Synergistic effect between defect sites and functional groups on the hydrolysis of cellulose over activated carbon

Article abstract of DOI:10.1002/cssc.201402928

The chemical oxidation of activated carbon by H₂O₂ and H₂SO₄ is investigated, structural and chemical modifications are characterized, and the materials are used as catalysts for the hydrolysis of cellulose. Treatment with H₂O₂ enlarges the pore size and imparts functional groups such as phenols, lactones, and carboxylic acids. H₂SO₄ treatment targets the edges of carbon sheets primarily, and this effect is more pronounced with a higher temperature. Adsorption isotherms demonstrate that the adsorption of oligomers on functionalized carbon is dominated by van der Waals forces. The materials treated chemically are active for the hydrolysis of cellulose despite the relative weakness of most of their acid sites. It is proposed that a synergistic effect between defect sites and functional groups enhances the activity by inducing a conformational change in the glucan chains if they are adsorbed at defect sites. This activates the glycosidic bonds for hydrolysis by in-plane functional groups.

Full text of DOI:10.1002/cssc.201402928

CHEMSUSCHEM
FULL PAPERS
DOI: 10.1002/cssc.201402928
Synergistic Effect between Defect Sites and Functional
Groups on the Hydrolysis of Cellulose over Activated
Carbon
[a]
Guo Shiou Foo and Carsten Sievers*
The chemical oxidation of activated carbon by H O and H SO
4
dominated by van der Waals forces. The materials treated
chemically are active for the hydrolysis of cellulose despite the
relative weakness of most of their acid sites. It is proposed that
a synergistic effect between defect sites and functional groups
enhances the activity by inducing a conformational change in
the glucan chains if they are adsorbed at defect sites. This acti-
vates the glycosidic bonds for hydrolysis by in-plane functional
groups.
2
2
2
is investigated, structural and chemical modifications are char-
acterized, and the materials are used as catalysts for the hy-
drolysis of cellulose. Treatment with H O enlarges the pore
2
2
size and imparts functional groups such as phenols, lactones,
and carboxylic acids. H SO4 treatment targets the edges of
2
carbon sheets primarily, and this effect is more pronounced
with a higher temperature. Adsorption isotherms demonstrate
that the adsorption of oligomers on functionalized carbon is
Introduction
Cellulose, a linear polysaccharide of glucose monomers linked
with carboxylic acid and phenolic groups was not tested for its
catalytic activity.
by b-1,4-glycosidic bonds, is an abundant constituent in most
[
1]
types of biomass. The hydrolysis of cellulose to glucose has
been of great interest as the latter is a versatile intermediate
Kobayashi et al. showed that unacidified carbon with
oxygen functional groups also displays catalytic activity in the
[
2]
[7]
for sustainable chemicals and fuels. Mineral acids are inex-
hydrolysis of cellulose. It was further shown that the catalytic
[
3]
pensive and display high catalytic activity, but major draw-
backs are related to product separation, solvent recycling, and
equipment corrosion. Solid acid catalysts offer simplified sepa-
ration and recycling. Of all the types of solid acid catalysts, car-
bonaceous solid acids and amorphous carbon seem to be es-
activity decreases with the specific amount of functional
groups. The hydrolysis of miscanthus xylan was also effective
using mesoporous carbon nanoparticles functionalized with
[8]
weak acid sites. It was argued that strong acid sites such as
sulfonic groups could leach through ion exchange in the pres-
ence of salts, thus weak acid sites are preferable. However, the
mechanism for the hydrolysis of carbohydrates using weak
acid sites has not been elucidated fully.
[
4]
pecially effective. Particularly, it has been reported that
carbon imparted with sulfonic groups acts as an active, stable,
and reusable heterogeneous acid catalyst. Specifically, sulfonat-
[5]
ed carbon-based acids with a high surface area are preferred.
Katz and co-workers grafted b-glucan strands on silica and
alumina and found that weakly acidic OH groups with the
greatest interactions are able to hydrolyze the glycosidic
Many studies focused on the catalytic performance of
carbon catalysts, but changes in carbon structure after acid
treatment, the role of different functional groups, and mecha-
nistic information remain to be clarified. Besides the use of sul-
fonated carbon, experiments for the hydrolysis of cellulose
have been performed with different solid acids such as niobic
[9]
bond. They also demonstrated that glucan chains can adsorb
rapidly onto the surface of mesoporous carbon nanoparticles
[10]
within the confined pore space. On the basis of these re-
sults, it was concluded that the force exerted on the constrain-
ed polymer by the pore walls and OH groups leads to the hy-
drolysis of the glycosidic bond, which relieves some of the me-
[
4c,d,6]
acid, Nafion, and Amberlyst-15.
It was reported that the
hydrogen bonds between the phenolic OH groups in carbon
and the glycosidic bonds in b-1,4-glucan led to a higher cata-
lytic activity. However, the catalysts were compared with differ-
ent substrates. Furthermore, the unsulfonated carbon material
[9c]
chanical strain. However, no study has been undertaken to
correlate the changes in carbon structure and functional
groups as a result of chemical treatment to reactivity.
Here, we report the effects of chemical treatments of activat-
ed carbon (AC) with hydrogen peroxide (ACOH) and sulfuric
acid at 100 and 2008C (SAC100 and SAC200) as well as a treat-
ment with hydrogen peroxide followed by sulfuric acid at
[
a] G. S. Foo, Prof. C. Sievers
School of Chemical&Biomolecular Engineering
Georgia Institute of Technology
1008C (SACOH). Different types and concentrations of acid
3
11 Ferst Dr., NW, Atlanta, GA, 30332 (USA)
sites are imparted by each of these treatments. The effect of
the functional groups was determined by adsorption isotherms
and the hydrolysis of cellulose. Finally, we examine the origin
E-mail: carsten.sievers@chbe.gatech.edu
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201402928.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&1
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
of the catalytic activity, which can be linked to the synergistic
effect of functional groups and the immobilization of glucan
chains on edge/defect sites.
Table 1. Degree of graphitization of carbon catalysts.
Catalyst
G
ID1/(I +ID1+ID2)
AC
0.745
0.746
0.733
0.729
0.715
ACOH
SACOH
SAC100
SAC200
Results
Characterization
The Raman spectra of all carbon catalysts contained two com-
À1
plex peaks at n˜ =1352 and 1598 cm (Figure 1). These are as-
[12a]
is related to the abundance of edges.
Consequently, it is
suggested that H SO is able to break down the stacks of gra-
2
4
phene sheets and to create step sites at the edge or cavities
and that this effect is more pronounced at higher tempera-
À1
tures. The absence of a broad peak at n˜ =500 cm indicates
3
that amorphous sp carbon is not present to a significant
[14]
extent (Figure S1).
All of the XRD patterns of the carbon catalysts exhibited two
broad peaks at 2q=24 and 428 (Figure S2), which are attribut-
ed to the (002) and (100) or (101) planes of graphite crystalli-
[15]
tes, respectively. This shows that the chemical treatments
with H O2 or H SO4 did not change the crystallinity of the
2
2
carbon structure compared to that of AC to a measurable
extent.
13
The C direct-polarization magic-angle spinning (DP MAS)
NMR spectra of the carbon samples are shown in Figure S3.
The intense and broad peak at d=119 ppm is assigned to non-
Figure 1. Raman spectrum and results from fitting for AC.
[16]
protonated polyaromatic carbon in the core structure. The
network of activated carbon is highly complex. As such, the
broad peak also contains contributions from exterior carbon
species that have a chemical shift that ranges between d=110
sociated with the vibrational modes of disordered and ordered
[
11]
graphene sheets. However, as a result of the presence of var-
ious contributions, the spectra needed to be deconvoluted to
obtain the proper structural information (Figure 1). The two
peaks were fitted with four Lorentzian-shaped bands (D4, D1,
[17]
and 148 ppm. As a result of the large range of polyaromatic
carbon present, the shoulder caused by carboxylic and phenol-
ic carbon species at d=180 and 155 ppm can barely be ob-
À1
G, and D2) at n˜ =1210, 1350, 1575, and 1600 cm , respective-
À1 [12]
3
ly, and one Gaussian-shaped band (D3) at n˜ =1530 cm .
served. In addition, the low intensity of the resonance of sp -
These five components are assigned as polyenes, graphene
edges, graphitic carbon, graphene sheets, and amorphous
carbon, respectively.
hybridized carbon (d=0–40 ppm) suggests that most of the
2
carbon is polyaromatic and sp hybridized. Deconvolution was
attempted, but the resolution of the spectrum was not suffi-
cient to obtain reliable results.
It has been suggested that the ratio I /(I +I +I ) is a mea-
D1
G
D1
D2
sure of the degree of graphitization for carbonaceous materi-
The morphological properties of the carbon catalysts are
given in Table 2. Treatments with H O and H SO had a moder-
[
12b,13]
als, in which I is the integral of the fitted peak.
rameter was averaged over three
The pa-
2
2
2
4
spectra for each sample and is
shown in Table 1. The highest
Table 2. Surface area, concentration of stable acid sites, and catalytic activity of carbon catalysts.
values (i.e., the most organized
structures) were observed for AC
and ACOH, which indicates that
À1
[c]
Catalyst
SBET
[m g
Concentration of acid site [mmolg
]
Conversion
[%]
Glucose
yield [%]
2
À1
[a]
[a]
[b]
[c]
]
phenolic
lactonic
carboxylic+
sulfonic acid
[
a]
sulfonic acid
the treatment of H O2 did not
2
affect the structural integrity of
the sample. Samples treated
distilled water
AC
ACOH
SACOH
SAC100
SAC200
–
613
634
695
595
653
–
–
39
–
58
–
34
–
–
–
31
144
153
–
24
27
26
38
32
34
23
34
0
0
2
10
8
10
10
18
281
233
152
160
–
252
285
115
139
–
244
341
147
163
–
with H SO4 at 1008C (SACOH
2
and SAC100) had a lower degree
of
graphitization,
whereas
3
1
.8 mm acetic acid
.8 mm H SO
SAC200 had the least organized
structure as it was subjected to
a more severe treatment. The
disorder of the graphene sheets
2
4
–
–
–
–
–
[a] Determined by Boehm titration. [b] Determined by elemental analysis (S content). [c] Reaction time of 24 h
at 1508C and autogenic pressure.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
2
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Figure 3. FTIR spectra of a) AC, b) ACOH, c) SAC100, d) SAC200, and
e) SACOH.
À1
of sulfonic acid groups (31 mmolg ). However, there was an
Figure 2. Pore size distribution of carbon catalysts a) 0.45–1.00 nm deter-
mined by CO physisorption and b) 3–6 nm determined by N physisorption.
2 2
increase in the concentration of lactonic and carboxylic acid
groups and a decrease in phenolic groups. Acidification by
H SO at 1008C functionalized the carbon surface mainly with
2
4
phenolic, lactonic, and sulfonic acid groups. At a higher tem-
perature treatment (SAC200), more of these functional groups
are present on the surface. For reference, the concentration of
sulfonic acid groups before hot liquid water treatment was
also quantified (Table S2).
ate effect on the surface area of the carbon catalysts. The pore
volume distribution of the carbon catalysts obtained by both
CO and N physisorption is shown in Figure 2. AC oxidized by
2
2
H O showed a decrease in pore volume in the small-pore size
2
2
range (Figure 2a), whereas there was an increase in pore
volume in the larger-pore size range (Figure 2b). Treatment
with H SO4 after H O2 (SACOH) led to an increase in pore
The pH of the carbon catalysts (300 mg) suspended in
27 mL of deionized (DI) water as well as 3.8 mm acetic acid
and 1.8 mm sulfuric acid are given in Table 3. AC displayed the
highest pH of 8.4 because of the presence of base sites. The
suspension of chemically treated carbon displayed a pH be-
tween 4.4 and 5.0, whereas acetic acid and sulfuric acid had
a pH of 3.7 and 2.8, respectively.
2
2
[21]
volume compared to that of ACOH. In addition, there was an
increase in pore volume for pore sizes around 5 nm. Acidifica-
tion by H SO4 at 1008C (SAC100) did not change the pore
2
volume distribution, whereas acidification at 2008C (SAC200)
decreased the contribution of pores in the larger size range.
The pore surface area of the samples showed similar trends
(Table S1).
Table 3. pH values of slurries of 300 mg of each catalyst in 27 mL of
water and homogeneous acid solutions used as references.
FTIR spectra were measured to identify the types of func-
tional groups present on the carbon samples (Figure 3). All
Catalyst
pH value
À1
samples displayed a band at n˜ =1580 cm that is assigned to
AC
8.0
5.0
4.3
4.5
4.4
3.7
2.8
2
[18]
the in-plane vibration of sp -hybridized carbon (i.e., C=C).
ACOH
SACOH
SAC100
SAC200
À1
The broad features at n˜ =1260 and 1020 cm correspond to
aromatic dCH. For samples treated with H O (Figure 3b and
[
19]
2
2
À1
e), a peak at n˜ =1720 cm indicated the presence of carbonyl
groups (nC=O). For SAC100 and SAC200 (Figure 3c and d),
3
1
.8 mm acetic acid
.8 mm sulfuric acid
À1
a broad shoulder was observed at n˜ =1109 cm , which is as-
signed as the nsymSO for ionic sulfates, whereas the n SO vi-
3
asym
3
bration overlapped with the aromatic dCH vibration at approx-
Hydrolysis of cellulose
À1 [20]
imately n˜ =1230 cm . Although this peak was observed for
SACOH (Figure 3e), it had a low intensity. These results show
that different functional groups were imparted successfully to
the carbon structure by different chemical treatments.
Hydrolysis experiments were performed with cellulose that
was ball-milled for 48 h to reduce its crystallinity (Figure S4).
The resulting material had a slightly lower degree of polymeri-
[22]
For quantitative analysis, only functional groups that were
stable for 24 h in liquid water at 1508C were considered. For
ACOH, chemical oxidation by H O imparted phenolic, lactonic,
zation (Table S3) and part of it was soluble in water. The AC
catalyst gave 27% conversion and 0% glucose yield after 24 h
of reaction (Table 2). ACOH provided a similar conversion but
had a higher glucose yield of 2%. SACOH showed the highest
conversion of 38% and a glucose yield of 10%. SAC100 gave
2
2
and carboxylic acid groups (Table 2). Sequential treatment with
H O and H SO (SACOH) only imparted a small concentration
2
2
2
4
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
3
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
a glucose yield of 8%, whereas SAC200 had a higher reactivity
with a glucose yield of 10%, which is similar to SACOH. The
rest of the products were mainly oligomers, which could not
be quantified with the analytical setup used here.
For comparison, 27 mL of 3.8 mm acetic acid, which has the
same amount of carboxylic acids as SACOH (300 mg), was used
in the hydrolysis of cellulose. The glucose yield (10%) was
comparable to that of SACOH and SAC200. In addition, 1.8 mm
of H SO was used for comparison because it has the same
2
4
concentration of sulfate groups as SAC200. The catalyst exhib-
ited the highest glucose yield of 18%.
The hydrolysis reactions for the carbon catalysts were also
performed over various periods of time (Figure 4). SACOH ex-
Figure 5. Adsorption isotherms of a) glucose and b) cellobiose on AC. Insets:
Linear regression of Langmuir isotherm parameters.
and cellobiose were measured. An example of the adsorption
isotherms on AC is shown in Figure 5. The maximum uptake
values in triplicate experiments were within 10% of each other
for the same adsorbent and adsorbate, and all the coefficient
2
of determination (R ) values were at least 0.95. Adsorption co-
efficients are within 12–48%. The average maximum uptake
Figure 4. a) Conversion and b) glucose yield during cellulose hydrolysis at
1
508C using chemically treated carbon catalysts: (^) ACOH, (&) SACOH, (~)
(q ) was normalized by the BET specific surface area of each
m
SAC100, (*) SAC200.
adsorbent so that comparison between adsorbents can be
made. The values of the parameters are summarized in Table 4.
The adsorption coefficient (K) of glucose on AC was
À1
hibits the highest reactivity followed by SAC200, SAC100, and
ACOH. The conversion of cellulose and the glucose yield in-
creased significantly over the first 16 h. For the next 8 h, the
glucose yield remained approximately the same, and the in-
crease in conversion was not substantial. These trends are con-
sistent with reports in the literature, which indicated
37 Lmol . If cellobiose was adsorbed on AC, K increased sig-
À1
nificantly to 650 Lmol . However, if the bare carbon material
was functionalized, K decreased notably. For the adsorption of
glucose, K decreased by approximately 60% for the functional-
ized materials compared to AC. For the adsorption of cello-
that a plateau is reached for the hydrolysis of cellu-
[
4a,c,d]
lose using acidified carbon catalysts.
This sug-
Table 4. Langmuir adsorption constant and surface specific maximum uptake for glu-
cose and cellobiose on activated carbon catalysts and change in standard free energy
of adsorption between cellobiose and glucose.
gests that the remaining cellulose could have a recal-
citrant structure that is resistant to hydrolysis by acti-
[
4a]
vated carbon.
In addition, reactions with acetic
Adsorbent
Glucose
Cellobiose
DDG
[kcalmol ]
[
a]
[b]
[a]
[b]
À1
acid and sulfuric acid did not show a net degradation
of glucose after 24 h (Figure S5).
normalized q
K
normalized q
K
m
m
À2
À1
À2
À1
[mgm
]
[Lmol
]
[mgm
]
[Lmol ]
AC
0.135
0.178
0.173
0.192
0.201
37
13
14
11
16
0.171
0.187
0.187
0.200
0.200
650
379
437
393
402
À1.78
À1.94
À2.03
À2.11
À2.13
ACOH
SACOH
SAC100
SAC200
Adsorption isotherms
To understand how different functional groups on
the same carbon substrate influence the adsorption
of glucan molecules, adsorption isotherms of glucose
[
a] Maximum uptake. [b] Adsorption coefficient.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
4
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
biose, K also decreased but only by approximately 40%. The
change in the standard free energy of adsorption (DDG) be-
tween the adsorption of cellobiose and glucose was also calcu-
lated for each adsorbent (see Supporting Information and
terial (ACOH) in the small-pore size range, but the pore
volume in the larger-pore size range increased (Figure 2). This
suggests that oxidation by H O enlarges at least some of the
2
2
pores of the activated carbon. Generally, two routes are possi-
ble: an attack on existing defect sites and electrophilic addi-
À1
Table 4). A negative DDG value of À1.78 kcalmol was ob-
[27]
tained for AC. For the functionalized carbon materials, the
tion on unsaturated C=C bonds. However, Raman spectros-
copy results (Table 1) showed that the degree of graphitization
for ACOH remains the same as that of AC. This implies that the
ratio of graphene edges to graphitic carbon and graphene
sheets is largely the same, which suggests that the material
that is removed or redistributed could be amorphous in
nature.
value of DDG decreased to between À1.94 and À2.13 kcal
À1
mol . The normalized q of AC for the adsorption of glucose
m
À2
À2
is 0.135 mgm . This increased to around 0.17 mgm
for
À2
ACOH and SACOH, and 0.20 mgm for SAC100 and SAC200.
However, the normalized qm for the adsorption of cellobiose
did not increase significantly relative to that of AC. The frac-
tional increase was only 9–17%.
In the case of SACOH, acidification with H SO after H O
2
2
4
2
treatment increased the pore volume for existing pore sizes
relative to ACOH (Figure 2), whereas the degree of graphitiza-
tion decreased (Table 1). SAC100 had a similar pore size distri-
bution to AC, but the degree of graphitization was lower than
that of AC, which indicates that acidification with H SO at
Discussion
Effects of chemical oxidation on activated carbon
2
4
Chemical oxidation is a common approach to tailor the func-
1008C creates more edges in the carbon structure. This effect
is more pronounced at a high-temperature treatment as indi-
cated by the observation of the lowest degree of graphitiza-
tion for SAC200. This observation is in agreement with a previ-
ous study that showed that intercalated sulfuric acid can exfo-
liate the structure and attack the active sites of carbons at
[
21]
tionality of activated carbon catalysts. Methods to modify ac-
tivated carbon include acid, base, microwave, ozone, and
[
23]
plasma treatments. These methods are usually used to en-
hance the adsorption of organic and inorganic pollutants. It is
commonly known that oxidative modifications lead to decreas-
[
23,24]
[28]
es in surface area and pore volume.
Most studies in the lit-
temperatures higher than 1008C.
13
erature only focus on the quantification of the number of func-
tional groups on carbon catalysts and its correlation to reactivi-
Results from C DP MAS NMR spectroscopy showed that
a large range of polyaromatic carbon species was present in
the materials and that most of the carbon in the structure is
[
4c–e,25]
ty.
In contrast, the effect of chemical treatment on the
2
structure of carbon has not been studied extensively.
sp hybridized. In addition, the absence of a peak between n˜ =
À1
In the present study, FTIR spectroscopy and Boehm titration
results demonstrated that oxidation by H O functionalized the
400–500 cm in the Raman spectra of all the samples (Fig-
3
ure S1) also indicated that amorphous sp -hybridized carbon is
2
2
[14]
surface with phenolic, lactonic, and carboxylic acid groups for
ACOH, whereas treatment with H SO imparts phenolic, lacton-
not present in significant amounts. This suggests that the
functional groups that are imparted would have to be posi-
tioned around the edges and defect sites of the carbon struc-
2
4
ic, and sulfonic acid groups (SAC100). Acidification at a high
temperature of 2008C resulted in the presence of more func-
tional groups. In the case of SACOH, only a small concentration
[18,26,29]
ture, and in the same plane as the graphene sheet.
Spe-
cifically, the functional groups are not positioned perpendicu-
larly to the surface, which limits their accessibility for glucan
chains. A simple schematic representation of the structures
formed in the chemical oxidation process that can occur at the
edge or existing defect sites of a graphene sheet is shown in
Figure 6. An aromatic or unsaturated C=C bond can be at-
À1
of hydrothermally stable sulfonic groups (31 mmolg ) was re-
tained on the surface after treatment in liquid water at 1508C.
À1
Prior to the elution step, 150 mmolg of sulfonic acid groups
was present on the sample (Table S2). In agreement with previ-
ous studies, we propose that the treatment with H O results
2
2
[26]
in oxidation of sites with unsaturated and dangling bonds ter-
tacked by a reagent to produce a hydroxyl group. This can
be oxidized to form a quinone group or broken up to form
a carboxylic acid group in the presence of a strong oxidizing
agent. In close proximity, a hydroxyl group and a carboxyl or
[
15b,26]
minated with CH or CH groups.
It is suggested that this
2
is followed by chemical reactions at the hexatomic–hexatomic
boundaries to generate new defects after the initial points of
attack are functionalized. This suggests that the second chemi-
cal treatment with H SO results in sulfonic groups that occupy
[29]
carbonyl group can condense to form a lactone.
2
4
newly generated sites, which are easier to elude. It is speculat-
ed that the functional groups present from the first treatment
could create a reactive environment on the carbon surface
that are converted to porous domains during the acidification
with H SO , which results in a slightly higher surface area for
Adsorption properties and locality of functional groups
In studies on the hydrolysis of cellulose using functionalized
amorphous carbon, it has been claimed that phenolic groups
on carbon form hydrogen bonds with the oxygen atoms in
glucan chains, whereas carboxylic and sulfonic acid groups do
2
4
SACOH.
[4c,d]
XRD results showed no noticeable effect of treatments with
H O and H SO on the limited crystallinity of the activated
not function as adsorption sites.
However, this conclusion
was based on the comparison between different types of cata-
lysts, such as Nafion, niobic acid, and Amberlyst-15, and these
materials also differ in other properties.
2
2
2
4
carbon. However, the pore size distribution indicated that oxi-
dation by H O decreased the pore volume of the carbon ma-
2
2
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
5
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
The preference for the adsorption of oligomers can be fur-
ther illustrated by a comparison of the change in the standard
free energy of adsorption between cellobiose and glucose
À1
(
DDG). The value for AC was calculated to be À1.78 kcalmol ,
À1
which is close to the value of À1.4 kcalmol obtained by
[10]
Chung et al. for mesoporous carbon nanoparticles. For the
functionalized materials, the values of DDG decreased further
À1
by up to 0.3 kcalmol , which indicates an increase of the pref-
erence for the adsorption of cellobiose. As this decrease in the
free energy of adsorption is small compared to the strength of
[32]
a hydrogen bond, it is suggested that interactions with func-
tional groups are not responsible for the difference in adsorp-
tion between cellobiose and glucose. Besides solvation, DDG
also depends on the degree of contact on the molecular
[10]
level.
We speculate that the defect sites that are formed
during the chemical treatment of AC provide a spatial environ-
ment that favors the adsorption of cellobiose over glucose.
An increased uptake of glucose monomers was observed on
functionalized materials compared to that on AC. One possible
explanation is that glucose molecules are adsorbed in an ori-
entation with a smaller footprint on defect sites. Interestingly,
SACOH has the same normalized adsorption capacity as ACOH
even though it has more functional groups. As the functional
groups are located at the edge or defect sites of the graphene
sheets (vide supra), this suggests that they are localized in cavi-
ties of the carbon structure. Thus, in the case of SACOH, its
cavities or edges would have a higher density of functional
groups compared to those of ACOH, and there is a limit on
the amount of biomass molecules that can be adsorbed in
these cavities because of spatial limitations.
Figure 6. Schematic representation of the structure of functionalized carbon.
In this study, adsorption isotherms revealed that all of the
carbon materials are able to adsorb both glucose and cello-
biose (Table 4). AC is commonly considered to be a good sorb-
ent for biomass molecules because of its large surface area
A recent report by Fukuoka and co-workers demonstrated
that oxygenated functional groups on carbon do not affect the
adsorption capacity of glucans, and there was no difference in
[
30]
and pore volume. The adsorption process is driven by van
[10]
der Waals forces between glucans and the carbon surface.
Specifically, it is the interaction between CH groups and the
[31a]
the adsorption coefficient.
However, the functional groups
[
31]
polyaromatic rings. The adsorption of glucose on AC had an
were removed by heat treatment up to 1273 K, and this
method has been shown to also increase the cross-linking of
carbon sheets, which could result in a more uniform polyaro-
À1
adsorption coefficient of 37 Lmol . However, functionalization
of AC reduced the adsorption coefficient of glucose by approx-
imately 60% compared to that of AC. The large decrease is at-
tributed to competitive adsorption with water, which becomes
more pronounced as the polarity of the surface increases.
The larger number of CH groups in cellobiose resulted in an
[4b,6]
matic surface for adsorption.
Thus, it is challenging to attri-
bute the observed behavior to a single factor.
Mechanism of cellulose hydrolysis
À1
increase of the adsorption coefficient on AC to 650 Lmol .
This is attributed to additional van der Waals interactions with
the carbon surface. In the case of the functionalized materials,
the adsorption coefficients also decreased, but the decrease
was less pronounced (40%) compared to the adsorption of
glucose. This suggests that the stronger van der Waals interac-
tion allows cellobiose to compete more favorably with water
for adsorption sites, even on a relatively polar surface. By ex-
trapolation from the adsorption of glucose and cellobiose, the
adsorption coefficients would also increase with the length of
the glucan chain, which suggests that the functionalization of
the carbon surface has a limited impact on the adsorption of
oligomeric species, whereas the adsorption of monomeric glu-
cose is reduced by competition with water. This could explain
the limited degradation of glucose monomers in the hydrolysis
Reports in the literature have shown that weakly acidic func-
tional groups such as hydroxyl and carboxylic acids on other
types of catalysts (niobic acid and Amberlyst) are not active in
[4c,d]
the hydrolysis of cellulose.
However, if they are positioned
on a carbon surface, these weak functional groups are able to
catalyze the hydrolysis of carbohydrates if the glycosidic bonds
[7–9]
are forced to interact with these groups.
In this study, reac-
tivity results show that only the carbon materials treated
chemically yield glucose (Table 2). As no glucose is formed
over AC, the weak catalytic activity of ACOH can be ascribed
to the addition of the functional groups (i.e., phenolic, lactonic,
and carboxylic acid groups). However, the presence of these
groups alone provides limited reactivity. The higher catalytic
activity of SAC100 and SAC200 could be attributed to the for-
mation of sulfonic acid groups. However, the catalytic activity
[
4a]
of cellulose over sulfonated activated carbon.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
6
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
of SACOH is comparable to that of SAC200 even though it
only has a low concentration of sulfonic groups. In addition,
SACOH had only 14.5% more functional groups than ACOH,
but its glucose yield was approximately five times higher.
The improvement in catalytic activity for SACOH can be cor-
related with the degree of graphitization as obtained by
Raman spectroscopy (Table 1). Only treatment with H SO de-
On the basis of these findings, it is proposed that the ad-
sorption of glucan chains on defect sites or edges is critical for
catalytic activity because a conformational change is induced
in this environment. This enables the in-plane functional
groups to attack the exposed glycosidic bonds. A schematic
representation of the adsorption and hydrolysis process is
shown in Figure 7. The analysis of the adsorption isotherms in-
2
4
creased the degree of graphitization, which decreased more at
a higher treatment temperature. The disorder of the carbon
[
12a]
structure is attributed to abundant edges.
Thus, we suggest
that acidification with H SO creates more steps in the cavities
2
4
or edges of the carbon structure, and these types of defects
have a synergistic effect with the in-plane functional groups
(vide infra).
Kobayashi et al. ball-milled carbon and cellulose together,
and the reactivity of the resulting material was higher than
[7]
that from an experiment without ball-milling them together.
This suggests that intimate interaction between the reactant
and catalyst is a prerequisite for hydrolysis. They also showed
that salicylic acid, which contains adjacent carboxylic and phe-
nolic groups, is more effective in the hydrolysis of cellobiose
than other molecular catalysts such as m- and p-hydroxybenzo-
ic acid, in which the functional groups are further apart. This il-
lustrates the potential synergistic effects of a coordination site
in close proximity to the active site. However, ACOH and
SACOH have a similar concentration of weak acid sites, and
there is no reason to suggest a preferential formation of
paired active sites in SACOH. Furthermore, the pH values of
the aqueous slurry of SACOH and SAC200 were 4.3 and 4.4, re-
spectively, which should not be acidic enough to catalyze the
hydrolysis reaction. The pH values after reaction were within
Figure 7. Schematic representation of the adsorption and hydrolysis mecha-
nism.
0
.1 of the original values, which indicates that the functional
[21]
groups are stable in agreement with a previous report. It
was proposed that a pH lower than 4 is needed for this reac-
dicates that interactions between the CH groups and polycyclic
aromatic rings are responsible for the immobilization of long
glucan chains on defect sites. As the functional groups are lo-
calized, the probability of hydrolysis increases with the density
of functional groups in each defect site. Glucan chains can also
adsorb on polyaromatic domains of the carbon surface. How-
ever, we speculate that these species are much less susceptible
to the hydrolysis reaction because of the lack of functional
groups positioned appropriately.
[
33]
tion. However, the environment within the vicinity of a sul-
fonic acid group should still be acidic enough for the reaction
to occur. Moreover, the pH values of acetic acid and sulfuric
acid used are 3.7 and 2.8, respectively, yet the reactivity of
acetic acid did not exceed that of SACOH and SAC200
(Table 2). The slurries of the other carbon catalysts had
a higher pH because of the existence of base sites and the de-
localization of charge across the basal plane of the carbon sur-
[
34]
face.
In a related report, Pang et al. sulfonated carbon at elevated
temperatures and reported higher catalytic activity even
Yanovsky simulated the adsorption of organic polymer
chains on different carbon surfaces (with and without defects)
and demonstrated that adsorption on a carbon surface with
defects is energetically more favorable and that the chains are
[4e]
though the increase in acid density was not significant. How-
ever, beyond a sulfonation temperature of 2508C, the catalytic
activity decreased slightly while the acid site density continued
to increase. In the present study, it was found that the degree
of graphitization decreases with increasing sulfonation temper-
ature. In addition, the creation of defects and edges in the
carbon structure facilitates the hydrolysis of cellulose by a con-
formation change upon adsorption. We hypothesize that an in-
crease of the sulfonation temperature above 2508C leads to
severe degradation of the carbon structure, which becomes
less useful as the defects created could be too large for any ef-
fective conformation change of the glucan chain to occur.
Chung et al. also reported that there was no proportional cor-
[
35]
immobilized in this environment.
Furthermore, Gazit and
Katz demonstrated that weakly acidic hydroxyl groups on silica
and alumina are able to hydrolyze the b-glycosidic bonds by
[
9c]
grafting glucan strands on the surface. This indicates that
weakly acidic functional groups are able to perform hydrolysis
if they are interacting with a glucan chain in a constraint posi-
tion. This conclusion is in agreement with a report that con-
cludes that conformational changes in the cellulosic chains are
required for the activation of the glycosidic bond for hydroly-
[
36]
sis.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
7
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
relation between phenolic groups and yield for sulfonated and
unsulfonated carbon material in the hydrolysis of miscanthus
batch of ACOH was subjected to the same sulfuric acid treatment
at 1008C, and the product was named SACOH. To prevent elution
[21]
[
8]
of functional groups during reactions, all chemically treated sam-
ples were treated with hot liquid water at 1508C and autogenic
pressure for 24 h, collected by filtration, and washed with DI water
until the filtrate had a neutral pH.
xylan. This further supports the hypothesis that both func-
tional groups and defect sites are needed for the hydrolysis of
glucan chains to occur effectively.
Conclusions
Characterization
The effect of chemical oxidation on activated carbon was in-
vestigated. A simple chemical treatment with H O can enlarge
Raman spectra were collected by using a Confocal Raman Micro-
scope Alpa-Witek with a laser wavelength of 514 nm. Each sample
was spread across a glass slide. Ten scans were accumulated and
three spectra were obtained for each sample at different locations.
Peak fitting was performed with the GRAMS/AI software. XRD
measurements were performed by using a Philips X’pert diffrac-
tometer equipped with an X’celerator module using CuKa radia-
tion. Diffractograms were collected at incident angles from 2q=5
to 708 with a step size of 0.01678. C DP MAS NMR spectroscopic
measurements were performed by using a Bruker DSX 300 spec-
trometer. The samples were packed into a 4 mm zirconia rotor and
spun at 10 kHz. Adamantane was used as an external reference
material, and the low field peak was set as the reference (d=
2
2
pore size and impart hydrophilic functional groups such as
phenols, lactones, and carboxylic acids. Treatment with H SO4
2
decreases the order of graphitization by etching the edges of
the carbon structure, and this effect scales with temperature.
Only phenols, lactones, and sulfonic groups are functionalized
on the carbon material. Extrapolation of the results from ad-
sorption isotherms shows that the functional groups only have
a minor effect on the adsorption of long-chain oligomers,
which is dominated by van der Waals interactions between CH
groups of the sorbate and the surface. The materials treated
chemically are more active in the hydrolysis of cellulose com-
pared to homogenous catalysts even though they possess
only weakly acidic functional groups. This can be explained by
the synergistic effect of functional groups at defect sites and
edges of the carbon surface, on which the glycosidic bonds
are exposed and forced to interact with the in-plane functional
groups.
13
1
13
3
8.45 ppm). The resonance frequencies of H and C were 300.2
13
1
and 75.5 MHz, respectively. For C detection, high power H de-
13
coupling was used during the sampling of the C magnetization.
A short p/2 pulse (5 ms) was applied, and the recycle delay was 4 s.
Approximately 20000 scans were accumulated for each spectrum.
N and CO physisorption measurements were performed by using
2
2
a Micromeritics ASAP 2020 physisorption analyzer at 77 and 273 K,
respectively. Before analysis, approximately 150 mg of each sample
was degassed under vacuum for 4 h at 1508C. For N physisorp-
2
[37]
tion, the surface area was calculated using the BET method, and
the pore volumes were calculated using the Barrett–Joyner–Halen-
Experimental Section
[38]
Materials
da (BJH) method. For CO physisorption, the pore volume was
2
[39]
calculated using the DFT method. FTIR spectra were collected by
Activated charcoal (untreated, granular, 4–8 mesh, made from peat
bog), sulfuric acid (95.0–98.0%), glucose (>99.5%), cellobiose (>
using a Nicolet 8700 FTIR spectrometer with a MCT/A detector. For
À1
each spectrum, 64 scans were recorded at a resolution of 4 cm
.
9
8%), sodium hydroxide (>98%, pellets), potassium hydrogen
Each carbon sample was mixed with KBr at 1 wt%, pressed into
a self-supporting wafer, and loaded into a vacuum transmission
phthalate (>99.95%), sodium carbonate (>99%), sodium bicar-
bonate (>99.7%), potassium bromide (FTIR grade, >99%), acetic
acid (>99.7%), and cellulose (microcrystalline powder) were pur-
chased from Sigma Aldrich. Hydrogen peroxide (30% w/w aqueous
solution) and hydrochloric acid (36.5–38.0% w/w aqueous solution)
were purchased from BDH and Alfa Aesar, respectively. DI water
was further purified by using a Barnstead NANOpure ultrapure
FTIR chamber. The spectra were taken under vacuum (<
À6
1
0
mbar). All carbon samples and cellulose were sent to Atlantic
Microlab for elemental analysis to determine their sulfur and
carbon contents. Boehm titration was performed following proce-
dures reported in the literature to quantify the amount of function-
[21,40]
al groups present.
Briefly, each carbon sample (1.5 g) was
À1
water system to 18.2 MWcm .
added to aqueous solutions of NaOH, NaHCO , and Na CO
3
3
2
(50 mL), respectively. Each base solution had a concentration of
0
.05m. The slurries were shaken for 24 h, filtered, and 10 mL ali-
Synthesis of catalysts
quots of the filtrate were collected. For the NaHCO and NaOH sol-
3
Activated charcoal (2.5 g) was mixed with DI water (30 mL) in
a 45 mL Teflon-lined acid digestion vessel (Parr Instrument) and
loaded into a rotary oven at 2008C for 24 h. The material was col-
lected by filtration and washed with DI water. The resulting sample
utions, HClaq (20 mL, 0.05m) was added. For the Na CO solution,
2
3
HClaq (30 mL, 0.05m) was added. The solutions were purged with
N2 for 2 h and back titrated with a 0.05m solution of NaOH until
an endpoint of pH 7.0 was reached. The pH of each carbon catalyst
(300 mg) was measured in DI water (27 mL) after 1 h of stirring.
was named AC. Subsequently, AC (0.15 g) was added to H O2
2
(
8
(
8.0 g), and the suspension was stirred continuously and heated at
58C for 1 h in a beaker. The product was collected by filtration
Fischer Scientific, Grade Q5,5–10 mm), washed with DI water, and
During stirring, the suspension was degassed with N . The pH of
acetic acid (3.8 mm) and sulfuric acid (1.8 mm) was also measured.
2
The weight-average molecular weight (M ) and number-average
w
named ACOH. To prepare other samples, AC (2.5 g) was mixed
molecular weight (M ) of cellulose and ball-milled cellulose samples
n
with H SO (30 mL, 5m) in a 45 mL Teflon-lined acid digestion
were determined by gel permeation chromatography (GPC). This
was accomplished by derivatization of the cellulose samples using
pyridine and phenyl isocyanate and characterization by using
a Waters GPC system. The weight-average and number-average
degree of polymerization (DP , DP ) of the cellulose were obtained
2
4
vessel (Parr Instrument), loaded into a rotary oven, and heated at
00 and 2008C for 5 h, respectively. The product was collected by
1
filtration, washed with DI water until the filtrate had a neutral pH,
and the samples were named SAC100 and SAC200, respectively. A
w
n
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
8
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
by dividing Mw and M by 519, the molecular weight of the cellu-
[1] G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–4098.
n
[
[
[
2] A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney,
C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak, C. L. Liotta, J. R. Mie-
lenz, R. Murphy, R. Templer, T. Tschaplinski, Science 2006, 311, 484–489.
3] a) E. E. Harris, E. Beglinger, G. J. Hajny, E. C. Sherrard, Ind. Eng.Chem.
lose tricarbanilate monomer. The molecular weight of the derivat-
ized cellulose samples was determined by using a relative calibra-
tion curve. The calibration curve was created by fitting a third poly-
nomial equation to the retention volumes obtained from a series
of narrow-molecular-weight-distribution polystyrene standards. The
1
945, 37, 12–23; b) E. E. Harris, E. Beglinger, Ind. Eng.Chem. 1946, 38,
890–895.
2
curve fit had an R value of 0.9984.
4] a) A. Onda, J. Jpn. Pet. Inst. 2012, 55, 73–86; b) K. Nakajima, M. Hara,
ACS Catal. 2012, 2, 1296–1304; c) M. Kitano, D. Yamaguchi, S. Suganu-
ma, K. Nakajima, H. Kato, S. Hayashi, M. Hara, Langmuir 2009, 25, 5068–
Hydrolysis of cellulose
5075; d) S. Suganuma, K. Nakajima, M. Kitano, D. Yamaguchi, H. Kato, S.
Hayashi, M. Hara, J. Am. Chem. Soc. 2008, 130, 12787–12793; e) J. F.
Pang, A. Q. Wang, M. Y. Zheng, T. Zhang, Chem. Commun. 2010, 46,
As a pretreatment, a batch of cellulose (20 g) was ball-milled in
a rotary mill for 48 h at RT. The hydrolysis of ball-milled microcrys-
talline cellulose was performed in a 100 mL stirred reactor (Parr In-
strument). The catalyst (0.300 g), cellulose (0.270 g), and DI water
6935–6937.
[
5] Y.-B. Huang, Y. Fu, Green Chem. 2013, 15, 1095–1111.
[6] K. Fukuhara, K. Nakajima, M. Kitano, H. Kato, S. Hayashi, M. Hara, Chem-
SusChem 2011, 4, 778–784.
(
27 mL) were loaded into the vessel and heated to 1508C. The mix-
ture was stirred continuously at 500 rpm. After reaction, the aque-
ous phase was analyzed by HPLC with refractive index detection
[7] H. Kobayashi, M. Yabushita, T. Komanoya, K. Hara, I. Fujita, A. Fukuoka,
ACS Catal. 2013, 3, 581–587.
[
8] P.-W. Chung, A. Charmot, O. A. Olatunji-Ojo, K. A. Durkin, A. Katz, ACS
Catal. 2014, 4, 302–310.
(
RID; Agilent 1260 Infinity HPLC with a Grace Davison Prevail Car-
bohydrate ES column) to determine the yield of glucose using the
same method as for the adsorption isotherms. The conversion of
cellulose was determined by mass difference before and after reac-
tion. For this purpose, the samples were dried at RT for 12 h, and it
was assumed that the mass of the catalyst remained constant
during the reaction.
[
9] a) O. M. Gazit, A. Charmot, A. Katz, Chem. Commun. 2011, 47, 376–378;
b) O. M. Gazit, A. Katz, Langmuir 2012, 28, 431–437; c) O. M. Gazit, A.
Katz, J. Am. Chem. Soc. 2013, 135, 4398–4402.
[
10] P.-W. Chung, A. Charmot, O. M. Gazit, A. Katz, Langmuir 2012, 28,
15222–15232.
[11] A. C. Ferrari, J. Robertson, Phys. Rev. B 2000, 61, 14095–14107.
[
[
12] a) M. Lezanska, P. Pietrzyk, Z. Sojka, J. Phys. Chem. C 2010, 114, 1208–
216; b) A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. Poschl,
Carbon 2005, 43, 1731–1742.
13] O. Beyssac, B. Goffe, J. P. Petitet, E. Froigneux, M. Moreau, J. N. Rouzaud,
Spectrochim. Acta Part A 2003, 59, 2267–2276.
1
Adsorption isotherms
Standard glucose solutions were prepared in various concentra-
tions (0.05, 0.1, 0.15, 0.2, 0.3, 0.4, 0.5, and 0.6m). Standard cello-
biose solutions were also prepared in various concentrations (0.01,
[14] S. Prawer, K. W. Nugent, D. N. Jamieson, J. O. Orwa, L. A. Bursill, J. L.
Peng, Chem. Phys. Lett. 2000, 332, 93–97.
[
15] a) V. S. Babu, M. S. Seehra, Carbon 1996, 34, 1259–1265; b) A. Mani-
vannan, M. Chirila, N. C. Giles, M. S. Seehra, Carbon 1999, 37, 1741–
0
.02, 0.03, 0.04, 0.06, 0.08, 0.10, and 0.12m). Preweighed amounts
of carbon catalysts (ꢀ40 mg) were mixed with sugar solution
1.000 mL) in 1.5 mL microcentrifuge tubes, vortexed for 30 s, and
1
747.
16] H. N. Cheng, L. H. Wartelle, K. T. Klasson, J. C. Edwards, Carbon 2010, 48,
455–2469.
17] H. Knicker, K. U. Totsche, G. Almendros, F. J. Gonzalez-Vila, Org. Geo-
chem. 2005, 36, 1359–1377.
(
[
[
[
placed on a shaker table for 24 h. The slurries were then centri-
fuged for 10 min at 10000 rcf. The supernatant was removed and
analyzed by HPLC–RID to determine the concentration of the
sugar molecules. Accurate concentrations were determined based
on an eight-point calibration curve. HPLC–RID analysis was per-
formed by using an Agilent 1260 Infinity HPLC system using
a Grace Davison Prevail Carbohydrate ES column. The mobile
phase consisted of 75 vol% acetonitrile and 25 vol% water. The
2
18] C. L. Su, M. Acik, K. Takai, J. Lu, S. J. Hao, Y. Zheng, P. P. Wu, Q. L. Bao, T.
Enoki, Y. J. Chabal, K. P. Loh, Nat. Commun. 2012, 3, 1298.
[19] D. Lennon, D. T. Lundie, S. D. Jackson, G. J. Kelly, S. F. Parker, Langmuir
2002, 18, 4667–4673.
[
[
[
20] G. Socrates, Infrared and Raman Characteristic Group Frequencies,
À1
3rd ed., Wiley, West Sussex 2001.
flow rate was 1.0 mLmin , and the injection volume was 0.6 mL.
21] A. H. Van Pelt, O. A. Simakova, S. M. Schimming, J. L. Ewbank, G. S. Foo,
E. A. Pidko, E. J. M. Hensen, C. Sievers, Carbon 2014, 77, 143–154.
22] H. B. Zhao, J. H. Kwak, Y. Wang, J. A. Franz, J. M. White, J. E. Holladay,
Energy Fuels 2006, 20, 807–811.
The uptake was calculated by q=(V(C ÀC ))/m , in which q is the
0
e
cat
À1
specific uptake [molg ], V is the volume of sugar solution used
[
L], C and C are the initial and equilibrium concentrations of the
0 e
À1
sugar solution, respectively [molL ], and m is the mass of
cat
[23] A. Bhatnagar, W. Hogland, M. Marques, M. Sillanpaa, Chem. Eng. J. 2013,
219, 499–511.
[24] a) G. X. Yu, M. Jin, J. Sun, X. L. Zhou, L. F. Chen, J. A. Wang, Catal. Today
2013, 212, 31–37; b) G. G. Stavropoulos, P. Samaras, G. P. Sakellaropou-
los, J. Hazard. Mater. 2008, 151, 414–421.
carbon material used [g]. Three isotherms were obtained for each
molecule on each adsorbent. For each isotherm, the Langmuir iso-
therm parameters were determined by linear regression (see Sup-
porting Information).
[
25] a) X. H. Qi, H. X. Guo, L. Y. Li, R. L. Smith, ChemSusChem 2012, 5, 2215–
220; b) B. B. Chang, J. Fu, Y. L. Tian, X. P. Dong, J. Phys. Chem. C 2013,
17, 6252–6258.
2
1
Acknowledgements
[
26] J. Zhang, H. L. Zou, Q. Qing, Y. L. Yang, Q. W. Li, Z. F. Liu, X. Y. Guo, Z. L.
Du, J. Phys. Chem. B 2003, 107, 3712–3718.
The Renewable Bioproducts Institute is acknowledged for the use
of its facilities. We thank Johannes Leisen for experimental assis-
[
27] W. Li, Y. Bai, Y. K. Zhang, M. L. Sun, R. M. Cheng, X. C. Xu, Y. W. Chen, Y. J.
Mo, Synth. Met. 2005, 155, 509–515.
13
[28] Y. Hong, Z. Wang, X. Jin, Sci. Rep. 2013, 3, 3439.
[
[
tance with C DP MAS NMR spectroscopy. Funding from the
Georgia Institute of Technology and the U.S. Department of
Energy (grant DE-AC36-08-GO28308) is gratefully acknowledged.
We thank Matthew Yung for fruitful discussions.
29] H. P. Boehm, Carbon 1994, 32, 759–769.
30] a) J. W. Lee, T. O. Kwon, I. S. Moon, Carbon 2004, 42, 371–380; b) J. W.
Yoo, T. Y. Kim, S. Y. Cho, S. G. Rho, S. J. Kim, J. Ind. Eng. Chem. 2004, 10,
257–266.
[
31] a) M. Yabushita, H. Kobayashi, J.-y. Hasegawa, K. Hara, A. Fukuoka,
ChemSusChem 2014, 7, 1443–1450; b) S. E. Kiehna, Z. R. Laughrey, M. L.
Waters, Chem. Commun. 2007, 4026–4028; c) Z. R. Laughrey, S. E.
Keywords: acidity · biomass · carbon · hydrolysis · oxidation
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&
9
&
ÞÞ
These are not the final page numbers!

CHEMSUSCHEM
FULL PAPERS
www.chemsuschem.org
Kiehna, A. J. Riemen, M. L. Waters, J. Am. Chem. Soc. 2008, 130, 14625–
[38] E. P. Barrett, L. G. Joyner, P. P. Halenda, J. Am. Chem. Soc. 1951, 73, 373–
380.
[39] C. Lastoskie, K. E. Gubbins, N. Quirke, J. Phys. Chem. 1993, 97, 4786–
4796.
1
4633.
32] S. Y. Sheu, D. Y. Yang, H. L. Selzle, E. W. Schlag, Proc. Natl. Acad. Sci. USA
003, 100, 12683–12687.
[
2
[
[
33] O. Bobleter, Prog. Polym. Sci. 1994, 19, 797–841.
34] C. A. Leon y Leon, J. M. Solar, V. Calemma, L. R. Radovic, Carbon 1992,
[40] a) A. M. Oickle, S. L. Goertzen, K. R. Hopper, Y. O. Abdalla, H. A. Andreas,
Carbon 2010, 48, 3313–3322; b) S. L. Goertzen, K. D. Theriault, A. M.
Oickle, A. C. Tarasuk, H. A. Andreas, Carbon 2010, 48, 1252–1261.
30, 797–811.
[35] Y. G. Yanovsky, Nano-Modeling Structure and Micromechanical Properties
of Mesoscopic Composite Systems, Tsinghua University Press, Beijing,
2
006.
Received: September 1, 2014
Revised: November 10, 2014
Published online on && &&, 0000
[
[
36] C. Loerbroks, R. Rinaldi, W. Thiel, Chem. Eur. J. 2013, 19, 16282–16294.
37] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309–319.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&10
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
Understanding the hydrolysis of cellu-
lose: Activated carbon is treated chemi-
cally with hydrogen peroxide and sulfu-
ric acid to understand the effects of
functional groups and morphology on
the hydrolysis of cellulose. Defect sites
on the carbon surface have been identi-
fied as a particularly active environment.
Through a synergistic effect with func-
tional groups, glucan chains are hydro-
lyzed with an increased activity.
G. S. Foo, C. Sievers*
& – &&
&
Synergistic Effect between Defect
Sites and Functional Groups on the
Hydrolysis of Cellulose over Activated
Carbon
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 0000, 00, 1 – 11
&11
&
ÞÞ
These are not the final page numbers!