, 87.8%). Detailed characterization studies suggested
that RuMoP/CNT presented nice dispersion of nanoparticles on
outer surface of carbon-nanotubes, which significantly increased
the contact surface between the reactants and active sites, and
greatly minimize the diffusion limitation. Moreover, the emergence
responsible for Lewis acid sites according to the literature.[
22]
Moreover, pyridine adsorption (Py-IR) results of relevant catalysts
(
Figure S5 and S6) were supported to verify the hypothesis. The
RuMoP/C catalyst presented much more acid sites, especially
[23,
Brønsted acid sites, than that of Ru/C and RuMo/C catalysts.
2
4]
of small amount Mo
HRTEM analyses, also had
performance. According to relevant literatures, Mo
2
C on CNT catalysts, confirmed by XRD and
great effect on the HDO
C possessed
a
2
similar bulk electronic properties and catalytic activity to noble
metal catalysts such as Pt.[15c, 26] Mo-based transition-metal
catalysts (Mo C,Mo S, etc.) attracted much attention as an
2 2
inexpensive material with high catalytic activity in HDO
reaction.[27] Since the in-situ analysis for aqueous HDO catalytic
system would be challenging at present, the detailed synergy
catalytic mechanism will be further studied in future. Above all,
these results were keys to develop an effective catalyst for
renewable alkanes and higher alcohols from biomass-derived
feedstocks.
Conclusions
Figure 4. Typical TEM and HRTEM images (a, b, c, d), and corresponding EDX
elemental mapping of P, Ru and Mo atoms (e, f, g) of RuMoP/CNT spent sample.
In summary, phosphated ruthenium molybdenum catalysts were
developed to produce liquid fuels with high content of long-chain
alkanes (>C5) and higher alcohols (>C4) from biomass sorbitol
via aqueous HDO in a continuous trickle-bed reactor. The product
distribution was notably influenced by the sorbitol LHSV. The high
sorbitol LHSV was favorable for the higher mono-alcohols with
partial oxygen removal. 21.3% selectivity to liquid fuels with C4-
RuMoP/C
-
1
C6 higher alcohols content of 76.1 % was obtained at 3.0 h , 573
K. Moreover, RuMoP/CNT presented a low C-C bonding cracking
property in sorbitol conversion with high selectivity to C6 products
in gas-phase (C6 alkane, 74.7%) and oil-phase (C6 alkane and
alcohols, 87.8%). These results are important to understand the
catalytic mechanism for the HDO of biomass-derived polyol into
higher bio-alcohols and provide a guidance for the design of new
catalytic formulations.
RuMoP/CNT
3
73 423
473
523
573
623
673 723
Temperature (K)
3
-TPD analyses of two catalysts.
Figure 5. NH
Proposed catalytic mechanism
As reported, the HDO reaction is a catalytic process whereby the
total or partial removal of oxygen from biomass derivatives. [21]
Obviously, the results of sorbitol test indicated that the mono-
alcohol products were key intermediates in the HDO reaction of
sorbitol into alkanes and could be changed by adjusting the space
velocity of the sorbitol solution. Generally, the increase of reactant
space velocity could decrease the contact time, which would lead
to oxygenated intermediates such as mono-alcohol.[9b,
Therefore, the low C-C cracking property of metal catalyst would
enhance the yield and selectivity to higher alcohols. Herein, the
RuMoP catalyst presented a lower C-C cracking property with
higher C5-C6 products in gas- and oil-phase. The subsequent
characterizations over relevant catalysts (RuMoP/C and
Experimental Section
Sorbitol was supplied by Yuan-Ju biotechnology Reagent Co. Ltd.
(Shanghai, China). Ruthenium (Ⅲ) chloride hydrate (RuCl3•3H2O) was
purchased from Aladdin Co. Ltd. (Shanghai, China). Ammonium
heptamolybdate ((NH4)6Mo7O24) and phosphoric acid (H3PO4, 85 wt.%)
were from Tianjin Fuyu Fine Chemical Co. Ltd. (Tianjin, China). Moreover,
active carbon (AC, 70-120 mesh, Huanneng carbon Co., Ltd., Heibei,
China) and multi-walled carbon-nanotube (CNT, inner diameter 2-5 nm,
length 10-30 um, Chengdu Organic Chemicals Co., Ltd., Chinese
Academy of Sciences, Sichuan, China) were selected as support materials.
24]
RuMoP/CNT) suggested that Ru and partial reduced MoO
x
were
highly dispersed on the support with intimate contact, which would
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
-
1
Phosphated ruthenium molybdenum carbon catalysts (RuMoP/C and
RuMoP/CNT) were prepared by sequential impregnation with 1.5 wt.% Ru
loading amount and molar ratio of Ru/Mo=1 (Table S4, ICP analytic result).
Took the RuMoP/C catalyst as an example. Firstly, active carbon was pre-
treated by using phosphoric acid at 383 K overnight, then cooled to room-
temperature, filtered and dried at 393 K overnight before impregnation.
Secondly, the phosphated active carbon was incipient impregnated into
RuCl3 solution followed by ultrasonic treatment for 2 h. Thirdly, the mixture
was then transferred to an oven and dried at 393 K for 12 h, and the dry
sample was next heat treated in a muffle furnace under a nitrogen
atmosphere at 673 K for 4 h. Lastly, the second impregnation of Mo was
carried out in the similar process. Additionally, all catalysts were reduced
NH3/He flow (50 ml min ) for 1 h. After that, the sample was swept with
helium gas flow at 383 K until the baseline stabilized. The desorption of
NH3 was carried out in helium gas flow (30 cm3 min ) by increasing the
-1
-
1
temperature from 100 to 723 K at the heating rate of 10 K min . Infrared
spectra of pyridine adsorption (Py-IR) were also used for acidity study with
a Thermo Nicolet Nexus FT-IR spectrometer (Massachusetts, USA)
equipped with a liquid nitrogen cooled MCT detector.
Acknowledgements
3
in a H2 flow (150 cm /min) at 673 K for 4 h prior to the reaction.
This work was supported by grants from the Scientific and
Technological Research Projects of Henan Province
(182102210319), Henan Polytechnic University (B2018-42),
Outstanding Youth Foundation for Scientific and Technological
Innovation in Henan Province (184100510013), National Natural
Science Foundation of China (51536009), National key research
plan (2018YFB0604500) and Chinese Academy of Sciences “one
hundred talented plans”. Finally, thanks to the supports from
Lungang Chen, Haiyong Wang and Kai Wang in the experiment
and characterization sections.
HDO reaction was carried out in a tubular stainless-steel trickle-bed
reactor (inner diameter of 10 mm; length of 350 mm) at 4 MPa H2 pressure.
Before the reaction, 4 ml catalyst was loaded in the constant temperature
zone of the reactor with quartz sand and quartz wool as the filler materials
on the top and bottom of the catalyst bed, respectively. During reaction,
the sorbitol aqueous solution (20 wt.%) was pumped into the trickle-bed
reactor at different flow rates by a high-pressure liquid pump (HPLP). H2
flow (150 cm3 min ) was purged into the reactor at the same time. The
reactor set-up kept H2 pressure at 4 MPa with the use of pre- and rear-
pressure controller and operated in the co-current-flow of liquid and
hydrogen from top to bottom (Figure S7).
-1
Keywords: Biomass catalytic conversion • Higher alcohol •
Hydrodeoxygenation • Phosphated ruthenium-molybdenum
catalyst • Renewable alkane.
The reaction products were distributed in gaseous phase and liquid phase.
Gaseous sample was analyzed on-line by gas chromatography (Shimadzu
GC-2010, Rt@-Alumina Bond capillary column, 50 m×0.53 mm, ± 5%)
equipped with a thermal conductivity detector (TCD) and a flame ionization
detector (FID) via external standard method. Additionally, liquid samples
[1]
aD. Sun, S. Sato, W. Ueda, A. Primo, H. Garcia, A. Corma, Green
Chem. 2016, 18, 2579-2597; bJ. Lee, I. Ro, H. J. Kim, Y. T. Kim, E.
E. Kwon, G. W. Huber, Process Saf. Environ. Prot. 2017; cN. M.
Eagan, J. P. Chada, A. M. Wittrig, J. S. Buchanan, J. A. Dumesic,
G. W. Huber, Joule 2017, 1, 1-22; dK. Tomishige, Y. Nakagawa, M.
Tamura, Green Chem. 2017, 19, 2876-2924.
aR. Sundar, G. Saravanan, Thermal Science 2011, 15, 1215-1222;
bM. V. De Poures, A. P. Sathiyagnanam, D. Rana, B. R. Kumar, S.
Saravanan, Appl. Therm. Eng. 2017, 113, 1505-1513; cN. Yilmaz,
A. Atmanli, Fuel 2017, 210, 75-82.
(
oil phase and aqueous phase), were condensed and collected in an ice-
water condensing tank and then weighted at an analytical balance (±
.5 %), further detected by a Waters Alliance e2695 HPLC (± 2.27%) with
0
[
2]
UV-Vis (Waters 2489) and Refractometer (Waters 2414) detector by using
a Shodex SH1011 column (8× 300 mm). 0.005 M H2SO4 as mobile phase
-
1
flowed at a rate of 0.5 ml·min . GC-MS (Thermo Trace GC Ultra with a
Polaris-Q ion trap mass spectrometer) equipped with a HP-INOWx
capillary column was applied to analyze the species in aqueous samples
and a HP-5MS capillary column was applied to analyze the species in the
oil phase samples. For the analyses of oil samples, the elements including
C, H and O, were measured by Vario EL Ⅲ elemental analyzer (± 0.2 %).
The higher heat value (HHV) was calculated according to the formula
developed by Channiwala and Parikh[28]. Meanwhile, the HHV of oil
samples were also measured by bomb calorimeter WZR-1T-
Cll&IKAC2000 (± 0.2 %).
[3]
K. J. Pellow, R. L. Wingad, D. F. Wass, Catal. Sci. Technol. 2017,
7, 5128-5134.
[
4]
aC. Tang, L. Wei, X. Man, J. Zhang, Z. Huang, C. K. Law, Combust.
Flame 2013, 160, 520-529; bK. Liu, J. Pritchard, L. Lu, R. van Putten,
M. Tiny Verhoeven, M. Schmitkamp, X. Huang, L. Lefort, C. J. Kiely,
E. J. M. Hensen, E. A. Pidko, Chem. Commun. 2017, 53, 9761-
9
764; cS. Fu, Z. Shao, Y. Wang, Q. Liu, J. Am. Chem. Soc. 2017.
[
5]
P. H. Galebach, D. J. McClelland, N. M. Eagan, A. M. Wittrig, J. S.
Buchanan, J. A. Dumesic, G. W. Huber, ACS Sustainable Chem.
Eng. 2018, 6, 4330-4344.
E. L. Kunkes, D. A. Simonetti, R. M. West, J. C. Serrano-Ruiz, C. A.
Gartner, J. A. Dumesic, Science 2008, 322, 417-421.
[6]
[
7]
aS. Liu, Y. Okuyama, M. Tamura, Y. Nakagawa, A. Imai, K.
Tomishige, Green Chem. 2016, 18, 165-175; bS. Liu, Y. Okuyama,
M. Tamura, Y. Nakagawa, A. Imai, K. Tomishige, ChemSusChem
N2 adsorption–desorption experiments were measured using
a
QUADRASORB SI analyzer equipped with Quadra Win software system
at liquid nitrogen temperature (77 K). The crystalline structure of the
catalysts was characterized by X-ray diffraction (XRD) (X Pert Pro MPD
with Cu Kα (λ = 0.154 nm) radiation, Philip) operated at 40 kV and 100 mA.
X-ray photoelectron spectroscopy (XPS) analyses of the samples were
obtained using a Leybold LH 10 spectrometer equipped with a single-
channel detector, employing Al Ka radiation (1253.6 eV, anode operated
at 10 KV X 23 mA). The C1s peak position was set at 284.5 eV and used
as reference to locate the other peaks. The microstructure and metal
distribution were characterized by scanning electron microscopy (SEM)
and transmission electron microcopy (TEM). The total acidity and acid
strength distribution were determined by temperature-programmed
desorption of ammonia (NH3-TPD) with a Micromeritics chemisorbs 2750
pulse chemisorption system to estimate the acidity (100 mg catalyst). The
2
015, 8, 628-635.
Q. Xia, Y. Xia, J. Xi, X. Liu, Y. Wang, Green Chem. 2015, 17, 4411-
417.
[
[
8]
9]
4
aY. Yang, Q. Liu, D. Li, J. Tan, Q. Zhang, C. Wang, L. Ma, RSC Adv.
2017, 7, 16311-16318; bF. Sun, L. Chen, Y. Weng, T. Wang, S. Qiu,
Q. Li, C. Wang, Q. Zhang, L. Ma, Catal. Commun. 2017, 99, 30-33;
cY. Weng, S. Qiu, C. Wang, L. Chen, Z. Yuan, M. Ding, Q. Zhang,
L. Ma, T. Wang, Fuel 2016, 170, 77-83.
Y. Weng, T. Wang, S. Qiu, C. Wang, L. Ma, Q. Zhang, L. Chen, Y.
Li, F. Sun, Q. Zhang, ChemCatChem 2017, 9, 774-781.
K. L. Ho, D. J. Lee, A. Su, J. S. Chang, Int. J. Hydrog. Energy 2012,
[
10]
[11]
12]
37, 15569-15574.
[
aY. Ni, W. Peng, A. Sun, W. Mo, J. Hu, T. Li, G. Li, J. Ind. Eng.
Chem. 2010, 16, 503-505; bX. Li, L. Feng, Z. Liu, B. Zhong, D. B.
Dadyburjor, E. L. Kugler, Ind. Eng. Chem. Res. 1998, 37, 3853-3863.
Y. Weng, T. Wang, P. Duan, C. Wang, F. Wang, Q. Liu, L. Chen, H.
Wang, Z. Liang, L. Ma, Energy Technol. 2017.
[13]
3
-1
sample was firstly pre-treated by helium flow (30 cm min ) at 573 K for 1
h, and then cooled to 323 K and ammonia-saturated in a stream of 10 %
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
[
14]
aS. Hanelt, G. Orts-Gil, J. F. Friedrich, A. Meyer-Plath, Carbon 2011,
Luque, C.-J. Zhong, H. Ye, Z. Yang, B. H. Chen, J. Mater. Chem. A
2017, 5, 7869-7875.
49, 2978-2988; bS. Aftab, H. Nawaz, M. Siddiq, J. Chem. Soc. Pak.
2
013, 35, 308-313.
[21]
[22]
Y. T. Kim, J. A. Dumesic, G. W. Huber, J. Catal. 2013, 304, 72-85.
J. Cui, J. Tan, Y. Zhu, F. Cheng, ChemSusChem 2018, 11, 1316-
1320.
aS. Koso, H. Watanabe, K. Okumura, Y. Nakagawa, K. Tomishige,
Appl. Catal., B 2012, 111-112, 27-37; bJ. Guan, G. Peng, Q. Cao,
X. Mu, J. Phys. Chem. C 2014, 118, 25555-25566.
aY. Roman-Leshkov, C. J. Barrett, Z. Y. Liu, J. A. Dumesic, Nature
2007, 447, 982-U985; bR. M. West, E. L. Kunkes, D. A. Simonetti,
J. A. Dumesic, Catal. Today 2009, 147, 115-125; cA. V. Kirilin, A. V.
Tokarev, E. V. Murzina, L. M. Kustov, J. P. Mikkola, D. Y. Murzin,
ChemSusChem 2010, 3, 708-718.
[
15]
aM. Grilc, G. Veryasov, B. Likozar, A. Jesih, J. Levec, Appl. Catal.,
B 2015, 163, 467-477; bP. Liang, H. Gao, Z. Yao, R. Jia, Y. Shi, Y.
Sun, Q. Fan, H. Wang, Catal. Sci. Technol. 2017, 7, 3312-3324; cJ.
Li, L. Liu, Y. Liu, M. Li, Y. Zhu, H. Liu, Y. Kou, J. Zhang, Y. Han, D.
Ma, Energy Environ. Sci. 2014, 7, 393-398.
[23]
[24]
[
16]
aH. Xiong, L. L. Jewell, N. J. Coville, ACS Catal. 2015, 5, 2640-
2658; bE. Santillan-Jimenez, M. Perdu, R. Pace, T. Morgan, M.
Crocker, Catalysts 2015, 5, 424-441; cY. Qin, P. Chen, J. Duan, J.
Han, H. Lou, X. Zheng, H. Hong, RSC Adv. 2013, 3, 17485.
aO. A. Rusu, W. F. Hoelderich, H. Wyart, M. Ibert, Appl. Catal., B
[
[
17]
18]
2
015, 176-177, 139-149; bL. Atanda, S. Mukundan, A. Shrotri, Q.
[25]
aJ. Wang, H. Liu, S. Yang, J. Zhang, C. Zhang, H. Wu, Appl. Surf.
Sci. 2014, 316, 443-450; bN. Li, G. A. Tompsett, G. W. Huber,
ChemSusChem 2010, 3, 1154-1157.
Ma, J. Beltramini, ChemCatChem 2015, 7, 781-790.
aC. Sassoye, G. Muller, D. P. Debecker, A. Karelovic, S.
Cassaignon, C. Pizarro, P. Ruiz, C. Sanchez, Green Chem. 2011,
[26]
[27]
A. Mehdad, R. E. Jentoft, F. C. Jentoft, J. Catal. 2017, 347, 89-101.
aM. Grilc, B. Likozar, J. Levec, Appl. Catal., B 2014, 150-151, 275-
287; bG. Veryasov, M. Grilc, B. Likozar, A. Jesih, Catal. Commun.
2014, 46, 183-186; cB. Hočevar, M. Grilc, M. Huš, B. Likozar, Appl.
Catal., B 2017, 218, 147-162; dJ. Han, J. Duan, P. Chen, H. Lou, X.
Zheng, H. Hong, ChemSusChem 2012, 5, 727-733; eY. Qin, L. He,
J. Duan, P. Chen, H. Lou, X. Zheng, H. Hong, ChemCatChem 2014,
6, 2698-2705.
13, 3230; bR. Fang, H. Liu, R. Luque, Y. Li, Green Chem. 2015, 17,
4183-4188.
[
[
19]
20]
aJ. Jae, W. Zheng, A. M. Karim, W. Guo, R. F. Lobo, D. G. Vlachos,
ChemCatChem 2014, 6, 848-856; bY. Li, Q. Zhang, N. Zhang, L.
Zhu, J. Zheng, B. H. Chen, Int. J. Hydrog. Energy 2013, 38, 13360-
13367.
aW. Wang, S. Guo, I. Lee, K. Ahmed, J. Zhong, Z. Favors, F. Zaera,
M. Ozkan, C. S. Ozkan, Sci. Rep. 2014, 4, 4452; bL. Zhu, S. Shan,
V. Petkov, W. Hu, A. Kroner, J. Zheng, C. Yu, N. Zhang, Y. Li, R.
[28]
P. Duan, B. Wang, Y. Xu, Bioresour. Technol. 2015, 186, 58-66.
…
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201801214
FULL PAPER
This article is protected by copyright. All rights reserved.