Oxidative hydrolysis of cellobiose to glucose

Article abstract of DOI:10.1007/s10562-010-0532-8

Cellobiose hydrolysis into glucose was chosen as a model system for cellulose breakdown to investigate glycosidic bond cleavage. The hydrolysis was enhanced by increased acidity in an inert gas medium, while air-assisted hydrolysis with a neutral solution achieved over 70% glucose yield. Hydrogen peroxide, as a stronger oxidant than air, converted cellobiose to carboxyl compounds, which lowered the glucose selectivity. At 150 °C, the selectivity from cellobiose to glucose was very low on porous γ-Al₂O ₃ supported catalysts, even lower than without a catalyst. When the active metals were prepared on non-porous supports such as spherical alumina (α phase), the overall yield of glucose was dramatically improved at 120 °C. Similar improvements were obtained for another disaccharide model, sucrose, which achieved greater than 90% sucrose conversion with selectivity in excess of 90% at 80 °C. Graphical Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-010-0532-8

Catal Lett (2011) 141:498–506
DOI 10.1007/s10562-010-0532-8
Oxidative Hydrolysis of Cellobiose to Glucose
Weiling Deng
Steven T. Christensen
Christopher L. Marshall
•
Rodrigo Lobo
•
Worajit Setthapun
•
•
•
Jeffrey W. Elam
Received: 28 September 2010 / Accepted: 1 December 2010 / Published online: 5 January 2011
Ó Springer Science+Business Media, LLC (outside the USA) 2010
Abstract Cellobiose hydrolysis into glucose was chosen
as a model system for cellulose breakdown to investigate
glycosidic bond cleavage. The hydrolysis was enhanced by
increased acidity in an inert gas medium, while air-assisted
hydrolysis with a neutral solution achieved over 70% glu-
cose yield. Hydrogen peroxide, as a stronger oxidant than
air, converted cellobiose to carboxyl compounds, which
lowered the glucose selectivity. At 150 °C, the selectivity
from cellobiose to glucose was very low on porous c-Al O
supported catalysts, even lower than without a catalyst.
When the active metals were prepared on non-porous sup-
ports such as spherical alumina (a phase), the overall yield
of glucose was dramatically improved at 120 °C. Similar
improvements were obtained for another disaccharide
model, sucrose, which achieved greater than 90% sucrose
conversion with selectivity in excess of 90% at 80 °C.
1 Introduction
In 2009, the Energy Information Administration (EIA) [1]
projected that the use of renewable energy sources such as
ethanol will increase steadily at 3.3% per year. It was
expected by EIA that more than 40% of the ethanol will
come from cellulosic biomass by 2030. Ligno-cellulose is
the most abundant organic compound in nature. The biggest
obstacle to producing ethanol and other transportation fuels
from ligno-cellulosic biomass is the high costs associated
with the many processing steps [2]. Economical and clean
processes with limited steps are needed to convert biomass
to liquid fuels [3]. For future fuel and chemical production,
the great challenge is to develop an environmentally friendly
method of cellulose breakdown that is cost competitive with
petroleum-driven chemical commodities.
2
3
Cellulose hydrolysis into glucose is a key step for
biomass conversion. Glucose is a renewable feedstock
molecule that can be converted into fuels, chemical com-
modities, foods, and medicines [4]. A commercial method
of breaking down cellulose is an enzymatic process, which
has some drawbacks such as the low thermal stability of the
enzymes and problems with separation and recovery of the
enzymes from the products. Although the concentrated
acid-catalyzed hydrolysis of cellulose was industrialized
almost a century ago, reactor corrosion and waste acid
treatment have driven the search for a more environmen-
tally friendly alternative. Dilute sulfuric acid hydrolysis has
a glucose yield of 70% [5], and the chemistry behind it is
not well understood and still a research focus today. Solid
acid catalysts have been investigated to replace mineral
acids in cellulose hydrolysis [6–8]. However, the hydro-
thermal instability of catalysts tends to produce large
Keywords Cellobiose ꢀ Sucrose ꢀ Glucose ꢀ
Oxidative hydrolysis ꢀ Non-porous support
W. Deng ꢀ R. Lobo ꢀ W. Setthapun ꢀ C. L. Marshall (&)
Chemical Sciences and Engineering Division, Argonne National
Laboratory, 9700 South Cass Avenue, Argonne, IL 60439, USA
e-mail: marshall@anl.gov
S. T. Christensen ꢀ J. W. Elam
Energy Systems Division, Argonne National Laboratory,
700 South Cass Avenue, Argonne, IL 60439, USA
9
Present Address:
W. Setthapun
Chiang Mai Rajabhat University, Chiang Mai 50300, Thailand
2
-
amounts of SO4
and by-products in the resultant
solution.
1
23

Catalytic Oxidative Hydrolysis of Disaccharides
499
The non-enzymatic methods reported in the literature to
Table 1 Catalyst list and preparation methods
break down cellulose into C -sugars, to syn gas (CO ? H ),
6
2
Catalyst
Metal loading
wt%)
Preparation
and to chemicals include gasification [9], pyrolysis [10],
ionic liquid catalysis [11], and hydrogenolysis over precious
metal catalysts [12–16]. While these results are encourag-
ing, the process requires costly hydrogen as a feed gas and,
in some cases, suffers from low selectivity.
(
c-Al
VO
Au/c-Al
Pt/c-Al
Pd/c-Al
sp-Al
Pt/sp-Al
Au/sp-Al
2
O
3
0
4
3
5
5
0
3
2
–
x
/c-Al
2
O
3
Incipient wetness impregnation
Electrostatic adsorption
Incipient wetness impregnation
Incipient wetness impregnation
–
2 3
O
2
O
3
Instead of using hydrogen, Schutt et al. [17] employed
wet oxidation of cellulose in a monolith reactor with a Pd
catalyst and reported a glucose yield of 40%. However,
also formed were large quantities of acetic acid, as well as
malic, succinic, and oxalic acids.
2 3
O
2 3
O
2
O
3
Atomic layer deposition
Atomic layer deposition
2 3
O
Cellulose is composed of hundreds to thousands of
D-glucose units, which are linked by b-(1,4)-glycosidic
bonds. A simple model system for cellulose is cellobiose, a
dimer of D-glucose units. Studies on cellobiose hydrolysis
into glucose can help understand the cleavage of the gly-
cosidic bond. Cellobiose has been used as a model system
in several hydrolysis studies [18–20]. The maximum yield
of glucose was 36.8% at 320 °C and 5076 psi in the sub-
critical water region [21]. A single-step conversion of
cellobiose into sorbitol was demonstrated by using Ru
catalyst under 580 psi hydrogen [19]. Solid acid catalysts
such as propylsulfonic acid-functionalized silica can com-
pletely hydrolyze cellobiose at 175 °C with some degra-
dation of glucose (*20%) under the same conditions [22].
In this paper, we explore a different approach, oxidative
hydrolysis using air to assist the cellobiose hydrolysis. The
goal is to convert cellobiose to glucose while minimizing
the breakage of non-glycosidic bonds, including the C–C,
C–OH, and C–O bonds in the glucose ring. The reaction
parameters (temperature, solution pH, duration time, gas
medium and pressure, and catalyst type) were investigated
to optimize the cellobiose conversion and glucose selec-
tivity. For comparison, another disaccharide model com-
pound, sucrose (isomer of cellobiose), was tested in
selected conditions as well.
Aldrich, Inc.) and oxalic acid (99%, Aldrich, Inc.) in a 1:2
weight ratio. After impregnation in the c-Al O support, the
2
3
samples were dried in air at 100 °C overnight and calcined in
static air at 400 °C for 4 h.
2.2 Preparation of Au/c-Al O by Electrostatic
2
3
Adsorption
The conventional impregnation method to prepare a gold
catalyst would produce large particles due to its tendency
to agglomeration. In this work, Au/c-Al O was prepared
2
3
by electrostatic adsorption [23]. Briefly, c-Al O was sus-
2
3
pended in deionized (D.I.) water. Then, 1 M HNO solu-
3
tion was used to adjust the pH of the slurry to 2, which was
lower than the point of zero charge of c-Al O (typically,
2
3
6–8). A desired amount of chloroauric acid (HAuCl ꢀ3H O,
4
2
Aldrich, Inc.) was dissolved in D.I. water, then dropwise
added into the pH-adjusted slurry. The positively charged
-
surface can adsorb anionic metal complex ions (AuCl ) .
4
After aging for 1 h, the precipitate was filtered and washed,
then dried at 100 °C overnight and calcined in static air at
400 °C for 4 h.
2.3 Preparation of Pt/sp-Al O and Au/sp-Al O
2 3 2 3
by Atomic Layer Deposition
2
Experimental
A non-porous support of spherical alumina (sp-Al O ,
2
3
Table 1 list the preparation methods and metal loadings of
the catalysts used in this work.
a phase, Alfa Aesar, NanoDur) was chosen to deposit Pt or
Au by atomic layer deposition (ALD). For the preparation
of Pt/sp-Al O , the sp-Al O support was spread in a
2
3
2 3
2
.1 Preparation of Pd/c-Al O , Pt/c-Al O , and VO /
stainless steel tray covered by a stainless steel mesh to
contain the powder while allowing easy access by the
precursors [24]. The Pt coating was accomplished by using
alternating exposures to (methylcyclopentadienyl) trim-
2
3
2
3
x
c-Al O by Incipient Wetness Impregnation
2
3
The c-Al O support was impregnated with a solution of
2
3
Pd(NO ) or Pt(NH ) (NO ) with appropriate concentra-
3
ethylplatinum (MeCpPtMe ) and oxygen at 300 °C [25] in
2
3 4
3 2
3
tion, whose volume equaled the total pore volume of the
a viscous-flow ALD reactor [26]. The MeCpPtMe was
3
support. The VO /c-Al O catalyst was prepared by incipi-
ent wetness impregnation of c-alumina (226 m /g, BASF)
with an aqueous solution of ammonium metavanadate (99%,
contained in a stainless steel bubbler heated to 50 °C to
increase the vapor pressure of the precursor. The reactant
x
2 3
2
-
exposures were 200 s with partial pressures of 1 910 psi
3
123

5
00
W. Deng et al.
-
0.05 torr) for MeCpPtMe and 4 9 10 psi (0.20 torr)
3
(
used to detect possible unsaturated acids. Each sample was
diluted 10 times by D.I. water, then 20 ll of the diluted
sample was injected into the HPLC, and the products were
analyzed by an external standard method using peak area
response.
3
for oxygen. Nitrogen purge periods of 50 s were used
between reactant exposures. The sample mass was mea-
sured with an analytical balance before and after the ALD
to determine the Pt loading.
The Au/sp-Al O was prepared in a similar manner with
2
The conversion of cellobiose was calculated by
Conversion ð% ) ¼
3
a precursor of dimethyl (acetylacetonate) gold (III)
Au(acac)Me ) except that the temperature for bubbler and
ðinitial moles of cellobioseꢁ moles of cellobiose in products)
(
2
ꢂ 100%
ðinitial moles of cellobiose)
The selectivity of glucose was calculated by
dosing was maintained at 40–50 °C. The as-prepared cat-
alysts were tested without pretreatment.
ðmoles of glucose in productsÞ
ðinitial moles of cellobiose ꢁ moles of cellobiose in products) ꢂ 2
Selectivity ð% ) ¼
ꢂ 100%
2
.4 Cellobiose Hydrolysis Reaction
The yield of glucose was calculated by
The hydrolysis reactions were conducted in a stainless-
steel autoclave (50 ml, Autoclave Engineers, Inc.). The
reactor was equipped with a magnet-drive stirrer, and the
temperature was maintained by an electric heating mantle
with a temperature controller. Typically, 30 g of 10 wt%
cellobiose (C98%, Sigma-Aldrich) in water solution and/or
ðmoles of glucose in productsÞ
Yield ð% ) ¼
ꢂ 100%
ðinitial moles of cellobiose) ꢂ 2
2.6 Temperature-Programmed Reduction
with Hydrogen (H -TPR)
2
0
.1 g catalyst were used in each reaction. After the cello-
The H -TPR was conducted in an Altamira AMIx90
2
biose solution and/or catalyst were added, the reactor was
purged with Ar four times before charging the required gas
or gas mixtures. The reaction was held at each reaction
temperature for 2, 5, or 8 h, then cooled down to room
temperature by an internal cooling water loop in the reactor
prior to depressurization.
instrument equipped with an electrically heated tube fur-
nace. An on-line mass detector (Dynacor) analyzed the
effluent of the reactor to determine the H consumption and
2
the evolution of other gaseous products. The as-prepared or
the used catalysts after reaction in fine powder form were
diluted in SiC and placed inside a stainless steel flow
reactor and heated at a rate of 8 °C/min from room tem-
perature to ca. 650 °C in a 3.5%H /He gas.
2
2
.5 Product Analysis
The gas products were analyzed by an Agilent HP-5890
gas chromatograph (GC) equipped with a thermal
conductivity detector to examine the CO and CO2
formation.
3 Results and Discussion
3.1 Mechanism of Cellobiose Hydrolysis
High pressure liquid chromatography (HPLC, Agilent
Figure 1 shows the proposed mechanism of cellobiose
degradation by either hydrolysis or hydrogenation. Over-
oxidation of cellobiose to form aldehydes, ketones, car-
bonyl compounds, and acids is not included. Overall, the
1
100 series) was used to separate and quantify the liquid
products after separation from the solid catalysts by fil-
?
tration. A 300 9 7.8 mm Rezex RNM-carbohydrate Na
0
(
8%) column (OOH-0136-K0) separated the reaction
hydrolysis reaction takes place via path A (1,4 bond
products. A refractive index (RI) detector quantified the
concentrations of cellobiose, glucose, xylose, fructose, and/
or other derivative acids, such as glucuronic acid and
gluconic acid. The flow rate of the water mobile phase was
cleavage). The major product is glucose. This pathway is
corroborated by the presence of a small amount of HMF
and 2-furfural, which are the typical products from glucose
through isomerization and dehydration. The absence of
glucopyranosyl-glucitol indicates that oxygen (1,5 bond)
removal from the ring is not possible. If hydrolysis occurs
through both cleavage A and B together, sorbitol will be
the final product. However, HPLC analysis did not identify
0
.4 mL/min. The temperature of the column and that of the
RI detector were both set at 55 °C. A UV–Vis detector at
80 nm wavelength quantified 5-hydroxymethyl-2-furfural
HMF) and 2-furfural. Another wavelength, 210 nm, was
2
(
1
23

Catalytic Oxidative Hydrolysis of Disaccharides
501
Fig. 1 Proposed mechanism of
cellobiose degradation through
0
1
,4 cleavage and 1,5 cleavage
any sorbitol production. Hydrogenation products through
cleavage A or cleavage A ? B are included in Fig. 1 based
on results from Yan et al. [19]. By using either Pd catalyst
or Ru catalyst, one can control the hydrogenation selec-
tivity to glucose or sorbitol.
1
00
80
60
40
20
0
Air
Air
3
.2 Optimal Temperature for Cellobiose Hydrolysis
Ar
H
2
To locate the optimal temperature for the cellobiose
hydrolysis reaction, experiments were carried at reaction
temperatures from 100 to 300 °C. The pressure in the
reactor was set at 1200 psi in Ar gas. No reaction was
observed at 100 °C. The cellobiose conversion increased
with temperature, reaching 93.9% at 200 °C. However,
the glucose selectivity at both 200 °C (24.6%) and
Air
Ar
H
2
H2
Ar
2
5
8
Reaction time (h)
Fig. 2 Glucose yield in cellobiose hydrolysis reaction in various
gases. Reaction conditions: temperature, 150 °C; total pressure,
3
00 °C (5.7%) was much less than that at 150 °C
1
2
00 psi; 30 g of 10 wt% cellobiose/H O solution
(
92.7%). Significant black char formation inside the
reactor in reactions conducted at 200 and 300 °C was
the major indication of the loss of glucose selectivity.
Other byproducts (including fructose, HMF, and 2-fur-
fural) were also identified by HPLC. These results are
consistent with a previous study in which the chemistry
for the decomposition pathways of cellulose and glu-
cose in water indicated that glucose undergoes isomer-
ization to form fructose, which can then undergo
dehydration to form HMF [21]. In this temperature
effect study, those byproducts only contributed to
B10% of the selectivity. Therefore, for all experiments,
a reaction temperature of 150 °C or lower was chosen
for the remaining studies.
3.3 Effect of Atmosphere on Cellobiose Hydrolysis
Most previous work regarding conversion of cellobiose or
cellulose to C -sugars used precious metal catalysts under
6
reductive conditions [12, 13, 19]. Here, the glucose yields
at 2, 5, and 8 h reaction over H , Ar, and air were
2
compared (Fig. 2). Those gases were used separately in
the 150 °C reaction, and the pressure was set at 100 psi
(at room temperature). In the H medium, the glucose yield
2
increased slowly to 26.5% after 8 h reaction. In the Ar
medium, the glucose yield was a little higher than in H2,
but still less than 35% at the end of the 8 h reaction.
123

5
02
W. Deng et al.
It appeared that air-assisted hydrolysis gave much better
However, at 100 °C, hydrolysis of the glycosidic bond only
occurred very slowly. In this work, the fast hydrolysis and
beginning of oxidation of the functional groups began at
150 °C, which is far from the pyrolysis condition. In
addition to the hydrolysis route through the glycosidic
bond, oxidation may happen at this temperature, depending
on the oxidation strength of the system during the reaction.
glucose yield compared to Ar or H . The yield reached a
2
maximum of 72.6% at 5 h reaction, and it remained nearly
constant with an additional 3 h under the same condition,
indicating the reaction was at equilibrium. Even though the
glucose yields were the same for the air-assisted reaction at
5
and 8 h, the glucose selectivity differed markedly (94.5%
for the 5 h and 77.5% for the 8 h reaction), suggesting that
the decomposition rate of glucose was higher than the
cellobiose hydrolysis rate. In fact, significant conversion of
glucose under thermal conditions was found in the wet
oxidation of carbohydrate-containing waste streams [22].
Hydrogen should not have a significant cellobiose con-
version under the neutral condition (pH 7); this observation
was confirmed by another group [19] not using any cata-
lyst. The glucose yield being *20% under hydrogen is due
to the weak acidic solution used (pH 5.5 in D.I. water).
Acid-catalyzed hydrolysis can explain why, in Fig. 2, both
hydrogen and the inert gas argon give similar results.
However, air significantly promotes the reaction. It is
plausible that the enhanced solubility of oxygen in the
solution due to the pressure (100 psi) weakens the glyco-
sidic bond, which helps the hydrolysis reaction. This oxi-
dation condition is so mild that no over-oxidation product,
such as carboxyl compounds, is formed.
3.4 Effect of Oxidant on Cellobiose Hydrolysis
It was unexpected that the hydrolysis of cellobiose was
significantly enhanced by air compared to Ar or H under the
2
same conditions. Under oxidizing conditions, CO and/or
CO would be expected. However, we found that the highest
2
selectivity to glucose was 94.5%, while the total selectivity
to the byproducts CO , HMF, and furfural was less than 5%
2
with the addition of air. To study the effect of the oxidation
strength on the reaction, three concentrations of hydrogen
peroxide solution were applied: 1, 3, and 10%. As shown in
Fig. 3, all three hydrogen peroxide conditions gave very
high cellobiose conversion ([90%). However, the selectiv-
ity to glucose decreased dramatically with concentration,
resulting in only 17% selectivity for the 10% H O condi-
2
2
tion. Char formation was severe when hydrogen peroxide
was used. At 1% H O , after 5 h reaction, 4% CO was
2
2
2
Even though the glucose selectivity is lower under
oxidizing condition, this finding is encouraging because no
other researchers have reported that oxidation can break the
glyclosidic bond in cellulose with the product stopping at
glucose without further decomposition. We propose that
under high pressure (100–1200 psi), more dissolved oxy-
present in the gas product, which was only attributed to less
than 1% of the selectivity from cellobiose. Upon heating,
hydrogen peroxide will decompose to oxygen and water.
Thus, the partial pressure of oxygen was varied by using
different hydrogen peroxide concentrations. We estimated
there would be 44, 70, and 89% O in the system (as gas
2
0
gen in solution weakens the 1,4 bond (Fig. 1), which
phase) if using 1, 3, and 10% H O , respectively, for com-
2
2
fosters hydrolysis without the need for a strong acidic
solution. Haskins and Hogsed [27] proposed that under
hydrogen peroxide/sodium hydroxide (strongly oxidizing
system), the hydroxyl group on position 2 or 3 may
plete decomposition of H O . However, this was not the case
2 2
100
100
80
60
40
20
0
undergo oxidation and then facilitate the 1,4 ether bond
cleavage. However, this mechanism should produce
ketones or derivative aldol dehydration products, which we
do not see under our conditions. Therefore, the –C2 or –C3
position activation mechanism is not valid in our mild
oxidation condition when using 100 psi air.
8
6
4
0
0
0
Lojewska et al. [28] used in situ Fourier transform
infrared (FTIR) spectroscopy to study the accelerated aging
of paper (cellulose) under a humid air atmosphere. They
found that the oxygen-rich condition gave rise to an intense
growth of carbonyl vibration at 250 °C, with two broad
maxima appearing in the FTIR spectrum at 1730 and
20
0
0
2
Ar
Air
1%H
2
O
2
3%H
2
O
2
10%H
2
O
-
1
1
620 cm . Both frequencies are from C=O stretching.
Oxidants
The former is due to the aldehyde/carbonyl vibration while
the latter is due to the conjugated carbonyl groups vibra-
tion. At this temperature, the reaction ran partly through
pyrolysis forming the gaseous products CO and CO2.
Fig. 3 Cellobiose hydrolysis reaction in various oxidants. Reaction
conditions: temperature, 150 °C; total pressure, 100 psi; reaction
time, 5 h; 30 g of 10 wt% cellobiose/H O solution. Squares cellobi-
2
ose conversion, triangles glucose selectivity
1
23

Catalytic Oxidative Hydrolysis of Disaccharides
503
since the total pressure in the system did not change much
during the reaction time. Other than the severe char forma-
tion, the lower glucose selectivity at H O conditions may
100
100
80
2
2
8
6
0
0
also be due to the formation of low-molecular-weight
organic acids from glucose. This hypothesis was corrobo-
rated by the pH measurement of the product solution after
the reactions using hydrogen peroxide. Typically, the final
solution became more acidic, with a pH of around 2–3
60
40
40
(
initial pH was 5.5 for D.I. water). Analysis by liquid
chromatography-mass spectrometry (LC-MS) is ongoing to
further identify and quantify the product selectivity.
Illman [29] detected a small amount of H O during
20
20
2
2
wood degradation by brown-rot fungi. Infrared spectros-
copy and carboxyl determinations with methylene blue
showed that carboxyl groups were present in the degraded
cellulose. Thus, H O is related to carboxyl compound
0
0
no
γ-Al
O
2 3
Pd
Pt
Au
V
Catalyst
2
2
2 3
Fig. 4 Comparison of various c-Al O supported catalysts in cello-
biose hydrolysis reaction under inert gas. Reaction conditions:
temperature, 150 °C; total pressure, 1200 psi Ar; 30 g of 10 wt%
production. This is consistent with our findings: the solu-
tion changed to more acidic after reaction when using
2
cellobiose/H O solution; reaction time, 8 h. Squares cellobiose
various H O solutions. Also, the H O did not decompose
2
2
2
2
conversion, triangles glucose selectivity
completely to water and oxygen since the head pressure of
the reactor did not change significantly. The H O in
2
2
1
00
100
80
60
40
20
0
solution may attack the primary and/or secondary hydroxyl
to form –C=O or –COOH. It can also cleavage the ether
linkage within the two glucose ring and result in oxidation
to form acids. Separation and identification of the carboxyl
group in the products are underway using LC-MS.
8
0
60
3
.5 Effect of Catalyst and Support on Cellobiose
Hydrolysis
4
0
0
0
2
Yan et al. [19] recently reported hydrolysis of cellobiose by
heterogeneous catalysts, in which Pt-, Pd-, and Ru-supported
catalysts were compared under 580 psi (40 bar) of H . They
2
no
γ-Al O
Pd
Pt
Au
V
2
3
found that a Ru catalyst selectively converted cellobiose to
Catalyst
sorbitol. In our study, we looked at the selectivity of c-Al O
2
3
supported Pd, Pt, Au, and VO catalysts in the reaction from
x
2 3
Fig. 5 Comparison of various c-Al O supported catalysts in cello-
biose hydrolysis reaction with air assistance. Reaction conditions:
temperature, 150 °C; total pressure, 100 psi Air; 30 g of 10 wt%
cellobiose to glucose under an inert condition (Fig. 4) and
under an oxidizing condition (Fig. 5). A c-Al O support was
2
3
2
cellobiose/H O solution; reaction time, 8 h. Squares cellobiose
also included for comparison.
conversion, triangles glucose selectivity
In Fig. 4, Ar was used as the gas medium, and the pres-
sure was set at 1200 psi. The reactions were conducted over
The cellobiose conversions (all [90% except for the gold
catalyst) for all the catalysts were similar to the reaction
without catalyst, which was promoted by air. Similar to the
results in the inert gas medium (Fig. 4), the glucose selec-
tivity was lower over c-Al O supported Pd, Pt, VO , and Au
8
h at 150 °C. The catalyst activity was ranked from low to
high in the order of c-Al O \ Pd \ Pt \ Au \ VO . The
2
3
x
VOx catalyst completely converted all the cellobiose.
However, the selectivity to the desired product, glucose,
over these catalysts showed the opposite trend. In fact, the
color of the residue inside the reactor after the reaction
indicated the degree of severeness of char formation. In both
2
3
x
catalysts than that with no catalyst. The CO gas formed in
2
the reaction was quantified as *10% in the gas phase,
equivalent to 0.6% selectivity from cellobiose decomposi-
tion, suggesting that significant char was the major contri-
bution of the cellobiose conversion.
the Au and VO systems, the char color was pure black,
x
while in the Pt and Pd cases the color was lighter.
The catalyst performance in cellobiose hydrolysis at
The H -TPR data show that the amount of surface
2
150 °C under the oxidizing condition (air) is shown in
Fig. 5. The pressure was set at 100 psi at room temperature.
oxygen on the fresh c-Al O supported catalysts decreases
2
3
in the order VO [ Au [ Pt [ Pd ꢃ c-Al O , which is
x
2 3
123

5
04
W. Deng et al.
the same trend as the catalyst activity shown in Fig. 4.
Thus, the available oxygen species on the catalyst surface
dictates the activity in the reaction. With the same ratio-
nale, the selectivity to glucose followed the reverse order
60
40
oxidative hydrolysis
direct hydrolysis
(
Figs. 4 and 5) because the higher that the oxidation
strength of the catalyst is, the easier it can further oxidize
glucose to other products. The catalyst/Ar system had
lower glucose selectivity than the catalyst/air system due to
more char formation. This observation is corroborated by
2
0
0
the H -TPR results for the used catalysts. After the reaction
2
in Ar atmosphere, more m/z = 15 (from CH ), m/z = 27
4
γ-Al O
Pt/γ-Al O
sp-Al O
Pt/sp-Al O
3
2
3
2
3
2
3
2
(
from C H ), m/z = 43 (from C H ) species eluted when
2 4 3 8
Fig. 6 Cellobiose conversion comparison over various catalysts at
20 °C. Reaction condition: 10% cellobiose/H O. Oxidative hydro-
lysis: 100 psi air; direct hydrolysis: 100 psi Ar
compared to the ones after air reaction. Upon heating in
hydrogen, char will release the above-mentioned mass
species.
1
2
Evolution of methane during the H -TPR experiment can
2
reaction temperature was lower, 120 °C, the thermal deg-
radation of cellobiose was quenched with no conversion. For
the Pt catalyst prepared by the conventional impregnation
give a qualitative indication of the amount of carbon
deposited on the catalyst samples after reaction. Results of
methane evolution from samples used for the conversion of
cellobiose showed a greater amount of carbon deposition
when the reaction was carried out under argon as opposed to
method on porous supports, such as c-Al O , the reaction
2 3
rates are low, and the selectivity of the model compound
cellobiose to glucose is low, as shown in Fig. 6. It was
speculated that cellobiose is too large to enter the catalyst
pores, leading to poor contact with the catalytic surface. Our
computational studies indicated that the critical size of
cellobiose in aqueous solution is at least 12 9 7 9 10
angstroms. In addition, under liquid-phase reaction condi-
tions, the catalyst pores are likely filled with water. Diffu-
sion of the cellobiose, therefore, to the active site is slow,
giving low conversion. In addition, diffusion of the smaller
glucose product is much faster. Thus, the products diffuse to
air; hence, it is clear that O plays a major role in preventing
2
the formation of heavy carbon deposits on the surface. The
evolution of methane also shows that, for the samples tested
under argon, the amount of carbon on the surface follows the
order Pd \ Pt \ Au \ VO (Table 2). This trend is similar
x
to that of the activity of these catalysts during the conversion
of cellobiose, suggesting that the increase of activity of these
metals is likely the result of a greater production of heavy
carbon-containing species (char) on the surface of the cat-
alyst, and that the low selectivity is, in part, due to the for-
mation of char on the catalyst.
the catalyst interior, where they are over-oxidized to CO
significantly lowering the selectivity.
,
2
The reaction temperature for the tests in Figs. 4 and 5
is 150 °C, at which the thermal degradation of cellobiose is
already significant. With the catalysts, dehydration is
enhanced leading to substantial char formation. When the
To address the diffusion limitation issue arising from the
use of porous materials, Pt catalyst was prepared on a non-
porous support, spherical alumina (sp-Al O , a phase), by
2 3
ALD. This technique produces very small and uniform Pt
nanoparticles, in the range of 1–2 nm, as characterized by
high resolution transmission electron microscopy (HRTEM)
Table 2 Methane evolution from used c-Al
determined by H -TPR
2 3
O supported catalysts
[
30], consistent with the CO chemisorption measurement
2
that was used to estimate the metal dispersion. Figure 6
shows that Pt on sp-Al O has four times higher activity than
Catalyst
Reaction
Methane released Temperature of
2
3
atmosphere (Torr s/g-catalyst) maximum peak (°C)
on c-Al O under oxidative hydrolysis, indicating the
2
3
-
5
intrinsic activity of Pt in this reaction by eliminating the
diffusion limitation. Furthermore, without the assistance of
air, the direct hydrolysis of cellobiose by Pt/sp-Al O is
c-Al
VO
Au/c-Al
Pt/c-Al
Pd/c-Al
c-Al
VO
Au/c-Al
Pt/c-Al
Pd/c-Al
2
O
3
Air
Air
Air
Air
Air
Ar
9.14 9 10
397, 631
653
-4
-5
-5
-5
-5
-4
-4
-5
-5
x
/c-Al
2
O
3
1.1 9 10
3.0 9 10
3.4 9 10
5.4 9 10
7.8 9 10
2.0 9 10
1.8 9 10
4.7 9 10
3.9 9 10
2
3
2
O
3
483
similar to that of the support sp-Al O itself, implying that
2
3
2
O
3
494
this reaction is not catalyzed by Pt.
2
O
3
421, 596
579
2 3
O
3.6 Sucrose Conversion by Oxidative Hydrolysis
x
/c-Al
O
2 3
Ar
587
2
O
3
Ar
581
The oxidative hydrolysis rate is increased and the overall
yield of glucose is dramatically improved for oxidative
hydrolysis of cellobiose to glucose at 120 °C (*50%
2
O
3
Ar
520
2
O
3
Ar
426, 585
1
23

Catalytic Oxidative Hydrolysis of Disaccharides
505
1
00
and the understanding of the chemical steps and role of the
catalyst.
oxidative hydrolysis
direct hydrolysis
8
6
4
2
0
0
0
0
0
4
Conclusions
Oxidizing conditions are preferable to inert or reducing
conditions for the hydrolysis of cellobiose to glucose at
1
50 °C and 100 psi pressure. Air-assisted hydrolysis can
0
achieve over 70% glucose yield by weakening the 1,4
glycosidic bond. The amount of surface oxygen on the
catalyst dictates its activity in the reaction, and higher
catalyst oxidation strength leads to further decomposition
of glucose. The use of a non-porous support enables the
study of the intrinsic activity of Pt and Au catalysts in the
hydrolysis of disaccharide sugars, cellobiose and sucrose.
The significant activity of these catalysts under oxidative
hydrolysis conditions indicates a new route for cellulosic
material breakdown.
γ
-Al
2
O
3
Pt/γ
2
-Al O
3
sp-Al
2
O
3
2 3
Pt/sp-Al O
Fig. 7 Sucrose conversion comparison over various catalysts at
8
0 °C. Reaction conditions: 10% sucrose/H
2
O. Oxidative hydrolysis:
2 3
and Pt/c-Al O
1
both have zero conversion
00 psi air; direct hydrolysis: 100 psi Ar. c-Al
O
2 3
conversion, *50% selectivity). Similar improvements are
obtained for sucrose, another model compound for disac-
charide. Sucrose is an isomer of cellobiose and is com-
posed of one glucose unit and one fructose unit linked by a
glycosidic bond. The asymmetric structure of sucrose
Acknowledgment This work is supported by U.S. Department of
Energy, Office of Basic Energy Sciences, Division of Chemical
Sciences, Geosciences and Biosciences under contract FWP 57703.
Argonne is operated by UChicago Argonne, LLC, for the U.S
Department of Energy under contract DE-AC02-06CH11357.
(
5-member ring of glucose and 6-membrane ring of fruc-
tose) could promote a more facile degradation compared to
cellobiose. In Fig. 7, at 80 °C, Pt/c-Al O does not show any
activity for either direct hydrolysis or oxidative hydrolysis.
2
3
However, for the non-porous catalyst Pt/sp-Al O , the
3
2
conversion of sucrose under the oxidative condition reached
1%, with the selectivity of 92% to target products, glucose
References
9
and fructose.
1. Energy Information Administration (2009) Annual energy out-
look. Energy Information Administration, Washington
Gold supported on spherical alumina (a phase) yielded
similar results for the oxidative hydrolysis reaction for
sucrose at 80 °C. With the combination of air and catalyst,
the conversion reached 74% with a 98% selectivity to the
target monosaccharide sugar products, as shown in
Table 3. Some control experiments were conducted to
further look into the reaction (Table 3). First, replacing air
by argon while other conditions are the same no activity
was observed. Second, air without the Au/sp-Al O cata-
2
. Hamelinck CN, Faaij APC, den Uil H, Boerrigter H (2004)
Energy 29:11
3
. Ragauskas AJ, Williams CK, Davison BH, Britovsek G, Cairney
J, Eckert CA, Frederick WJ Jr, Hallett JP, Leak DJ, Liotta CL,
Mielenz JR, Murphy R, Templer R, Tschaplinski T (2006)
Science 311:5760
4
5
. Huber GW, Iborra S, Corma A (2006) Chem Rev 106:9
. Mok WS, Antal MJ, Varhegyi G (1992) Ind Eng Chem Res 31:1
6. Suganuma S, Nakajima K, Kitano M, Yamaguchi D, Kato H,
Hayashi S, Hara M (2008) J Am Chem Soc 130:38
2
3
7
8
9
. Onda A, Ochi T, Yanagisawa K (2008) Green Chem 10:10
. Onda A, Ochi T, Yanagisawa K (2009) Top Catal 52:6
. Yoshida T, Oshima Y (2004) Ind Eng Chem Res 43:15
lyst did not break down any sucrose. Further study of the
oxidative hydrolysis mechanism is underway. It is clear
that catalytic oxidative hydrolysis is potentially an attrac-
tive new route for the decomposition of cellulosic materials
10. Carlson TR, Vispute TP, Huber GW (2008) Chem Sus Chem 5:5
11. Mascal M, Nikitin EB (2008) Chem Int Ed 47:41
2. Dhepe PL, Fukuoka A (2007) Catal Surv Asia 11:4
3. Fukuoka A, Dhepe PL (2006) Angew Chem Int Ed 45:31
4. Ji N, Zhang T, Zheng M, Wang A, Wang H, Wang X, Shu Y,
Stottlemyer AL, Chen JG (2009) Catal Today 147:2
5. Huber GW, Chheda JN, Barrett CJ, Dumesic JA (2005) Science
3
1
1
1
Table 3 Sucrose conversion at 80 °C by Au/sp-Al
O
2 3
catalyst
1
1
1
Catalyst
Gas medium, Sucrose Selectivity to glucose
00 psi conversion (%) and fructose (%)
08:5727
1
6. Kunkes EL, Simonetti DA, West RM, Serrano-Ruiz JC, Gartner
CA, Dumesic JA (2008) Science 322:5900
7. Schutt BD, Abraham MA (2004) Chem Eng J 103:1–3
Au/sp-Al
Au/sp-Al
None
2
O
O
3
3
Air
Ar
74.2
0
98.3
0
2
18. Park J, Park S (2002) Korean J Chem Eng 19:6
1
9. Yan N, Zhao C, Luo C, Dyson PJ, Liu H, Kou Y (2006) J Am
Chem Soc 128:27
Air
0
0
Conditions temperature, 80 °C; total pressure, 100 psi; 30 g of
0 wt% sucrose/H O solution
20. Bootsma JA, Shanks BH (2007) Appl Catal A Gen 327:1
21. Kruse A, Gawlik A (2002) Ind Eng Chem Eng 42:2
1
2
123

5
06
W. Deng et al.
0
2
2
2. Patrick TA, Abraham MA (2000) Environ Sci Technol 34:16
3. Miller JT, Kropf AJ, Zha Y, Regalbuto JR, Delannoy L, Louis C,
Bus E, Bokhoven JA (2006) J Catal 240
4. Libera JA, Elam JW, Pellin MJ (2008) Thin Solid Films 516:18
5. Aaltonen T, Ritala M, Sajavaara T, Keinonen J, Leskela M
28. Lojewska J, Mis kowiec P, Lojewska T, Proniewicz LM (2005)
Polym Degrad Stab 88:512
29. Illman BL (1991) Oxidative degradation of wood by brown-rot
fungi. An American Society of Plant Physiologists Series 6.
American Society of Plant Physiologists, Rockville, MD
30. Setthapun W, Williams WD, Kim SM, Feng H, Elam JW,
Rabuffetti FA, Poeppelmeier KR, Stair PC, Stach EA, Ribeiro FH,
Miller JT, Marshall CL (2010) J Phys Chem 114:21
2
2
(2003) Chem Mater 15:9
2
2
6. Elam JW, Groner MD, George SM (2002) Rev Sci Instrum 73:8
7. Haskins JF, Hogsed MJ (1950) J Org Chem 15:6
1
23