Unique approach for transforming glucose to C3 platform chemicals using metallic iron and a Pd/C catalyst in water

Article abstract of DOI:10.1246/bcsj.20160114

In the utilization of biomass for fuels and chemicals, the key reactions involved are hydrogenolysis and deoxygenation using hydrogen. Unfortunately, the industrial use of molecular hydrogen is limited because of various drawbacks, such as high cost and considerable hazards associated with high-pressure operation. In this study, a unique chemical manufacturing process was proposed for inducing the hydrogenolysis and deoxygenation of biomass carbohydrates using hydrogen generated in situ from the reaction between metallic iron and water. From the results obtained, hydrogen generated in situ by metallic iron particles (hydrogen-generating agent) combined with a carbon-supported palladium catalyst (hydrogenation catalyst) transforms glucose to C3 platform chemicals, such as propylene glycol, hydroxyacetone, and lactic acid. Moreover, reaction conditions and mechanism were also evaluated. With the use of the proposed system, value-added chemicals were produced from biomass carbohydrates by using renewable sources of energy (such as hydrogen generated from the reaction between iron and water) without the complete dependence on fossil resources.

Full text of DOI:10.1246/bcsj.20160114

Unique Approach for Transforming Glucose to C3 Platform Chemicals
Using Metallic Iron and a Pd/C Catalyst in Water
Yoshiaki Hirano,* Yuka Kasai, Kunimasa Sagata, and Yuichi Kita
Department of Chemical Science and Engineering, Graduate School of Engineering, Kobe University,
1-1 Rokkodaicho, Nada-ku, Kobe, Hyogo 657-8501
E-mail: hirano@phoenix.kobe-u.ac.jp
Received: April 1, 2016; Accepted: June 1, 2016; Web Released: June 8, 2016
Yoshiaki Hirano
Yoshiaki Hirano received his M.E. degree from the Tokyo Institute of Technology in 1995. In 1995, he joined
Nippon Shokubai Co., Ltd. He received his Master of Management of Technology degree (2010) from the
Tokyo University of Science. In 2012, he moved to the endowed chair “Sustainable Chemistry” (Nippon
Shokubai), established in the Department of Chemical Science & Engineering at Kobe University as an
Assistant Professor. His research focuses on practical utilization technologies for renewable biomass
resources to create new industrial fields by sustainable chemistry.
Abstract
Carbohydrates, such as sugars, starch, cellulose, and hemi-
cellulose, have an empirical formula of C (H O) (where m
In the utilization of biomass for fuels and chemicals, the
key reactions involved are hydrogenolysis and deoxygenation
using hydrogen. Unfortunately, the industrial use of molecular
hydrogen is limited because of various drawbacks, such as high
cost and considerable hazards associated with high-pressure
operation. In this study, a unique chemical manufacturing
process was proposed for inducing the hydrogenolysis and
deoxygenation of biomass carbohydrates using hydrogen gen-
erated in situ from the reaction between metallic iron and
water. From the results obtained, hydrogen generated in situ by
metallic iron particles (hydrogen-generating agent) combined
with a carbon-supported palladium catalyst (hydrogenation
catalyst) transforms glucose to C3 platform chemicals, such as
propylene glycol, hydroxyacetone, and lactic acid. Moreover,
reaction conditions and mechanism were also evaluated. With
the use of the proposed system, value-added chemicals were
produced from biomass carbohydrates by using renewable
sources of energy (such as hydrogen generated from the reac-
tion between iron and water) without the complete dependence
on fossil resources.
m
2
n
can be different from n) and high oxygen content. For making
them suitable for use as fuels and chemicals, it is necessary
to decrease the oxidation state of biomass-derived materials
5
,6
via key reactions such as hydrogenolysis or deoxygenation,
which result in the cleavage of C­C and C­O bonds by
hydrogen or the removal of oxygen atoms from a molecule. For
instance, the use of noble- or transition-metal catalysts has been
reported for the conversion of carbohydrates, such as glycerol,
6
glucose, sucrose, sorbitol, and cellulose, to various polyols.
Nevertheless, these methods still exhibit various drawbacks,
such as high cost, considerable hazards associated with high-
pressure operation, low efficiency of hydrogen resulting from
undesired side reactions, significant dependence on fossil
resources, requirement for high-energy-intensive processes, and
geographical constraints associated with hydrogen storage and
transportation. In this regard, a number of methods have been
proposed and intensively developed for the sustainable pro-
duction of hydrogen: electrolysis of water using surplus power
from natural resources (photovoltaics, hydraulic power, and
7
­9
wind power); thermolysis and photolysis of water by solar
7
­9
energy; thermochemical and biological conversion of bio-
7­9
mass (gasification, pyrolysis, and fermentation);
chemical
1
. Introduction
looping of metals or using a metal oxide redox system, which
1
0,11
For several industries, biomass carbohydrates are the most
abundant sustainable resources of both energy and organic
carbon. Since the last decade, because of the depletion of fossil
resources as well as concerns surrounding greenhouse gas
emissions, biomass has been rapidly utilized for the production
of transportation fuels and value-added chemicals.1 In most
cases, biomass has been utilized in three methods: power gener-
ation, thermochemical conversion (gasification or liquefaction),
and biological conversion.
is commonly referred to as the steam iron process.
Zero-
valent metals, such as Ge, Mo, W, Fe, Co, Ni, Cu, Zn, Al, Mn,
as well as their oxides, have been previously investigated, and
promising results have been demonstrated for the production of
1
2­15
hydrogen by thermal water splitting.
­4
A number of studies have been reported on reduction,
hydrogenation, and hydrogenolysis using hydrogen generated
in situ. For example, encouraging results have been observed
in previously reported studies on the use of iron and zinc for the
1026 | Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan

reduction of organic and inorganic compounds such as nitro-
transformed to value-added chemicals (C2­C5 polyols and
alkanes) via hydrogenolysis by hydrogen generated in situ
1
6,17
18,19
20
arenes,
aryl chlorides,
aldehydes, and chlorinated
2
1
hydrocarbons. Recently, Jin and co-workers have reported
from water, forming high-valence iron oxides (FeO ). In the
x
that CO is reduced to formic acid or methanol by hydrogen
second step, the iron oxides can be reduced by a process
powered by renewable sources of energy, such as solar thermal
2
generated in situ by the oxidation of metal powder (such as
Fe, Zn, Al, and Mn) in water.2
2­27
In addition, they have
energy,
from biomass gasification,
biomass-derived tar or biomass itself.
48,49
solar hydrogen,
12,13
reducing gas (CH , CO, H )
4 2
5
0,51
reported a series of experiments on the conversion of biomass-
derived materials to value-added chemicals under similar con-
and carbothermal reduction by
5
2­54
2
8­31
ditions.
Feng and co-workers have reported the reduction
In particular, it is possible to perpetuate the FeO redox cycle
x
of CO to formic acid, acetic acid, and phenol with water in
by using renewable energy. This new proposed process differs
from conventional hydrogenation or reduction using pressur-
ized hydrogen because it does not require an external hydrogen
source. In addition, this method does not require high-purity
and high-pressure hydrogen, and the supply of the hydrogen
generated in situ prevents the consumption of energy for the
production, storage, and transformation of hydrogen, as well
as any hazardous risk involved with its handling. Thus, the
hydrogen-free reaction system can be operated under mild pres-
sure, resulting in improved, safer operation, as well as being
cost-effective. Our proposed system could facilitate the sus-
tainable production of industrial chemical products using re-
newable energy (in situ generation of hydrogen) from biomass
carbohydrates as sustainable carbon resources without the
dependence on fossil resources.
2
the presence of iron nanoparticles under mild hydrothermal
3
2,33
conditions.
Moreover, ever since Dumesic and co-workers
demonstrated that hydrogen can be produced from sugars and
34
alcohols using a platinum-based catalyst in water, aqueous-
3
5­39
phase reforming (APR)
tion (CTH)
and catalytic-transfer hydrogena-
3
7,40­42
using hydrogen donor molecules have been
investigated as efficient technologies for hydrogenation through
the in situ generation of hydrogen for converting biomass-
derived carbohydrates and bio-oil to fuels and chemicals. How-
ever, APR and CTH are hypothesized to exhibit disadvantages
in terms of carbon efficiency, attributed to the consumption of a
raw-material substrate or a hydrogen donor. More recently,
some reduction and hydrogenation or hydrogenolysis reaction
systems with CO and water have been employed as a source of
4
3
44
hydrogen for the reduction of nitro compounds, alkynes,
In this study, we demonstrated that the coexistence of metal-
lic iron particles (hydrogen-generating agent) and a carbon-
supported palladium catalyst (Pd/C, a hydrogenation catalyst)
enables the transformation of glucose (a biomass-based
carbohydrate) to C3 platform chemicals, such as propylene
glycol (PG), hydroxyacetone (HA), and lactic acid (LA), using
4
5
46
aldehydes, and carboxylic acids, and for the conversion of
4
7
cellulose.
In this study, a unique chemical manufacturing process was
proposed, which induces the hydrogenolysis and deoxygena-
tion of biomass carbohydrates using hydrogen generated in situ
via the chemical looping of a metal oxide redox. Typically,
chemical looping occurs because of reduction and oxidation.
As iron exhibits an advantage with respect to the thermody-
5
5­57
58­60
61­63
hydrogen generated in situ; PG,
HA,
and LA,
examples of value-added products, are important chemicals
for a wide range of applications in various industries, and
recently, the development of novel chemical processes for their
production from biomass has also attracted significant atten-
tion. Glucose (C H O ) is the most abundant natural hexose,
14
namics of reactions conducted in water, as well as reduction
starting from iron oxide,15 which is one of the most abundant
metals on earth; it is the most promising material for producing
hydrogen. A method well known for obtaining hydrogen from
metallic iron and water, referred to the steam iron process
6
12
6
and it is of significance to directly synthesize C3 platform
chemicals from glucose. We also demonstrated the reusability
of our system via reuse experiments and evaluated reaction
conditions and mechanism.
(
3Fe + 4H2O ꢀ Fe3O4 + 4H2), has long been employed.
Hence, we select iron metal and iron oxides as materials for
chemical looping. As shown in Figure 1, the proposed process
involves two steps. In the first step, biomass carbohydrates are
2
. Experimental
2
.1 Materials. All materials were purchased from com-
mercial suppliers and used as received without any pretreat-
ment or purification. Metallic iron particles with dimensions of
6
0­80 nm (NM-0019-UP, 99.9%); 5­9 ¯m (93-2601, 99.9%);
and 75 ¯m (00737, 200 mesh, 99+%) were purchased from
Ionic Liquids Technologies GmbH, Germany; Strem Chem-
icals, Inc., USA; and Alfa Aesar, UK; respectively. A carbon-
supported 5 wt % palladium catalyst (Pd/C) was obtained from
N. E. ChemCat Corporation, Japan. Glucose, fructose, HA, PG,
glycerol, LA, 1,2-butanediol, 1,2-hexanediol, glycolic acid, and
acetic acid of special grade were acquired from Wako Pure
Chemical Industries, Ltd., Japan. Calibration standard gases
were purchased from GL Sciences Inc., Japan.
2
.2 Reaction Apparatus and Procedure. Experiments
Figure 1. Hydrogenolysis by hydrogen generated in situ
via the chemical looping of a metal oxide redox. FeOx
denotes iron oxide (divalent or trivalent), and Fe represents
iron metal.
were conducted in a 64 mL stainless-steel autoclave (TSV-1,
Taiatsu Techno Corporation, Japan) or a 100 mL Hastelloy auto-
clave (MMJ-100, OM Lab-Tech Co. Ltd., Japan). In a typical
Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan | 1027

run using the 64 mL autoclave, 180 mg of glucose (1 mmol),
35 mg of Fe metal particles (6 mmol), 30 mg of the 5 wt % Pd/
Fe and Ru metal particle sizes were obtained by scanning
electron microscopy (SEM; S-4800, Hitachi High-Technologies
Corporation, Japan) and field-emission electron microscopy
(FE-TEM; JEM-2100F, JEOL Ltd., Japan), respectively. The
used iron oxide (1 g) after two rounds of reuse experiments
was heated with lignocellulosic biomass of empty fruit bunch
(EFB) powder (0.5 g; elemental composition: C, 51.1 wt %; H,
6.1 wt %; O, 42.5 wt %; N, 0.3 wt %; dry ash free) in a flow
3
C catalyst (Pd, 15 ¯mol), and 36 mL of ultrapure water were
loaded into the autoclave under anaerobic conditions. Second,
the reactor was purged with nitrogen three times for expelling
air, pressurized with 0.5 MPa of nitrogen, and heated to the
designated temperature, which was maintained constant for
the duration of the intended reaction, with stirring at 600­720
rpm. The temperature inside the reactor was monitored using a
thermocouple. Approximately 30 min was required to reach the
desired temperature, and approximately 1 h was required to cool
the reaction vessel. “Reaction time” indicates the time between
attaining the intended temperature and the beginning of cooling.
For reuse experiments, the catalysts were collected by filtration
and washed several times with water. Before the reuse experi-
ments, the recovered catalysts were reduced in a flow apparatus
3
¹1
apparatus with nitrogen (100 cm min ) at 1273 K for 2 h.
3
. Results and Discussion
3.1 Generation of Hydrogen from Fe Metal Particles and
Water. In a preliminary experiment, some iron metal particles
with different particle sizes generated hydrogen under hydro-
thermal conditions, as shown in Figure 2. The amount of the
generated hydrogen increased over time, and iron nanoparticles
with dimensions of 60­80 nm were the most effective for
hydrogen generation in terms of the formation rate and amount
of generated hydrogen because their specific surface area was
greater than those of larger iron particles. Hence, 60­80 nm iron
nanoparticles are selected as the hydrogenation-generating
agent in this study.
3.2 Transformation of Glucose Using Hydrogen Gener-
ated in situ. Figure 3 shows the transformation of glucose
using Fe metal particles or a Pd/C catalyst or the combination
of both Fe metal particles and a Pd/C catalyst (see Table S1 for
details on the analysis). All of the glucose was converted to
100 mol %-C.
The Fe metal particles resulted in a 19.3% yield of HA and a
16.4% yield of LA, while the use of only Pd/C resulted in a
20.2% yield of HA and a low yield (1.7%) of PG. Surprisingly,
the combination of Fe metal particles and Pd/C catalyst
resulted in the drastic increase in the PG yield (22.7%) along
with a decrease in the formation of HA and LA. The total yield
of C3 chemicals (PG, HA, and LA) was 34.6%. Notably, glu-
cose was selectively transformed to PG even though the hydro-
gen generated in situ exhibited very low pressure (approxi-
mately 0.20­0.30 MPa). Thus, the Fe + Pd/C system enables
3
¹1
with pure hydrogen (40 cm min ) at 723 K for 4 h, followed by
drying under vacuum at 343 K.
2
.3 Analytical Methods. After each reaction, the gaseous
products were collected in a gas sampling bag and analyzed by
gas chromatography (GC; GC-8A, Shimadzu Corporation,
Japan) equipped with both silica gel and molecular sieve 5A
columns, and a thermal conductivity detector. The liquid prod-
ucts were separated by centrifugation, decantation, and filtra-
tion using 0.45 ¯m filters and then analyzed using two high-
performance liquid chromatography systems (HPLC; 600 sys-
tem, Waters Corporation, USA; Prominence system, Shimadzu
Corporation, Japan) equipped with refractive index detectors.
An Aminex HPX-87H column (ø7.8 © 300 mm; mobile phase:
¹
1
¹1
aqueous solution of 5 mmol L formic acid; 0.7 mL min ;
13 K), and an Aminex HPX-87P column (ø7.8 © 300 mm;
3
¹
1
mobile phase: water; 1.0 mL min ; 358 K), both manufactured
by Bio-Rad Laboratories Inc., USA, were used. The products
were identified using standard substances and quantified by
an external standard. The total amount of organic carbon dis-
solved in water (liquid products) was determined using a total
organic carbon analyzer (TOC-LCSH/CSN, Shimadzu Corpora-
tion, Japan). The yields of the products (mol %-C) were
calculated on the basis of carbon content:
number of C atoms in product
Yield ½mol %-Cꢀ ¼
number of C atoms in substrate
number of moles of product
ꢁ
initial number of moles of substrate
ꢁ
100
ð1Þ
The catalysts were analyzed by X-ray powder diffraction
XRD) for determining the crystalline form using a diffractom-
(
eter (SmartLab, Rigaku Corporation, Japan) with Cu Kα
radiation. Diffraction peaks were assigned in accordance with
the Joint Committee of Powder Diffraction Standards (JCPDS)
card file data. X-ray fluorescence (XRF) and inductively cou-
pled plasma atomic emission spectroscopy (ICP-AES) were
employed for determining the catalyst composition and amounts
of dissolved metallic ions by using an X-ray fluorescence
spectrometer (ZSX Primus II, Rigaku Corporation, Japan) with
Rh Kα radiation. Brunauer­Emmett­Teller (BET) analysis was
conducted for determining the specific surface area of the cata-
Figure 2. Generation of hydrogen using Fe metal particles
of various sizes (60­80 nm, 5­9 ¯m, and 75 ¯m). Con-
ditions: Fe, 335 mg (6 mmol); water, 36 mL; 453 K; 20 h;
2
¹1
lysts by performing N adsorption at 77 K using an adsorption
2
N , 0.5 MPa. Specific surface area: 60­80 nm, 8.7 m g ;
2
2
¹1
analyzer (BELSORP-mini II, MicrotracBEL Corp., Japan). The
5­9 ¯m, 0.47 m g ; 75 ¯m, could not be measured.
1028 | Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan

Figure 3. Transformation of glucose using Fe metal
particles or a Pd/C catalyst or the combination of both
Fe metal particles and Pd/C catalyst. Conditions: glucose,
Figure 4. Reuse of the Fe + Pd/C catalyst system. Con-
ditions: glucose, 360 mg; Fe, 670 mg; 5% Ru/C, 120 mg;
water, 40 mL; 453 K; 20 h. The gas yield is not available
for Reuse 2.
1
4
80 mg; Fe, 335 mg; 5% Pd/C, 30 mg; water, 36 mL;
53 K; 20 h; N2, 0.5 MPa. FeOx + Pd/C: FeOx from Fe
after 20 h at 453 K; 5% Pd/C, 30 mg; N2, 0.5 MPa; H2,
.2 MPa. Products: propylene glycol (PG); hydroxyacetone
HA); lactic acid (LA); other chemicals (1,2-butanediol,
,2-hexanediol, glycolic acid, and acetic acid); gases
methane, ethane, propane, butane, carbon monoxide, and
0
(
1
(
carbon dioxide). Maximum error: «2%. Liquid and
gaseous product yields were determined by HPLC and
GC analysis, respectively.
the transformation of glucose to PG, HA, and LA with the
selective formation of PG, as shown in eq 2.
OH
OH
O
Pd/C
Fe/H
HO
HO
OH
O
OH
OH
OH
2
O
OH
OH
O
Glucose
Propylene glycol Hydroxyacetone Lactic acid
ð2Þ
Interestingly, the activity of the in situ hydrogen system for
PG formation was greater than that observed by the use of
an external hydrogen source. A reaction was conducted using
external hydrogen with iron oxide (FeOx) generated from the
metallic iron particles (Figure 4, the right-most column). The
yield of PG from gaseous hydrogen was 5.6%, which is less
than that (22.7%) observed with the use of in situ hydrogen.
Recently, Jin and co-workers have reported a similar phenom-
enon for the Pd/C-catalyzed conversion of glucose to PG via
water splitting with Zn.29
Figure 5. XRD patterns of iron materials: (a) fresh iron
metal, (b) iron oxide after two rounds of reuse, (c) Fe + Pd/
C reduced after three rounds of reuse; (d) Fe + Pd/C after
four rounds of reuse. Peak assignment: , Fe; , Fe3O4.
3.3 Reusability of the Fe + Pd/C System. The reusability
of the Fe + Pd/C system was examined by reuse experiments,
and Figure 4 shows the results. Using a fresh catalytic system,
the yield of PG increased from 17.3 to 23.9% in the first round
of reuse (Reuse 1), and it remained constant at around 23% in
the second and third rounds of reuse (Reuse 2 and 3). During
the reaction, the oxidation state of metallic iron was changed,
transforming to iron oxides. As shown in Figure 5, diffraction
peaks assigned to magnetite, Fe O (JCPDS card No. 01-085-
(Figure 5c). These results imply that hydrogen is generated
from the oxidation of metallic iron to magnetite by water.
In separate experiments, the regeneration of the used iron
oxide to metallic iron was demonstrated by thermal treatment
with lignocellulosic biomass as a practical method. As shown
in Figure S1, from the XRD pattern of the treated sample, the
used iron oxide was regenerated to metallic iron by thermal
treatment with lignocellulosic biomass. Iron oxide is hypothe-
3
4
1
436), and metallic iron, Fe (JCPDS card No. 01-087-0721),
were observed in the XRD patterns of the iron particles used
in the reuse experiment (Figures 5b and 5d), while diffraction
peaks corresponding to metallic iron were observed in the XRD
patterns of fresh iron metal (Figure 5a) and the reduced system
sized to be directly reduced by syngas (H and CO) or char
2
formed during the thermal decomposition of biomass, although
the method to regenerate the used iron oxide requires further
investigation.
Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan | 1029

The mechanism of the iron­water reaction under hydro-
6
4­66
thermal conditions has been deduced from previous studies.
Previously reported studies on zero-valent iron nanoparticles
2+
have demonstrated that during a liquid-phase reaction, Fe is
formed as iron hydroxide on a solid surface in water at ambi-
2
1,65
ent temperature.
decomposed to FeO under nitrogen at approximately 473 K,
while Fe(OH) converts to magnetite (Fe O ) in the absence of
In addition, Fe(OH) is well known to be
2
6
4
2
3
4
oxygen at temperatures of greater than 373 K (the sum of eqs 4
and 5: 3Fe(OH) ¼ Fe O + H + 2H O); this transformation
2
3
4
2
2
is known as the Schikorr reaction:6
5,66
Fe þ 2H2O ! FeðOHÞ2 þ H2
FeðOHÞ2 ! FeO þ H2O
FeO þ H2O ! Fe3O4 þ H2
ð3Þ
ð4Þ
ð5Þ
3
Total reaction:
Fe þ 4H2O ! Fe3O4 þ 4H2
3
ð6Þ
The particle sizes of Fe and Pd from the reuse experiments
appeared to be mostly unchanged (Table S3). Leaching of
metals from the reuse experiments was also investigated from
XRF and ICP-AES spectra. In all of the reuse experiments (see
Figure 4, violet dotted line), the dissolved iron species was
detected at a concentration of 520­640 ppm, indicating that the
Fe species is possibly involved in the reaction. The roles of the
Fe species are discussed in Section 3.5. On the other hand, the
leaching of Pd metal species was not observed (not shown
here). These results demonstrate that the Fe + Pd/C system is
recyclable not only as a chemical-looping material (hydrogen-
generating agent) but also as a hydrogenation catalyst.
3
.4 Effects of Reaction Conditions. The effects of reac-
tion conditions on the transformation of glucose using the Fe +
Pd/C system were evaluated. Figure 6a shows the effects of
temperature. The maximum yield of C3 chemicals was ob-
tained at 433 K. HA was the major product obtained in 31.1%
yield at 413 K. PG was the major product obtained in 22.7%
yield at 453 K and in 23.9% yield at 473 K. At higher
temperatures (473 and 493 K), the yields of gaseous products
increased with the simultaneous decrease of the C3 chemical
yield. At 493 K, the C3 chemicals further decreased with the
increase of unknown products, assumed to be condensation
6
7
products. Thus, the optimal temperatures for the formation
of HA and PG are 413 and 453 K, respectively. As needed, HA
or PG can be synthesized by setting the temperature of the
Fe + Pd/C system.
Figure 6. Effects of reaction conditions on the transforma-
tion of glucose in the Fe + Pd/C system. Conditions: glu-
cose, 180 mg; water, 36 mL; (a) Fe, 335 mg; 5% Pd/C, 30
mg; 20 h, (b) 5% Pd/C, 30 mg; 453 K; 20 h, (c) Fe, 335 mg;
Figures 6b and 6c show the effects of the amounts of Fe
metal particles and Pd/C for examining the selective formation
of PG. When Pd/C was used without the addition of Fe metal
particles, the yield of PG was 1.7%. The addition of Fe metal
particles drastically promoted the formation of PG along with
an increase in the amount of hydrogen generated, while the
formation of HA decreased (Figure 6b); the effect attained
saturation after the Fe amount reached 335 mg. The addition of
Pd/C promoted the formation of PG along with a decrease in
the formation of HA and LA (Figure 6c); the promotion effect
attained saturation after the amount of Pd/C reached 30 mg.
Thus, the optimal amounts of Fe metal particles and Pd/C for
the formation of PG are 335 and 30 mg, respectively.
4
53 K; 20 h.
3.5 Reaction Mechanism®Roles of Metallic Iron Nano-
particles and Pd/C Catalyst. The roles of metallic iron and a
Pd/C catalyst in the transformation of glucose were investi-
gated. In addition, the changes in product yield with time were
evaluated for the Fe + Pd/C system (see Figure S2). The
results indicate that fructose and HA are the intermediate
products in the formation of PG, and hydrogenation from HA
1030 | Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan

Scheme 1. Possible reaction pathway for the transformation of glucose using the Fe + Pd/C system.
Table 1. Transformation of various substrates using Fe or Pd/C, or both Fe and Pd/Ca),b)
Time Conversion
Product yield/mol %-C
Entry
Substrate Catalyst
/h
/mol %-C
Fructose
HA
PG
Others
Gas
1
2
3
4
5
6
Glucose
Glucose
Glucose
HA
HA
HA
Fe + Pd/C*
Fe
Pd/C*
Fe + Pd/C
Fe
0
69.2
59.0
15.1
72.7
12.6
0.2
22.3
23.7
0.0
®
9.3
7.7
0.0
®
0.0
0.0
0.0
64.2
4.9
0.0
2.5
0.0
0.0
0.0
0.0
0.0
0.4
0.1
0.5
0.1
0.0
0.2
0
0
4
4
4
®
®
®
®
Pd/C + H2 0.4 MPa
a) Reaction conditions: glucose, 180 mg; HA, 148 mg; Fe, 335 mg; 5% Pd/C, 30 mg or *60 mg; water, 36 mL;
N2, 0.5 MPa; 453 K. b) Abbreviations: HA, hydroxyacetone; PG, propylene glycol; Others, sum of 1,2-
butanediol, 1,2-hexanediol, acetic acid, glycolic acid, and lactic acid; Gas, sum of C1­C4 alkanes, CO2, and CO.
to PG is the rate-determining step, as demonstrated in our
previous study using a ZnO + Ru/C system.68 It is hypothe-
sized that by using the Fe + Pd/C system, the transformation
of glucose to PG also involved the isomerization of glucose to
fructose, retro-aldol reaction of fructose to trioses (glyceralde-
hyde and HA can be formed from each other via consecutive
keto­enol tautomerization ), dehydration of glyceraldehyde to
pyruvaldehyde, and hydrogenation of pyruvaldehyde to PG via
HA (see Scheme 1).
The roles of both Fe and Pd/C during intermediate steps in
the transformation from glucose to PG were also investigated.
First, the isomerization of glucose to fructose was evaluated.
As shown in Table 1, the use of Fe + Pd/C as well as only Fe
afforded fructose in yields of 22.3 and 23.7%, respectively, at
and co-workers have suggested that metal cations (including
ferrous ions) catalyze a series of reaction steps, involving the
isomerization of glucose to fructose, the retro-aldol reaction
(the selective cleavage of the C3­C4 bond) of fructose to
produce trioses, and the formation of lactic acid via pyruval-
6
3
dehyde. The role of homogeneous and heterogeneous Fe
species was evaluated in separate experiments for the formation
of HA and LA from glucose (see Table S2). Based on these
results, homogeneous Fe species, such as ferrous ions, has been
speculated to catalyze the transformation of glucose to LA via
pyruvaldehyde, and pyruvaldehyde is immediately converted to
HA in the presence of a hydrogenation catalyst and hydrogen.
Next, HA was used as a starting material. The yield of PG
was 64.2% when the Fe + Pd/C system was used at 433 K for
4 h. No PG was formed using only Pd/C, and a yield of 4.9%
for PG was obtained using only Fe. The Fe + Pd/C system
clearly exhibited a synergistic effect for catalyzing the hydro-
genation of HA to PG. Based on results reported in previous
studies on the hydrogenation of carbonyl compounds over
63
4
53 K for 0 h (reaction time of 0 implies just after increasing
the temperature), while no fructose was formed when only Pd/
C was used. Thus, it is clear that Fe catalyzes the isomerization
of glucose to fructose. As discussed in Section 3.3, the reac-
tion mixture of glucose at 453 K for 20 h contained dissolved
iron species at a concentration of around 600 ppm. The con-
centration of ferrous ions (Fe2 ) in the reaction mixture was
specifically determined to be approximately 600 ppm using a
gas detector tube by the colorimetric method with 1,10-
phenanthroline, indicating that a majority of the dissolved Fe
species are estimated to be present as divalent cations. Wang
6
9­71
metal catalysts,
the Fe species is speculated to activate the
+
carbonyl group of HA molecules via the donation of electrons
from unshared electron pairs of oxygen atoms to the electron-
deficient Fe species, in agreement with the hydrogenation of
the carbonyl group of HA molecules over the Pd metal catalyst.
Richard and co-workers have reported that the activity and
Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan | 1031

selectivity of platinum for the liquid-phase hydrogenation of
cinnamaldehyde to cinnamyl alcohol are improved by the addi-
3
4044.
4
G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106,
A. J. Ragauskas, C. K. Williams, B. H. Davison, G.
7
0
tion of FeCl2. They have also demonstrated that a part of iron
atoms are deposited on Pt, as evidenced by the composition
analysis of the metal particles determined by an energy-
dispersive X-ray emission spectrometer attached to a scanning
transmission electron microscope with a field-emission gun
Britovsek, J. Cairney, C. A. Eckert, W. J. Frederick, Jr., J. P.
Hallett, D. J. Leak, C. L. Liotta, J. R. Mielenz, R. Murphy, R.
Templer, T. Tschaplinski, Science 2006, 311, 484.
5
P. M. Mortensen, J.-D. Grunwaldt, P. A. Jensen, K. G.
Knudsen, A. D. Jensen, Appl. Catal., A 2011, 407, 1.
A. M. Ruppert, K. Weinberg, R. Palkovits, Angew. Chem.,
Int. Ed. Engl. 2012, 51, 2564.
R. M. Navarro, M. C. Sánchez-Sánchez, M. C. Alvarez-
Galvan, F. del Valle, J. L. G. Fierro, Energy Environ. Sci. 2009, 2,
5.
(FEG-STEM). For determining whether the Fe species was
6
involved in the hydrogenation of HA to PG using the Pd/C
catalyst, the distribution of iron and palladium atoms was
measured by an energy-dispersive X-ray spectrometer (EDS)
attached to a field-emission transmission electron microscope
7
3
(FE-TEM) (see Figure S3). The results revealed that dissolved
8
A. Thursfield, A. Murugan, R. Franca, I. S. Metcalfe,
Energy Environ. Sci. 2012, 5, 7421.
J. Turner, G. Sverdrup, M. K. Mann, P.-C. Maness, B.
Fe species are deposited on the palladium catalyst, as has been
7
0
reported previously. Based on these results, metallic Fe and
Fe species appear to play multiple roles in the transformation
of glucose to PG, containing not only a hydrogen-generating
agent, but also a degradation catalyst of glucose to trioses, and
in the promotion of the hydrogenation of HA to PG with a
Pd/C catalyst.
9
Kroposki, M. Ghirardi, R. J. Evans, D. Blake, Int. J. Energy Res.
2008, 32, 379.
10 V. Hacker, R. Fankhauser, G. Faleschini, H. Fuchs, K.
Friedrich, M. Muhr, K. Kordesch, J. Power Sources 2000, 86, 531.
11 K. Otsuka, T. Kaburagi, C. Yamada, S. Takenaka, J. Power
Sources 2003, 122, 111.
4
. Conclusion
12
13
14
T. Kodama, N. Gokon, Chem. Rev. 2007, 107, 4048.
A. Steinfeld, Sol. Energy 2005, 78, 603.
M. Thaler, V. Hacker, Int. J. Hydrogen Energy 2012, 37,
In this study, we developed a unique process for the hydro-
genolysis or deoxygenation of biomass carbohydrates with
hydrogen generated in situ using metallic iron and a palladium
catalyst system. The Fe + Pd/C system led to the transforma-
tion of glucose to C3 platform chemicals (PG, HA, and LA) in
2
800.
1
5
K. Otsuka, S. Murakoshi, A. Morikawa, Fuel Process.
Technol. 1983, 7, 203.
C. Boix, N. Poliakoff, J. Chem. Soc., Perkin Trans. 1 1999,
1
6
3
4.6% yield, with PG in 22.7% yield, at 453 K, as well as C3
11, 1487.
platform chemicals in 34.4% yield, with HA in 31.1% yield
at 413 K, without the addition of hydrogen from an external
source. We hypothesized that iron played multiple roles®
generating hydrogen, catalyzing the degradation of glucose to
trioses, and assisting hydrogenation®while the palladium
catalyst promoted the hydrogenation step, resulting in the for-
mation of PG. We believe that this technique can be a com-
plementary process and an alternative to most current applica-
tions of hydrogenation and reduction, which are limited by the
hazards of high-pressure operation of hydrogen and the high
cost of hydrogen storage and transformation. Currently, further
studies on the enhancement of product yield and selectivity,
improvement in the reusability of the Fe + Pd/C system, the
reason for high activity of hydrogen generated in situ, the
oxidation state of the active iron species, and the synergetic
effect of Fe species with Pd/C are underway in our laboratory.
1
7
L. Wang, P. Li, Z. Wu, J. Yan, M. Wang, Y. Ding, Synthesis
003, 13, 2001.
S. Mukhopadhyay, G. Rothenberg, D. Gitis, M. Baidossi,
D. E. Ponde, Y. Sasson, J. Chem. Soc., Perkin Trans. 2 2000, 1809.
S. Mukhopadhyay, G. Rothenberg, D. Gitis, Y. Sasson,
Org. Lett. 2000, 2, 211.
S. Mukhopadhyay, G. Rothenberg, H. Wiener, Y. Sasson,
2
1
8
1
9
2
0
New J. Chem. 2000, 24, 305.
21 X. Li, D. W. Elliott, W. Zhang, Crit. Rev. Solid State Mater.
Sci. 2006, 31, 111.
22 Z. Huo, M. Hu, X. Zeng, J. Yun, F. Jin, Catal. Today 2012,
194, 25.
23 F. Jin, Y. Gao, Y. Jin, Y. Zhang, J. Cao, Z. Wei, R. L. Smith,
Jr., Energy Environ. Sci. 2011, 4, 881.
2
4
F. Jin, X. Zeng, J. Liu, Y. Jin, L. Wang, H. Zhong, G. Yao,
Z. Huo, Sci. Rep. 2014, 4, 4503.
L. Lyu, F. Jin, H. Zhong, H. Chen, G. Yao, RSC Adv. 2015,
, 31450.
L. Lyu, X. Zeng, J. Yun, F. Wei, F. Jin, Environ. Sci.
Technol. 2014, 48, 6003.
X. Zeng, M. Hatakeyama, K. Ogata, J. Liu, Y. Wang, Q.
2
5
5
The authors gratefully acknowledge the financial support
of JSPS KAKENHI (Grant-in-Aid for Exploratory Research)
Grant No. 26620147 and an endowment fund by Nippon
Shokubai Co., Ltd.
2
6
2
7
Gao, K. Fujii, M. Fujihira, F. Jin, S. Nakamura, Phys. Chem.
Chem. Phys. 2014, 16, 19836.
Supporting Information
2
8
Z.-B. Huo, J.-K. Liu, G.-D. Yao, X. Zeng, J. Luo, F.-M. Jin,
Appl. Catal., A 2015, 490, 36.
J. Wang, G. Yao, Y. Wang, H. Zhang, Z. Huo, F. Jin, RSC
Adv. 2015, 5, 51435.
J. Xiao, Z. Huo, D. Ren, S. Zhang, J. Luo, G. Yao, F. Jin,
Some figures and tables for the reaction results and the
characterization of the Fe and Pd materials. This material is
available on http://dx.doi.org/10.1246/bcsj.20160114.
2
9
3
0
References
Process Biochem. 2015, 50, 793.
1
J. N. Chheda, G. W. Huber, J. A. Dumesic, Angew. Chem.,
Int. Ed. Engl. 2007, 46, 7164.
A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411.
31 L. Xu, Z. Huo, J. Fu, F. Jin, Chem. Commun. 2014, 50,
6009.
32 C. He, G. Tian, Z. Liu, S. Feng, Org. Lett. 2010, 12, 649.
2
1032 | Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan

3
3
G. Tian, H. Yuan, Y. Mu, C. He, S. Feng, Org. Lett. 2007,
, 2019.
R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 2002,
18, 964.
R. R. Davda, J. W. Shabaker, G. W. Huber, R. D. Cortright,
J. A. Dumesic, Appl. Catal., B 2005, 56, 171.
C. A. Fisk, T. Morgan, Y. Ji, M. Crocker, C. Crofcheck,
S. A. Lewis, Appl. Catal., A 2009, 358, 150.
A. Martin, U. Armbruster, I. Gandarias, P. L. Arias, Eur. J.
Lipid Sci. Technol. 2013, 115, 9.
D. Roy, B. Subramaniam, R. V. Chaudhari, Catal. Today
010, 156, 31.
S. Yin, A. K. Mehrotra, Z. Tan, Biomass Bioenergy 2012,
7, 228.
52 N. Kobayashi, L.-S. Fan, Biomass Bioenergy 2011, 35,
1252.
53 F. Li, L. Zeng, L.-S. Fan, Fuel 2010, 89, 3773.
54 S. Y. Luo, J. Fu, J. Renewable Sustainable Energy 2015, 7,
10.
55 M. A. Dasari, P.-P. Kiatsimkul, W. R. Sutterlin, G. J.
Suppes, Appl. Catal., A 2005, 281, 225.
56 X. Wang, L. Meng, F. Wu, Y. Jiang, L. Wang, X. Mu,
Green Chem. 2012, 14, 758.
57 Z. Zhang, J. E. Jackson, D. J. Miller, Appl. Catal., A 2001,
219, 89.
9
4
3
4
3
5
3
6
3
7
3
8
2
4
2
2
58 T. Deng, H. Liu, J. Mol. Catal. A: Chem. 2014, 388­389,
3
9
66.
59 A. K. Kinage, P. P. Upare, P. Kasinathan, Y. K. Hwang,
J.-S. Chang, Catal. Commun. 2010, 11, 620.
60 C. R. Vitasari, G. W. Meindersma, A. B. de Haan, Chem.
Eng. Res. Des. 2015, 95, 133.
61 M. Dusselier, P. Van Wouwe, A. Dewaele, E. Makshina,
B. F. Sels, Energy Environ. Sci. 2013, 6, 1415.
62 P. Mäki-Arvela, I. L. Simakova, T. Salmi, D. Yu. Murzin,
Chem. Rev. 2014, 114, 1909.
63 Y. Wang, W. Deng, B. Wang, Q. Zhang, X. Wan, Z. Tang,
Y. Wang, C. Zhu, Z. Cao, G. Wang, H. Wan, Nat. Commun. 2013,
4, 2141.
64 J. D. Bernal, D. R. Dasgupta, A. L. Mackay, Clay Miner.
Bull. 1959, 4, 15.
65 U. R. Evans, J. N. Wanklyn, Nature 1948, 162, 27.
66 M. Ma, Y. Zhang, Z. R. Guo, N. Gu, Nanoscale Res. Lett.
2013, 8, 16.
67 M. Möller, P. Nilges, F. Harnisch, U. Schröder,
ChemSusChem 2011, 4, 566.
68 Y. Hirano, K. Sagata, Y. Kita, Appl. Catal., A 2015, 502, 1.
69 V. Ponec, Appl. Catal., A 1997, 149, 27.
70 D. Richard, J. Ockelford, A. Giroir-Fendler, P. Gallezot,
Catal. Lett. 1989, 3, 53.
4
0
D. Scholz, C. Aellig, I. Hermans, ChemSusChem 2014, 7,
68.
4
1
Y. Wu, F. Gu, G. Xu, Z. Zhong, F. Su, Bioresour. Technol.
013, 137, 311.
Y. Xu, J. Long, Q. Liu, Y. Li, C. Wang, Q. Zhang, W. Lv, X.
Zhang, S. Qiu, T. Wang, L. Ma, Energy Convers. Manage. 2015,
9, 188.
L. He, L.-C. Wang, H. Sun, J. Ni, Y. Cao, H.-Y. He, K.-N.
Fan, Angew. Chem., Int. Ed. Engl. 2009, 48, 9538.
S.-S. Li, X. Liu, Y.-M. Liu, H.-Y. He, K.-N. Fan, Y. Cao,
Chem. Commun. 2014, 50, 5626.
4
2
8
4
3
4
4
4
5
A. Tlili, J. Schranck, H. Neumann, M. Beller, Chem.®Eur.
J. 2012, 18, 15935.
4
6
L. Yu, X.-L. Du, J. Yuan, Y.-M. Liu, Y. Cao, H.-Y. He,
K.-N. Fan, ChemSusChem 2013, 6, 42.
J. 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.
C. Hargus, R. Michalsky, A. A. Peterson, Energy Environ.
Sci. 2014, 7, 3122.
P. Krenzke, K. Krueger, N. Leonard, S. Duncan, R. D.
Palumbo, S. Möller, J. Sol. Energy Eng. 2010, 132, 034501.
4
7
4
8
4
9
5
0
R. W. Breault, E. R. Monazam, J. Ind. Eng. Chem. 2015,
71 W. Yu, H. Liu, M. Liu, Q. Tao, J. Mol. Catal. A: Chem.
1999, 138, 273.
2
9, 87.
5
1
S. Luo, C. Yi, Y. Zhou, Renewable Energy 2011, 36, 3332.
Bull. Chem. Soc. Jpn. 2016, 89, 1026–1033 | doi:10.1246/bcsj.20160114
© 2016 The Chemical Society of Japan | 1033