A metal-free, carbon-based catalytic system for the oxidation of lignin model compounds and lignin

Article abstract of DOI:10.1002/cplu.201300439

Nitrogen-containing graphene material (LCN) has been identified as an effective catalyst for the oxidation of β-O-4 and α-O-4 types of lignin model compounds in the presence of tert-butyl hydroperoxide, to provide aromatic aldehydes, acids and other organic chemicals in high yield. The transformations of five lignin model compounds over LCN were investigated systematically. Instrumentation analysis, kinetic study and radical trapping experiments highlight the mechanistic features of the reaction, including: 1) the reaction pathway starts by benzylic C-H or C-OH bond activation, followed by C_α-C_β or C_α-O bond cleavage, and finally further oxidation of intermediate aromatics; and 2) the reaction follows a free-radical mechanism with all the key steps involving radical species. In addition, the LCN proved to be a highly stable catalyst; no significant activity decrease was observed for four consecutive runs, and X-ray photoelectron spectroscopy analysis indicates negligible decrease in the content of the active nitrogen species in the catalyst. Notably, this new catalytic system can be extended to the oxidative depolymerisation of real lignin, to produce a significant portion of liquefied, low-molecular-mass products. Low mass, high conversion: Nitrogen-containing graphene-based carbon material has been used to establish a metal-free catalytic system for the oxidation of α-O-4 and β-O-4 types of lignin model compounds in the presence of tert-butyl hydroperoxide (see figure). In the oxidative depolymerization of organosolv lignin, a significant portion of liquefied, low-molecular-mass product is produced. Copyright

Full text of DOI:10.1002/cplu.201300439

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300439
A Metal-Free, Carbon-Based Catalytic System for the
Oxidation of Lignin Model Compounds and Lignin
Yongjun Gao,[a] Jiaguang Zhang,[a] Xi Chen,[a] Ding Ma,[b] and Ning Yan*[a]
Nitrogen-containing graphene material (LCN) has been identi-
fied as an effective catalyst for the oxidation of b-O-4 and a-O-
4 types of lignin model compounds in the presence of tert-
butyl hydroperoxide, to provide aromatic aldehydes, acids and
other organic chemicals in high yield. The transformations of
five lignin model compounds over LCN were investigated sys-
tematically. Instrumentation analysis, kinetic study and radical
trapping experiments highlight the mechanistic features of the
reaction, including: 1) the reaction pathway starts by benzylic
CÀH or CÀOH bond activation, followed by CaÀCb or CaÀO
bond cleavage, and finally further oxidation of intermediate ar-
omatics; and 2) the reaction follows a free-radical mechanism
with all the key steps involving radical species. In addition, the
LCN proved to be a highly stable catalyst; no significant activi-
ty decrease was observed for four consecutive runs, and X-ray
photoelectron spectroscopy analysis indicates negligible de-
crease in the content of the active nitrogen species in the cata-
lyst. Notably, this new catalytic system can be extended to the
oxidative depolymerisation of real lignin, to produce a signifi-
cant portion of liquefied, low-molecular-mass products.
Introduction
It is not unreasonable to speculate that biomass, such as cellu-
lose, lignin, even chitin, would be the major source for certain
fine chemicals or platform compounds in the post-fossil fuel
era.[1] Among various biomass materials, lignin could be seen
as the major resource for sustainable aromatic chemicals, be-
cause lignin is the most abundant, renewable aromatic biopo-
lymer on earth.[2] There are three major strategies to depoly-
merise lignin into an array of simple compounds: hydrogenoly-
sis,[3] hydrolysis[4] and oxidation.[5] Among these, the oxidative
depolymerisation of lignin is unique because it leads to highly
functionalised monomeric or oligomeric products, which can
be used directly as fine chemicals or as platform chemicals.[5c,6]
It has long since been established that lignin has a complex,
mildly branched polymeric structure composed of a few aro-
matic monomers randomly coupled and cross-coupled by CÀO
or CÀC bonds. The CÀO bond is the major linkage among the
monomers, with b-O-4, a-O-4 and 4-O-5 linkages being most
representative. The CÀC bond contributes a minor fraction
(about a third) to the linkages in lignin.[7] To generate com-
modity chemicals from lignin, these linkages have to be
broken down. Under oxidative conditions, the benzylic CÀH or
CÀOH bonds in lignin can be attacked relatively easily and
transformed to carbonyl groups. Then the CaÀO or CaÀCb
bond undergoes cleavage to afford depolymerised products.[5c]
The key to achieving the selective oxidation of lignin into
chemicals relies on the development of appropriate catalysis
technologies. The oxidation of lignin and associated model
compounds has been studied extensively.[2d,5c] A vast majority
of reported oxidative catalytic systems are based on using
metal-containing catalysts, and many of them suffer from low
efficiency, harsh reaction conditions and/or use toxic met-
als.[7a,8] Recently, a metal-free catalytic system consisting of 4-
acetamido-TEMPO (TEMPO=2,2,6,6-tetramethyl-1-piperidinyl-
oxy) in combination with HNO3 and HCl was developed to cat-
alyse the oxidation of benzylic alcohol as a lignin model com-
pound.[9] Nevertheless, efficient, metal-free catalytic systems for
the oxidation of lignin model compounds and real lignin
remain very limited. Also, the reaction pathway and the reac-
tion mechanism of lignin oxidation in metal-free catalytic sys-
tems remain elusive.
In this work, systematic evaluation of the oxidation of vari-
ous lignin model compounds over nitrogen-containing gra-
phene material (LCN) in the presence of organic peroxides was
conducted. Detailed product identification, kinetic study and
free-radical trapping experiments highlight the salient mecha-
nistic features of this new catalytic oxidation system for lignin
conversion. To our delight, this carbon-based catalytic system
is even able to transform organosolv lignin, which leads to
a significant portion of liquefied, partially depolymerised prod-
ucts. The depolymerised products can be further hydrolysed to
low-molecular-weight compounds, such as diethyl succinate
and diethyl phthalate, in acidic ethanol–water solution.
[a] Dr. Y. Gao, J. Zhang, X. Chen, Prof. N. Yan
Department of Chemical and Biomolecular Engineering
National University of Singapore
4 Engineering Drive 4, 117576 Singapore (Singapore)
E-mail: ning.yan@nus.edu.sg
[b] Prof. D. Ma
Beijing National Laboratory for Molecular Sciences
College of Chemistry and Molecular Engineering
Peking University, 202 Chenfu Road, Beijing 100871 (P. R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300439.
This article is part of the “Early Career Series”. To view the complete series,
visit: http://chempluschem.org/earlycareer.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 825

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
which according to DFT calculations[10] is responsible for the
catalytic activity of LCN, is the major constituent (58% accord-
ing to XPS fitting results). Overall, the synthesised LCN in this
study shows a graphene-based sheet structure containing cat-
alytically active, graphitic N species (>3 wt%).
Results and Discussion
LCN characterisation
Compared with normal graphene material, the exceptional cat-
alytic performance of LCN originates from its nitrogen
doping.[10] Therefore, the LCN in this study was prepared by
following a modified literature method by using chemical
vapour deposition (CVD) with pyrolytic graphene oxide as sub-
strate and acetonitrile as nitrogen source, in which the nitro-
gen content of LCN was maximised by elongated exposure to
acetonitrile.[10] The synthesised LCN can be regarded as a nitro-
gen-containing, layered carbon material possessing highly ex-
foliated graphitic structures. Transmission electron microscopy
(TEM) and scanning electron microscopy (SEM) analysis (see
Figure 1a,b) clearly show that the LCN possesses a graphitic
sheet structure. X-ray photoelectron spectroscopy (XPS; see
Figure 1c) exhibits only three sets of peaks, at around 284.8,
401.5 and 531.8 eV, corresponding to carbon, nitrogen and
oxygen, respectively. Based on XPS analysis, the nitrogen con-
tent in the LCN we used was 6.2 wt%, which is similar to that
of LCNs exhibiting excellent catalytic activity in the literature.[10]
Fitting results of the N1s XPS spectrum are shown in Fig-
ure 1d. The peaks at 398.6, 399.6, 401.5 and 403.5 eV are as-
signed to pyridinic N, pyrrolic N, graphitic N and oxidised N, re-
spectively.[11] Among these four nitrogen species, graphitic N,
Oxidation of lignin model compounds
Both the a-O-4 and the b-O-4 types of lignin model com-
pounds contain the benzylic CÀH or CÀOH bond. As such, our
strategy for lignin conversion is to utilise LCN as a catalyst to
activate the benzylic CÀH or CÀOH bond in the presence of an
oxidant. Previous works on the oxidation of lignin and its
model compounds employed oxygen and/or H2O2 as oxidant
owing to their low price.[5a,b,12] Our preliminary data (Table 1,
entry 1), however, indicated that H2O2 is not suitable in the
LCN-catalysed oxidation of lignin model compounds, presuma-
bly because of the fast decomposition of H2O2 under the ap-
plied conditions.
Therefore, tert-butyl hydroperoxide (TBHP), which is a more
stable oxidant widely used in industrial processes, was chosen
for further study. We started the work by evaluating the per-
formance of LCN in the oxidation of lignin model compounds
in the presence of TBHP. 1-(Benzyloxy)-2-methoxybenzene (a;
see Table 1) and benzyl phenyl ether (b) were selected as a-O-
4 type model compounds whereas 2-phenoxy-1-phenylethanol
(c), 2-(2,6-dimethoxyphenoxy)-1-
phenylethanol (d) and 2-(2,6-di-
methoxyphenoxy)-1-phenyletha-
nol (e) were selected as b-O-4
type lignin model compounds.
Both a-O-4 and b-O-4 types of
model compounds contain two
benzene rings. For clarity, the
benzene ring directly connected
with a carbon atom, that is, the
ring associating with the benzyl-
ic CÀH or CÀOH bonds, is denot-
ed as the A ring, whereas the
benzene ring directly connected
with an oxygen atom is denoted
as the B ring.
The results of the catalytic oxi-
dation of a-O-4 type model
compounds (a,b) over LCN are
compiled in Table 1. Only small
amounts of benzaldehyde 1
(0.6 mol%) and 2-methoxyphe-
nol 3 (0.2 mol%) were formed in
the blank experiment (Table 1,
entry 2), whereas noticeable con-
version and yield were achieved
by employing LCN as a catalyst.
As mentioned, the conversion of
compound a and the yields of
aromatic products were low
Figure 1. a) TEM and b) SEM images of LCN material; c) XPS spectrum and d) XPS N1s spectrum of LCN.
when H2O2 was used as the oxi-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 826

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
employed to extract the prod-
ucts from water) and escape GC
and GC–MS analysis. To verify
this assumption, the water
phase after extraction was
freeze-dried and analysed by
GC–MS (see Figure S1 in the
Supporting Information). Indeed,
a variety of dicarboxylic acids
and their derivatives, such as
oxalic acid, maleic anhydride,
maleic acid, fumaric acid and
muconic acid, were detected. In
a following experiment, 2-me-
thoxyphenol 3 was used as the
substrate instead of a under the
same reaction conditions. No 2-
methoxyphenol survived after
the reaction. These results dem-
onstrated that the 2-methoxy-
phenol generated from the
B ring is further oxidised into
water-soluble chemicals through
a series of reactions. The first
step is likely to be similar to the
reported Malaprade reaction, in
which o-quinone and methanol
are generated from 2-methoxy-
phenol and an oxidant.[13] Fol-
Table 1. The oxidation of a-O-4 type lignin model compounds under different conditions.[a]
Entry Substrate Conversion
[%]
Total yield
[mol%]
Yield [mol%]
1
2
3
4
5
6
7
1[b]
2[c]
3[d]
4
5[e]
6[f]
a
a
a
a
a
a
25.0
9.0
40.1
66.5
89.4
98.0
9.1
0.8
25.9
41.9
57.0
71.0, 7.2 (w1)
2.5 4.1
0.6 0
15.9 2.0
14.5 1.9
5.6 2.7
2.2
0.2
2.9 2.2
1.4 21.0
1.5 40.6
0
0
0
0
0.08
0.5
1.7
0
0
0
0
0
0.3
0
2.8
2.6
4.9
9.3
3.0 1.5, 0.8 (w2) 0.9 45.5, 0.3 (w3) 1.4
0.4, 0.7
(w4)
0.1
–
7[g]
8[e]
a
b
85.4
68.6
54.9
74.2
6.9 2.4
13.2 2.9
1.5 40.4
19.0 33.7
0.6
0.5
3.0
4.9
[a] Reaction conditions: substrate (0.5 mmol), LCN (0.01 g), TBHP (6 equiv, 70 wt% in water), H2O (3 mL), 808C,
12 h. The conversion and yield were determined by GC analysis (only organic-phase products were quantified).
[b] H2O2 (20 equiv) was used as oxidant. [c] No catalyst was used. [d] 3 equiv TBHP. [e] Reaction time is 24 h.
[f] 12 equiv TBHP was used and the yields of compounds in the water phase (w1–w4) were quantified. [g] Re-
action temperature 1208C.
dant, even if H2O2 was used in a large excess (20 equiv;
Table 1, entry 1). TBHP, on the other hand, exhibited superior
oxidative ability under similar reaction conditions. Three equiv-
alents of TBHP at 808C led to 40% conversion with a variety of
aromatic chemicals detected (Table 1, entry 3). The conversion
of a further increases on increasing the amount of TBHP, reac-
tion temperature and/or reaction time (Table 1, entries 3–7).
The catalytic system is also applicable with model com-
pound b, affording comparable substrate conversion and prod-
uct yield (Table 1, entry 8). After the reaction, seven aromatic
compounds (1–7) were identified, with benzaldehyde 1 and
benzoic acid 4 as the major products. Products 1–6 are mono-
aromatic compounds formed by oxidative CaÀO bond break-
age, which highlights that LCN is an effective catalyst for the
oxidative decoupling of a-O-4 type lignin model compounds
using TBHP as an oxidant. Product 7, which is generated by
benzylic CÀH bond oxidation without CÀO bond cleavage, was
also detected but in a small amount.
lowing that, the o-quinone reacts with TBHP to form muconic
acid, which is further oxidised to other dicarboxylic acids (Fig-
ure S2). Quantitative analysis of one experiment indicates
a combined yield of approximately 10 mol% for water-soluble
products, including maleic anhydride (w1, 7.2 mol%), hexadie-
noic acid (w4, 0.7 mol%) and two others (Table 1, entry 6).
The tendency for over-oxidation of the B ring product(s) into
dicarboxylic acids and their derivatives is closely related to the
nature of the substituent groups (R1 and R2). Starting from
b (R1 =R2 =H), the yield of phenol (product from the B ring)
reaches 19.0 mol%, which indicates that the ortho-methoxy
group enables the molecule to be more susceptible towards
over-oxidation owing to the electron-donating property of the
methoxy group.
The LCN-catalysed oxidations of lignin b-O-4 model com-
pounds (c–e) were also investigated (Table 2). Compound c
was selected for reaction condition optimisation. Not surpris-
ingly, no reaction occurred without adding LCN. At 808C for
12 hours, the conversion of c reaches 69.3% in the presence of
LCN. The major product is ketone ether 11 generated from the
oxidation of the benzylic CÀOH bond without CÀO or CÀC
bond cleavage. Benzoic acid 4 was obtained as the dominant
monomer along with other monomeric products, such as
phenyl formate 8 and phenol 3 (Table 2, entry 1). Notably, the
fact that 4 was the major product from the A ring indicates
that the major reaction pathway involves CaÀCb bond cleav-
age, which is further corroborated by the existence of 8 in the
product. An increase in the reaction time to 24 hours led to an
A close examination revealed that the mono-aromatic prod-
ucts (1, 2, 4–6) are generated from the A ring whereas 3 is
generated from the B ring. In principle, the combined yield of
the products from the A ring should equal that of the products
from the B ring. However, the yield of 3 (2-methoxyphenol in
this case) was significantly lower than the combined yield of 1,
2 and 4–6, which suggests that product(s) from the B ring un-
derwent further transformation into polar, non-aromatic com-
pounds. These chemicals stay preferentially in the water phase
upon organic solvent extraction (note that ethyl acetate was
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 827

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tions. As shown in Figure 2a, the
initial reaction rate of a is high,
reaching 53% conversion after
3 hours. Afterwards, the conver-
sion keeps increasing smoothly.
Table 2. The oxidation of b-O-4 type lignin model compounds under different conditions.[a]
Compounds 1,
3 (R1 =H, R2 =
OMe), 4 and 7 (R1 =H, R2 =OMe)
are the major products, and they
exhibit very interesting selectivi-
ty changes as the reaction
evolves. The yield of the dimeric
product 7 (R1 =H, R2 =OMe)
reaches its maximum at t=1 h
and drops afterwards. The yields
of monomeric products 1 and 3
Entry Substrate T [8C] t [h] Conversion [%] Total yield [mol%]
Yield [mol%]
1
2
3
4
5
8
9
10 11
1
2
3
4
5
c
c
c
d
e
80
80
12 69.3
24 88.0
68.7
73.5
97.4
62.7
72.9
0.1 0.6 1.0 16.0 4.6 0.4 0.7 1.0 44.3
0.2 1.2 3.1 24.5 3.3 0.3 0.2 3.1 37.6
0.2 2.1 18.3 45.3 2.0 0.2 0.3 18.3 10.7
120 24 91.8
120 24 95.8
120 24 quant.
0.1 4.3 5.1 42.4 0.5
0.1 8.7 33.5
0
0
0
14.6
0
0
10.3
16.0
increase in the first
3 and
0
0
0.5 hours, respectively, and then
decrease. On the contrary, the
[a] Reaction conditions: substrate (0.5 mmol), LCN (0.01 g), TBHP (6 equiv, 70 wt% in water), H2O (3 mL). The
conversion and yield were determined by GC analysis (only organic-phase products were quantified). quant. =
quantitative.
yield of
4 keeps increasing
during the entire process. These
increased conversion of c and in-
creased yield of monomers, but
11 remained as the major prod-
uct. By increasing the reaction
temperature to 1208C, CaÀCb
bond cleavage is significantly
promoted: the A ring is mainly
converted into
4 (45.3 mol%)
whereas the B ring is mainly con-
verted into 3 (18.3 mol%) and
10
(Table 2, entry 3).
Two other b-O-4 type lignin
model compounds, and e,
(18.3 mol%)
monomers
d
were also converted into mono-
meric compounds with satisfac-
tory yield under optimised con-
ditions (Table 2, entries 4 and 5).
Figure 2. Product distributions for the conversion of a) a and b) c over LCN as functions of time. Reaction condi-
tions: substrate (0.5 mmol), TBHP (6.0 mmol, 70 wt% in water), LCN (0.01 g), H2O (3 mL), 808C; the conversion and
yield were determined by GC.
The major difference is that the yields for the major products
from the B ring, that is, compounds 3 and 10, are considerably
lower. This result is in full agreement with the observations of
the LCN-catalysed a-O-4 type model compound conversion, in
which we identified that the ortho-methoxy substituents de-
crease the stability of the aromatic ring significantly. The ex-
periments shown in Table 2 demonstrate effective catalytic oxi-
dation of b-O-4 type lignin model compounds in the presence
of LCN and TBHP, despite the fact that a slightly higher tem-
perature is required to achieve reasonable monomer yields.
results suggest that the oxidative decoupling of a over LCN
first occurs at the position of the benzylic CÀH bond to gener-
ate dimeric product 7 (R1 =H, R2 =OMe). Cleavage of 7 at the
position of CaÀO generates products 1 and 3 (R1 =H, R2 =
OMe). Product 1 is further oxidised into its acid form 4, which
is a stable end product. Product 3 is highly unstable under the
applied conditions and is decomposed, in agreement with the
observation of dicarboxylic acids and associated derivatives
after the reaction.
Kinetic study employing c as the substrate provides similar
information (Figure 2b). Compound c was converted smoothly.
The dimeric product 11 (R1 =R2 =H) is the main product within
the first 12 hours, which indicates that the reaction starts from
the oxidation of the benzylic CÀOH bond. Following that, the
decomposition of 11 (R1 =R2 =H) leads to the formation of aro-
matic monomers, which is the rate-determining step in this
case. Similar to a, 4 is the stable end product after oxidation of
c via 1 as an intermediate.
Kinetics and reaction mechanism
Kinetic studies on the oxidation of compounds a and c over
LCN in a period of 24 hours at 808C were conducted (see
Figure 2 and detailed data in Table S1). In both cases, no signif-
icant drop in activity was observed in the entire period, thus
indicating that LCN remains active under these reaction condi-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 828

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
A vast majority of catalytic systems for lignin oxidation in-
volve free radicals during the reaction.[5c] It is also widely re-
ported that TBHP acts as a free-radical initiator, either generat-
ing tert-butoxide and hydroxyl radicals under microwave irradi-
ation[6a] or reacting with metal catalysts to form tert-butylper-
oxy radicals.[3g] As such, it is not unreasonable to speculate
that LCN-catalysed lignin model compound oxidation in the
presence of TBHP follows a free-radical mechanism. To identify
the radical species produced during the reaction, radical trap-
ping experiments with TEMPO as a radical scavenger were con-
ducted. Compounds a and c were used as substrates and
TEMPO was added into the system 5–10 minutes after the re-
action to trap the active intermediate species.
cals to produce 3. In the presence of excess TBHP, 1 can be fur-
ther oxidised to 4 whereas 3 decomposes to dicarboxylic acids
and its derivatives in a similar manner to the Malaprade reac-
tion. Besides, 4 can react with tert-butanol to form tert-butyl
benzoate 5.
With b-O-4 type lignin model compounds, the initiation step
and reaction mechanism are similar to the case of a. In radical
trapping experiments, 12, 14, 15 and 16 adducts were detect-
ed (see Scheme 1 and Figure S3), thus indicating that the ho-
molytic cleavage happens at the CaÀCb bond to afford benzoyl
radicals and phenoxyl methyl radicals. These radicals can be
further converted to benzoic acid 4 and phenyl formate 8. In
the presence of excess TBHP, 8 can be converted to phenol 3
and carbon dioxide (Scheme 2b).
Employing a as the substrate, four TEMPO adducts, denoted
as 12, 13, 14 and 15 (see Scheme 1 and Figure S3), were de-
tected by GC–MS, thus demonstrating the formation of at least
four radical species during the reaction. Benzoyl radicals and 2-
methoxyphenoxyl radicals are derived from the A and B rings
Recycling experiment
A significant advantage of using LCN as the catalyst is its facile
recovery. Evaluation of the long-term catalytic stability of LCN
was undertaken at a reaction temperature of 1208C employing
c as the substrate. After one batch reaction, LCN was filtered
out of the reaction mixture, washed and reused. Four batch re-
actions were performed without any further treatment. In the
first run a high conversion of 85% was obtained with LCN. To
our delight, the activity of LCN remains essentially constant in
the second batch reaction and only decreases by about 5% in
the fourth batch reaction (Figure 3a and Table S2). Interesting-
ly, a change in product distribution was observed during the
recycling experiments. For example, the yield of 4 decreased
from 29.0 to 19.4 mol%, and correspondingly the yield of 11
(R1 =R2 =H) increased from 28.3 to 37.6 mol%, after four batch
reactions. Considering 11 is formed through benzylic CÀOH
bond oxidation, whereas 4 is formed by CaÀCb bond cleavage,
it appears that the catalytic activity of LCN for the oxidation re-
action was maintained upon recycling whereas the ability to
break the CÀC bond decreases gradually.
Scheme 1. The TEMPO adducts detected by GC–MS in the radical trapping
experiment using a and c as the substrates.
of the substrate, respectively (trapped as 12, 13), and methyl
radicals and tert-butoxyl radicals are formed from TBHP
(trapped as 14, 15). By combining the kinetic study, the radical
trapping experiments and the previous literature reports,[3g,6a]
a plausible mechanism for the oxidation of lignin model com-
pound a was proposed (Scheme 2a). In the first step, the tert-
butoxyl radical forms from TBHP to initiate the reaction, which
could as well further decompose to acetone and methyl radi-
cals.[14] Meanwhile, LCN reacts with TBHP to afford LCN-OH,
which contains active oxygen species at the carbon atoms
neighbouring the graphitic nitrogen.[10] Then, the benzyl hy-
drogen, which is the least stable site of the substrate, gets ab-
stracted by the tert-butoxyl radical to afford benzylic radicals
and tert-butanol. The benzylic radicals convert to benzylic alco-
hol by OH transfer from LCN-OH, which concurrently regener-
ates the LCN catalyst. On repeating this step a dehydration re-
action product 7 is formed. The above mechanism involves
similar steps to the reported oxidation of alkyl-substituted aro-
matics with TBHP as oxidant.[6a,10] In the next step, the homo-
lytic cleavage of 7 takes place at the aliphatic CaÀO bond to
form benzoyl and 2-methoxyphenoxyl radicals, which have
been trapped by TEMPO. Benzoyl radicals react with hydrogen
radicals or methyl radicals in the next step, and the homolytic
cleavage of 7 takes place at the aliphatic CaÀO bond to form
benzoyl and 2-methoxyphenoxyl radicals, which have been
trapped by TEMPO. Benzoyl radicals react with hydrogen radi-
cals or methyl radicals to generate 1 and 2, respectively,
whereas 2-methoxyphenoxyl radicals react with hydrogen radi-
To shed light on structural/compositional changes of the cat-
alyst during the recycling, XRD and XPS were used to charac-
terise LCN after recycling. In the XRD pattern of LCN (Fig-
ure 3b), the peak at 2q=26.48 is attributed to reflection on
the (002) planes of well-ordered graphene, which is very simi-
lar to the diffraction pattern of graphite. The shoulder peak at
about 2q=308 and the broad hump at 2q=178 reflect the dis-
ordered state of graphene sheets.[15] LCN after reaction exhibits
almost identical diffraction patterns to that of LCN, which sug-
gests that the crystal structure of LCN does not change after
recycling. According to the XPS spectra (Figure 3c,d), the nitro-
gen content in LCN after reaction decreased slightly from 6.2
to 5.4 wt% relative to fresh LCN. Nevertheless, curve-fitting
analysis indicates that the content of graphitic nitrogen is still
the predominant nitrogen species (3.3 wt%), similar to unused
LCN, which explains the negligible decrease in catalytic activity
during repeated oxidation reactions. Compared with fresh LCN
samples, the oxygen content in used LCN increased significant-
ly (from 3.5 to 14.4 wt%), not unreasonably though, as we pro-
pose the formation of LCN-OH between LCN and TBHP to be
the first step in the reaction mechanism.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 829

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Scheme 2. Proposed mechanism of oxidative depolymerisation of a) a and b) c catalysed by LCN.
Oxidation of birch wood lignin
the results were compared with that of the starting material.
The average molecular weight (MW) of the oil fraction is 4800,
much lower than that of the starting material (MW 250000).
The MW of the solid fraction is similarly small, despite exhibit-
ing a wider MW distribution. GPC analysis indicates a significant
depolymerisation of organosolv lignin over LCN in the pres-
ence of TBHP.
We conducted preliminary studies employing the LCN-based
catalytic system for the oxidation of organosolv lignin extract-
ed from birch wood. Considering the more complex structure
of real lignin, a higher temperature was employed (24 h,
12 equiv TBHP, 1408C). Two fractions, that is, an orange oil frac-
tion (45.8 wt%) and a brown solid fraction (33.6 wt%), were
obtained after the reaction (Figure S5). The orange oil fraction
is water soluble and obtained directly after filtration and
freeze-drying. The brown solid fraction is obtained by extract-
ing the remaining solid with ethanol. Both fractions were ana-
lysed by gel permeation chromatography (GPC; Figure S6) and
Figure 4 shows the FTIR spectra of the orange oil and the
brown solid fractions, as well as of unreacted lignin. The IR
spectra of the oil and the solid fractions are very different from
that of the unreacted lignin. First, the intensities of the bands
at 1713 and 1177 cmÀ1 increase remarkably and a shoulder
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 830

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
bonds in the a-O-4 linkage in
lignin are oxidised to C=O
groups without CaÀO bond
cleavage. As a result, it is evident
that the oxidation of birch lignin
occurs,
thereby
generating
products.
oxygen-enriched
Second, a new band arises at
about 1374 cmÀ1 in the product
spectra, which can be attributed
to the phenolic hydroxyl group
vibrations,[18] thus providing
strong evidence of the cleavage
of ether bonds in lignin. Further-
more, the intensity of the band
at 2963 cmÀ1 ascribed to CÀH
stretching vibrations in CH3 in-
creases, whereas the bands at
1514 and 1325 cmÀ1 ascribed to
the aromatic ring and CÀO
stretching vibration in syringyl
units decrease in the product
spectra, which indicates that the
aromatic structure and CÀO
bonds were partially decom-
posed.[19] The orange oil product
seems to be more oxidised and
its IR spectrum exhibits a stron-
ger band at 3209 cmÀ1 than the
brown solid. More encouraging-
Figure 3. a) Conversion and product distributions of recycling experiments. Reaction conditions: c (0.5 mmol),
TBHP (6.0 mmol, 70 wt% in water), LCN (0.01 g), H2O (3 mL), 1208C, 12 h; the conversion and yield were deter-
mined by GC. b) XRD patterns and c,d) XPS spectra of LCN and used LCN.
ly, 13C NMR analysis of the oil fraction revealed the
presence of small-molecule products (Figure S7).
Overall, LCN exhibits similar catalytic behaviour in the
oxidation of lignin model compounds and birch
wood lignin, that is, it catalyses the oxidation of ben-
zylic CÀH and CÀOH bonds into carbonyl groups,
breaks the CaÀCb and CaÀO bonds, and further catal-
yses the decomposition of some unstable aromatic
compounds. A blank experiment employing birch
lignin with TBHP revealed that the structure of the
lignin could be partly disrupted even without LCN,
but to a lesser extent, which highlights the positive
role of LCN in the oxidation depolymerisation of real
lignin (see Figure S8 for FTIR analysis).
We further performed acid-catalysed hydrolysis
with both fraction I (orange oil product) and fractio-
n II (brown solid product) in ethanol–water solution
at 1508C. For fraction I, diethyl succinate and other
low-molecular-weight products (MW<400, according
to GC–MS spectra) were identified in GC–MS analysis.
For fraction II, diethyl succinate, diethyl phthalate and other
low-molecular-weight products (MW<400, according to GC–
MS spectra) were obtained. Combined, 0.13 wt% of diethyl
succinate and 0.75 wt% of diethyl phthalate were obtained.
Notably, no such products were identified by GC–MS analysis
by direct hydrolysis of birch lignin under the same hydrolysis
conditions (see Figures S9–S11 for GC–MS spectra).
Figure 4. FTIR spectra of birch lignin and the oxidation products of birch lignin.
peak appears at approximately 1643 cmÀ1 in the spectra of the
products. The bands at 1713 and 1643 cmÀ1 are attributed to
non-conjugated and conjugated carbonyl stretching, respec-
tively, and the band at 1177 cmÀ1 is assigned to the carbonyl
stretching of conjugated ester groups.[16] The changes indicate
that an appreciable number of CÀH and/or CÀOH bonds in
lignin are oxidised to C=O groups.[17] Besides, a portion of CÀH
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 831

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
a JEM-6700F scanning electron microscope (JEOL). The sample was
immobilised on a copper substrate by conductive adhesives with-
out further processing. X-ray photoelectron spectroscopy (XPS)
data were obtained on a VG ESCALAB MKII spectrometer, using
a monochromated AlKa X-ray source (hn=1486.71 eV, 5 mA, 15 kV).
The data were converted into VAMAS file format and imported
into the CasaXPS software package, calibrated by the C1s signal
(285.0 eV) and further processed. X-ray diffraction (XRD) was per-
formed on a Bruker D8 advanced diffractometer using CuKa (l=
1.5406 ꢁ) radiation (40 kV, 30 mA cathodic current). Diffraction pat-
terns were recorded within a 2q range of 5–808 in a period of
32 min. Reaction products were analysed by gas chromatography
(GC) and gas chromatography–mass spectrometry (GC–MS) on an
Agilent 7890A gas chromatograph with a flame ionisation detector
and an Agilent 7890A-5975 GC–MSD instrument, both equipped
with HP-5 capillary columns (30 mꢂ250 mm). In the oxidation of
lignin, the solution of products was firstly frozen into ice in
a fridge at À208C and then was freeze-dried on a Martin Christ
Freeze Dryer with the vacuum at 0.31 mbar and ice condenser at
À428C.
Conclusion
By employing nitrogen-containing graphene material (LCN) as
a catalyst and tert-butyl hydroperoxide as an oxidant, we sys-
tematically studied the oxidative decoupling of a-O-4 and b-O-
4 types of lignin model compounds in water. Under optimised
conditions, high yields of monomeric products were obtained.
The reaction follows a free-radical mechanism and the struc-
tures of key intermediates were identified by free-radical trap-
ping experiments. Kinetic study and control experiments high-
light that the key step for a-O-4 model compound oxidation is
the aliphatic CaÀO bond cleavage, whereas the key step for
the b-O-4 type of lignin model compound is the cleavage of
the CaÀCb bond. The LCN proved to be highly stable and
could be used at least several times without appreciable de-
crease in activity.
This novel oxidative system can convert birch wood lignin
into depolymerised products including a significant portion of
liquefied product. These products may undergo further trans-
formations more easily than untreated lignin. Further study is
ongoing in our laboratory to enhance the yield of monomeric
chemicals from real lignin.
Preparation of graphene oxide
In a first step, graphene oxide (GO) was synthesised according to
a modified Hummer’s method.[22] Briefly, natural graphite powder
(5 g), sodium nitrate (5 g) and concentrated sulfuric acid (230 mL)
were added to a flask placed in an ice bath. Then KMnO4 (30 g)
was added slowly to the mixture under vigorous stirring to avoid
the temperature of the suspension exceeding 208C. The suspen-
sion was kept in the ice bath under stirring for 4 h, and was trans-
ferred into a water bath (358C) for 4 h under stirring. Afterwards,
distilled water (460 mL) was added slowly to the suspension. The
flask was transferred into a preheated oil bath for 4 h, with the
temperature set at 988C. The reaction system was diluted with dis-
tilled water (1000 mL) followed by addition of H2O2 solution (30%,
100 mL) to remove residual KMnO4 and MnO2. The solid product
was separated by filtration, washed repeatedly with 5% HCl solu-
tion until sulfate could not be detected by BaCl2, and finally
washed with water to pH 7. The product was dried in an air oven
at 608C for 120 h. Note, the major difference between this proce-
dure and the original Hummer’s method is that the duration of the
three main steps, that is, in the ice bath (08C), water bath (358C)
and oil bath (988C), in the synthesis is longer in this study to
enable the graphene oxide to have the highest oxygen content.
Experimental Section
Chemicals
Potassium permanganate (KMnO4), sulfuric acid (H2SO4), barium
chloride (BaCl2) and hydrogen chloride (HCl) were purchased from
Beijing Chemical Industry Group Co., Ltd. 2-Bromoacetophenone,
sodium nitrate (NaNO3), potassium carbonate (K2CO3), phenol,
guaiacol, hydrogen peroxide (30 wt%, H2O2), tert-butyl hydroperox-
ide (70 wt% in water, TBHP) and 2,2,6,6-tetramethyl-1-piperidiny-
loxy (TEMPO) were purchased from Sigma–Aldrich. Natural flake
graphite (99.99%) was purchased from Alfa Aesar. All these com-
mercially available chemicals were used as received. 1-(Benzyloxy)-
2-methoxybenzene (a), benzyl phenyl ether (b), 2-phenoxy-1-phe-
nylethanol (c), 2-(2,6-dimethoxyphenoxy)-1-phenylethanol (d), 2-
(2,6-dimethoxyphenoxy)-1-phenylethanol (e)[20] and organosolv
lignin[21] were prepared by following literature procedures.
Characterisation
Gel permeation chromatography (GPC) analysis was performed
with a system equipped with a Waters 2410 refractive index detec-
tor, a Waters 515 HPLC pump and two Waters styragel columns
(HT 3 and HT 4) with dimethylformamide as eluent at a flow rate of
1 mLminÀ1 at 258C. The raw data were processed using narrow
polystyrenes as calibrations on software Breeze. Fourier transform
infrared (FTIR) spectroscopy was performed on a Bruker Equinox 55
infrared spectrometer. The number of scans was 16 with a resolu-
tion of 4 cmÀ1 over the range of 4000–400 cmÀ1. Solid sample was
finely mixed with KBr before pressing into a pellet for measure-
ment. Oil sample was measured directly using KBr windows. Nucle-
ar magnetic resonance (NMR) spectra were recorded on a Bruker
400, AV400 NMR spectrometer using [D6]dimethyl sulfoxide as the
solvent. High-resolution transmission electron microscopy (HRTEM)
images were taken on a Tecnai F30 (FEI) field-emission transmission
electron microscope operated at an accelerating voltage of 300 kV.
Scanning electron microscopy (SEM) images were obtained on
Preparation of nitrogen-containing graphene-based carbon
LCN was prepared in a two-step process from GO. In a first step,
pyrolytic graphite oxide (PGO) was prepared through rapid heating
of GO in a tube furnace under a hydrogen atmosphere. The heat-
ing started from room temperature at a heating rate of 208CminÀ1
and was stopped once the temperature reached 8008C. In
a second step, the chemical vapour deposition (CVD) method was
applied to prepare LCN from PGO and acetonitrile, which acted as
the substrate and the nitrogen source, respectively. Acetonitrile
was introduced into the CVD system by bubbling nitrogen gas
(30 mLminÀ1) in acetonitrile at room temperature. The temperature
of the oven was kept at 8008C for 15 h.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 832

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Catalytic reaction
Acknowledgements
Typically, the substrate (0.5 mmol), catalyst (0.01 g), oxidant and
water (3 mL) were added in turn to a 35 mL glass reactor. After
sealing with a Teflon lid, the reactor was placed into an aluminium
block preheated at the designated temperature and kept at that
temperature for a period of time. After that, the reaction vessel
was cooled to room temperature. Cyclohexanone (0.05 g) was
added to the system as internal standard and the system was ex-
tracted with ethyl acetate (2 mLꢂ3). The organic phase was ana-
lysed by GC and GC–MS. The water phase was freeze-dried and
characterised by GC, GC–MS, FTIR spectroscopy and NMR spectros-
copy. For quantification, methanol (1 mL) was added to the solu-
tion together with 1,4-dioxane (0.010 g) as the internal standard,
after which the solution was analysed by GC.
This study was financially supported by the A*Star PSF project
(WBS: R-279-000-403-305) and a MOE Tier-1 project (WBS: R-279-
000-387-112).
Keywords: aromatic chemicals · graphene · heterogeneous
catalysis · lignin · oxidation
[1] a) G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106, 4044–4098;
b) F. M. Kerton, Y. Liu, K. W. Omari, K. Hawboldt, Green Chem. 2013, 15,
860–871; c) X. Chen, S. L. Chew, F. M. Kerton, N. Yan, Green Chem. 2014,
DOI: 10.1039/c3gc42436g; d) N. Yan, P. J. Dyson, Curr. Opin. Chem. Eng.
2013, 2, 178–183.
When birch lignin was used as the substrate, birch wood lignin
(0.05 g), catalyst (0.01 g), TBHP (12.0 mmol, 70 wt% in water) and
water (3 mL) were added in turn to a 35 mL glass reactor. After
sealing with a Teflon lid, the mixture was heated to a designated
temperature and kept at that temperature for a period of time.
After that, the reaction vessel was cooled to room temperature.
The solution was filtered and the filtrate was collected in a glass
vial (water-soluble products, fraction I). The carbon catalyst was
washed with ethanol and then isolated by filtration. The filtrate
was collected into another vial (ethanol-soluble products, fractio-
n II). Both fractions were freeze-dried and sent for further analysis,
including GC, GC–MS, GPC and FTIR spectroscopy. In the acid-cata-
lysed hydrolysis of the oxidation products of birch lignin, the sub-
strate (either fraction I or II) was dissolved in ethanol (2 mL) and
transferred into a glass tube (35 mL). Water (1 mL) and HCl
(6 mmol) were added in turn. After sealing with a Teflon lid, the re-
actor was placed into an aluminium block preheated at 1508C and
kept at that temperature for 5 h. After that, the reaction vessel was
cooled to room temperature. Cyclohexanone (0.02 g) was added to
the system as an internal standard and the system was extracted
with ethyl acetate (5 mL). The organic phase was analysed by GC
and GC–MS.
[2] a) F. G. Calvo-Flores, J. A. Dobado, ChemSusChem 2010, 3, 1227–1235;
b) P. Azadi, O. R. Inderwildi, R. Farnood, D. A. King, Renewable Sustaina-
ble Energy Rev. 2013, 21, 506–523; c) T. Q. Yuan, F. Xu, R. C. Sun, J.
Chem. Technol. Biotechnol. 2013, 88, 346–352; d) J. Zakzeski, P. C. A.
Bruijnincx, A. L. Jongerius, B. M. Weckhuysen, Chem. Rev. 2010, 110,
3552–3599; e) J. Wu, M. R. Rostami, E. S. Tzanakakis, Curr. Opin. Chem.
Eng. 2013, 2, 17–25.
[3] a) A. Toledano, L. Serrano, A. Pineda, A. A. Romero, R. Luque, J. Labidi,
Appl. Catal. B 2014, 145, 43–55; b) M. D. Garrett, S. C. Bennett, C. Harda-
cre, R. Patrick, G. N. Sheldrake, RSC Adv. 2013, 3, 21552–21557; c) Q.
Song, F. Wang, J. Cai, Y. Wang, J. Zhang, W. Yu, J. Xu, Energy Environ. Sci.
2013, 6, 994–1007; d) J. He, C. Zhao, J. A. Lercher, J. Am. Chem. Soc.
2012, 134, 20768–20775; e) N. Yan, C. Zhao, P. J. Dyson, C. Wang, L.-t.
Liu, Y. Kou, ChemSusChem 2008, 1, 626–629; f) N. Yan, Y. Yuan, R. Dyke-
man, Y. Kou, P. J. Dyson, Angew. Chem. 2010, 122, 5681–5685; Angew.
Chem. Int. Ed. 2010, 49, 5549–5553; g) J. Zhang, H. Asakura, J. v. Rijn, J.
Yang, P. Duchesne, B. Zhang, X. Chen, P. Zhang, M. Saeys, N. Yan, Green
Chem. 2014, DOI: 10.1039/c3gc42589d.
[4] a) C. Z. Li, M. Y. Zheng, A. Q. Wang, T. Zhang, Energy Environ. Sci. 2012,
5, 6383–6390; b) S. Jia, B. J. Cox, X. Guo, Z. C. Zhang, J. G. Ekerdt, Chem-
SusChem 2010, 3, 1078–1084; c) T. Yoshikawa, S. Shinohara, T. Yagi, N.
Ryumon, Y. Nakasaka, T. Tago, T. Masuda, Appl. Catal. B 2014, 146, 289–
297; d) A. Toledano, L. Serrano, J. Labidi, Fuel 2014, 116, 617–624.
[5] a) A. Azarpira, J. Ralph, F. Lu, BioEnerg. Res. 2014, 7, 78–86; b) T. Voitl, P.
Rudolf von Rohr, ChemSusChem 2008, 1, 763–769; c) H. Lange, S.
Decina, C. Crestini, Eur. Polym. J. 2013, 49, 1151–1173.
[6] a) S. Kang, X. Li, J. Fan, J. Chang, Renewable Sustainable Energy Rev.
2013, 27, 546–558; b) Y. Zhao, Q. Xu, T. Pan, Y. Zuo, Y. Fu, Q.-X. Guo,
Appl. Catal. A 2013, 467, 504–508.
Recycling experiments
[7] a) M. P. Pandey, C. S. Kim, Chem. Eng. Technol. 2011, 34, 29–41; b) F. S.
Chakar, A. J. Ragauskas, Ind. Crop. Prod. 2004, 20, 131–141; c) E. Adler,
Wood Sci. Technol. 1977, 11, 169–218.
[8] a) V. M. Roberts, V. Stein, T. Reiner, A. Lemonidou, X. Li, J. A. Lercher,
Chem. Eur. J. 2011, 17, 5939–5948; b) S. Kundu, J. Choi, D. Y. Wang, Y.
Choliy, T. J. Emge, K. Krogh-Jespersen, A. S. Goldman, J. Am. Chem. Soc.
2013, 135, 5127–5143; c) S. K. Hanson, R. T. Baker, J. C. Gordon, B. L.
Scott, D. L. Thorn, Inorg. Chem. 2010, 49, 5611–5618.
Recycling experiments were conducted at 1208C using 2-phenoxy-
1-phenylethanol (c) as the substrate. Other reaction conditions can
be found in the section “Catalytic reaction”. After reaction, the mix-
ture was extracted with ethyl acetate (2 mLꢂ3) and the organic
phase was analysed by GC and GC–MS. The solution was filtered
and the remaining solid (the catalyst) was washed with ethyl ace-
tate, ethanol and water in turn and reused for the next batch of re-
actions under the same conditions.
[9] A. Rahimi, A. Azarpira, H. Kim, J. Ralph, S. S. Stahl, J. Am. Chem. Soc.
2013, 135, 6415–6418.
[10] Y. Gao, G. Hu, J. Zhong, Z. Shi, Y. Zhu, D. S. Su, J. Wang, X. Bao, D. Ma,
Angew. Chem. 2013, 125, 2163–2167; Angew. Chem. Int. Ed. 2013, 52,
2109–2113.
[11] S. H. Lim, H. I. Elim, X. Y. Gao, A. T. S. Wee, W. Ji, J. Y. Lee, J. Lin, Phys. Rev.
B 2006, 73, 045402.
Free-radical trapping experiments
[12] B. Sedai, C. Dꢃaz-Urrutia, R. T. Baker, R. Wu, L. A. P. Silks, S. K. Hanson,
ACS Catal. 2011, 1, 794–804.
[13] E. Adler, S. Hernestam, Acta Chemica Scandinavica 1955, 9, 319–334.
[14] L. Liao, Z. Song, Y. Zhou, H. Wang, Q. Xie, H. Peng, Z. Liu, Small 2013, 9,
1348–1352.
[15] M. J. McAllister, J.-L. Li, D. H. Adamson, H. C. Schniepp, A. A. Abdala, J.
Liu, M. Herrera-Alonso, D. L. Milius, R. Car, R. K. Prud’homme, I. A. Aksay,
Chem. Mater. 2007, 19, 4396–4404.
[16] J.-X. Sun, X.-F. Sun, R.-C. Sun, P. Fowler, M. S. Baird, J. Agric. Food Chem.
2003, 51, 6719–6725.
All free-radical trapping experiments were conducted at 808C.
Other reaction conditions can be found in the section “Catalytic re-
action”. For a-O-4 type model compound conversion, TEMPO
(1 mmol) was added to the system after 5 min of reaction, after
which the reaction continued for 55 min. For b-O-4 type model
compounds, TEMPO (1 mmol) was added to the system 10 min
after the reaction commenced, after which the reaction continued
for 50 min. The reaction mixture was extracted with ethyl acetate
(2 mLꢂ3) and the organic phase was analysed by GC–MS.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 833

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[17] a) A.-P. Zhang, C.-F. Liu, R.-C. Sun, Ind. Crop. Prod. 2010, 31, 357–362;
b) M. S. Jahan, D. A. Chowdhury, M. K. Islam, S. M. Moeiz, Bioresour. Tech-
nol. 2007, 98, 465–469.
[21] L.-T. Liu, B. Zhang, J. Li, D. Ma, Y. Kou, Acta Phys. Chim. Sin. 2012, 28,
2343–2348.
[22] W. S. Hummers, R. E. Offeman, J. Am. Chem. Soc. 1958, 80, 1339–1339.
[18] R. Prado, X. Erdocia, J. Labidi, Chemosphere 2013, 91, 1355–1361.
[19] a) O. Derkacheva, D. Sukhov, Macromol. Symp. 2008, 265, 61–68; b) Y.
Sun, X. Qiu, Y. Liu, Biomass Bioenergy 2013, 55, 198–204.
[20] J. M. Nichols, L. M. Bishop, R. G. Bergman, J. A. Ellman, J. Am. Chem. Soc.
2010, 132, 12554–12555.
Received: December 31, 2013
Revised: March 3, 2014
Published online on April 1, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 825 – 834 834