Effect of Treatment Methods on Chitin Structure and Its Transformation into Nitrogen-Containing Chemicals

Article abstract of DOI:10.1002/cplu.201500326

Chitin treatment using different methods, including ball mill grinding, steam explosion, alkaline treatment, phosphoric acid, and ionic liquid (IL) dissolution/reprecipitation have been systematically investigated. The chitin structures were thoroughly investigated by using a series of analytical techniques, and the reactivity after each treatment was evaluated in dehydration and liquefaction reactions. The parallel studies enable direct comparisons of these methods and help to establish the structure-activity correlations. Ball mill grinding in dry mode was the most effective method, with the crystal size and the hydrogen-bond network being the two crucial factors in enhancing the reactivity. Remarkably, the yield of 3-acetamido-5-acetylfuran (3A5AF) from chitin dehydration increased to the highest amount (28.5%) after ball mill grinding (the previous record yield was 7.5% for untreated chitin).

Full text of DOI:10.1002/cplu.201500326

DOI: 10.1002/cplu.201500326
Full Papers
Effect of Treatment Methods on Chitin Structure and Its
Transformation into Nitrogen-Containing Chemicals
Xi Chen,[a] Yongjun Gao,[a] Lan Wang,[b] Hongzhang Chen,[b] and Ning Yan*[a]
Chitin treatment using different methods, including ball mill
grinding, steam explosion, alkaline treatment, phosphoric acid,
and ionic liquid (IL) dissolution/reprecipitation have been sys-
tematically investigated. The chitin structures were thoroughly
investigated by using a series of analytical techniques, and the
reactivity after each treatment was evaluated in dehydration
and liquefaction reactions. The parallel studies enable direct
comparisons of these methods and help to establish the struc-
ture–activity correlations. Ball mill grinding in dry mode was
the most effective method, with the crystal size and the hydro-
gen-bond network being the two crucial factors in enhancing
the reactivity. Remarkably, the yield of 3-acetamido-5-acetylfur-
an (3A5AF) from chitin dehydration increased to the highest
amount (28.5%) after ball mill grinding (the previous record
yield was 7.5% for untreated chitin).
Introduction
With increasing concern over fossil fuel depletion, significant
efforts have been paid to biomass resources for the sustaina-
ble production of biofuels and biochemicals.[1] To achieve an
economically feasible method of biorefinery, it is essential to
build diversified, high-value bioproduct chains. Currently, bio-
mass utilization is largely focused on woody biomass materials
such as cellulose and lignin.[2] Nevertheless, it is highly desira-
ble and beneficial to explore new types of biomass resources
to complement and expand the current biorefinery scheme.[3]
Chitin, which is the world’s most abundant amino-biopolymer
with 7 wt% biologically fixed nitrogen, represents a promising
raw material to produce value-added biochemicals, especially
nitrogen-containing (N-containing) compounds that are not
readily available from lignocellulosic biomass. The industrial
production of chitin is conducted by extraction of crab and
shrimp shells, which are shellfish waste in the fishing industry.
As such, chitin valorization not only has scientific importance
and potential economic value, but also environmental benefits.
The structure of chitin is similar to that of cellulose except
for the side chain at the C-2 position (see Scheme 1). Chitin is
a linear polymer of b(1!4)-linked 2-acetamido-2-deoxy-d-glu-
copyranose. The side chain at the C-2 position is either an
Scheme 1. The chemical structure of chitin polymer.
acetyl amide group or an amine group. Chitin is highly crystal-
lized with extensive hydrogen-bond networks among the poly-
mer chains, thus making it even more difficult to be dissolved
and converted than cellulose. So far, only a handful of studies
have been reported that dealt with chitin conversion into
chemicals.[4] The non-N-containing chemicals levulinic acid
(LA)[5] and 5-hydroxymethylfurfural (5-HMF)[4b] were obtained in
water with metal salt additive or concentrated ZnCl2 solution
with 11.5 and 9% yield, respectively. Direct chitin dehydration
and liquefaction into N-containing compounds were reported
very recently.[6] The N-containing furan derivative, 3-acetamido-
5-acetylfuran (3A5AF), was produced by chitin dehydration in
organic solvent or IL[6a,c] with a maximum yield of 7.5%. In
these studies, the yields of target products were relatively low.
In chitin dehydration there was no further increase in 3A5AF
by employing high temperatures and prolonged reaction
times. It indicates that the traditional technologies in the pe-
troleum refinery that are defined by high-temperature process-
es are inherently not suitable for biorefinery because of the
huge structural differences between fossil fuels and biomass
materials. Chitin is a highly rigid and functionalized polymeric
material, and severe side reactions will happen under harsh
conditions. The direct production of chitin into high-yield N-
containing products is still challenging.
[a] X. Chen, Dr. Y. Gao, Prof. N. Yan
Department of Chemical and Biomolecular Engineering
National University of Singapore
4 Engineering Drive 4, Singapore 117585 (Singapore)
E-mail: ning.yan@nus.edu.sg
[b] Dr. L. Wang, Prof. H. Chen
Institute of Processing Engineering
Chinese Academy of Sciences
1 Zhongguancun North 2nd Alley, Haidian, Beijing 100190 (P. R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201500326.
As far as we are concerned, the greatest challenge in bio-
mass conversion often arises from the high crystallinity and rig-
idly fixed polymer chains. Logically, a plausible way to enhance
Part of a Special Issue to celebrate Singapore’s Golden Jubilee. To view
the complete issue, visit: http://dx.doi.org/10.1002/cplu.v80.8.
ChemPlusChem 2015, 80, 1565 – 1572
1565
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
the reactivity of chitin is to damage the structure robustness
and create more flexible polymer chains prior to chemical
transformation. The advances in treatment methods[7] (e.g.,
ball mill, steam explosion, acid/alkaline impregnation, and
ionic liquid (IL) treatment) to process biomass materials have
been widely reported. These hold promise in applications to
enhance chitin-based refinery into N-containing chemicals. In
this study, a systematic comparative study of five different
methods on the structure change and subsequent chemical re-
activity of chitin biomass were conducted, which enables hori-
zontal analysis to identify the most effective method as well as
the key factors responsible for chitin conversion. After different
treatments, the native chitin and treated samples were com-
prehensively characterized by various techniques such as scan-
ning electron microscopy (SEM), Brunauer–Emmett–Teller (BET),
X-ray powder diffraction (XRD), Fourier transform infrared spec-
troscopy (FTIR), nuclear magnetic resonance (NMR) spectrosco-
py, and gel permeation chromatography (GPC). Subsequently,
reactivity was studied by investigating two different chemical
reactions including chitin dehydration and liquefaction. On the
basis of the structure changes and reaction performances, the
major structure factors associated with enhanced product
yields were summarized, and the relative effectiveness of vari-
ous treatment methods was compared. The results indicate
that different treatment methods endow chitin with substan-
tially different structure changes and reactivity. With the
proper method, chitin can be converted with much higher effi-
ciency under milder reaction conditions.
Results and Discussion
Abbreviations were given to the samples treated by different
methods. Samples processed by ball mill grinding, steam ex-
plosion, alkaline treatment, phosphoric acid, and ionic liquid
(IL) dissolution/reprecipitation are denoted as BM, SE, Acid,
Base, and IL samples, respectively. Note that there are two
modes (dry and wet) for the ball-mill method, thus resulting in
two BM samples. BM-1 refers to the sample processed in the
dry mode, and BM-2 refers to the sample obtained in the wet
mode. As a result, in the following part, the structure changes
of the six treated samples are discussed relative to the untreat-
ed chitin. The details of treatment procedures are provided in
the Experimental Section.
Figure 1. (a) XRD analysis of native and treated chitin samples. (b) FTIR anal-
ysis of native and treated chitin samples. (c) Solid-state 13C NMR spectra of
the native and BM dry chitin.
Analysis of structure changes using XRD, FTIR, and NMR
spectroscopy
a slight decrease (about a 10% drop) in the CI value was ob-
served for Acid and Base samples, whereas a negligible change
was noticed for other samples. Note that BM-2 is much less af-
fected than BM-1, which suggests that the treatment is very
sensitive to ball-mill modes. It is plausible that the presence of
water mitigates the contact attrition and thus decreases the
grinding efficiency. The decrease in the CI values of the Acid,
Base, and BM-1 samples was reported in previous studies with
cellulose/chitin,[8] whereas the result in our study suggested
ball milling to be significantly more efficient than other meth-
ods. The IL sample displayed a negligible change in crystal
structure, which was different from cellulose treatment by
XRD and FTIR analyses were conducted on the native chitin
and treated samples (as shown in Figure 1), and the calculated
crystalline index (CI) parameters are shown in Table 1 (an over-
all summary). The results of ball-mill-treated chitin and other
samples were compiled for comparison. In the XRD pattern of
native chitin, the major peak at 2q=198 represents the (110)
plane of crystalline chitin. Native chitin is highly crystalline
with a CI value of 91%. In Figure 1a, it is clear that the BM-
1 sample displayed significantly decreased crystallinity, and the
CI value was reduced strikingly from 91 to 28%. However, only
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1566
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
chemical shift of C-1 possibly indicated that partial
cleavage of chains occurred and more reducing ends
have been exposed.
Table 1. Overall summary of the structure of native and treated chitin.
Chitin
XRD
GPC
BET
FTIR
SEM
CI [%] MW [kDa] Surface area [m2 gÀ1
]
Hydrogen-bond Particle size
networks
Optimization of ball-milling parameters
native
BM-1
BM-2
SE
Acid
Base
IL
91
28
89
93
82
83
91
–
flfl
fl
7.4
2.3
6.8
6.2
5.2
1.9
6.2
–
–
decrease
unchanged
unchanged
decrease
decrease
unchanged
decrease
unchanged
unchanged
decrease
increase
unchanged
The BM-1 sample exhibited the most dramatic de-
crease in XRD signals. To further understand the ball-
milling process, the number of balls, the feedstock
mass, the ball-milling time, and speed were varied.
The CI value decreased clearly as numbers of balls in-
creased from 10 to 160 (see Figure S1 in the Support-
ing Information). Furthermore, the increase in sub-
strate loading reduced the milling efficiency, thus
fl
flfl
!
!
means of IL dissolution/reprecipitation. Significantly decreased
crystallinity was often observed[9] in the cellulose structure. For
chitin dissolution in ILs, contrary results in XRD analysis were
reported. Chitin regenerated from [BMim]Ac (BMim=1-butyl-3-
methylimidazolium chloride) was reported to show greatly de-
creased XRD signals, thus indicating the damage of crystalline
regions after regeneration.[10] In contrast, chitin regenerated
from [BMim]Cl and [AMim]Br exhibited strong XRD signals that
were comparable to that of the original chitin.[11] Our results
were in agreement with the latter reports. It was assumed that
regeneration of the crystalline region readily occurs upon pre-
cipitation and drying. As such, it appears that the mechanism
behind cellulose and chitin treated by IL is different, despite
their similar chemical structure.
leading to less decrease in crystallinity (see Figure S2 in the
Supporting Information). With an increase in time from one
half to two hours, the CI value decreased accordingly, whereas
insignificant changes were found when the ball-milling time
was increased further (see Figure S3 in the Supporting Informa-
tion). To ensure an efficient breakdown of the crystalline
region, a relatively high milling speed is needed. At 300 rpm,
the CI value showed a negligible decrease. Nevertheless, the
decrease became considerable when the speed increased to
600 and 650 rpm (see Figure S4 in the Supporting Informa-
tion). These results indicated that ball-milling parameters have
remarkable influences on chitin treatment and that operational
parameters should be carefully selected in future applications.
We also followed the temperature change after ball milling. In
the presence of 80 balls at 650 rpm, the temperature was 40.6,
50.5, and 50.88C, respectively after one half, one, and three
hours of milling. In the presence of 160 balls at 650 rpm, the
temperature increased to 538C after half an hour. The moder-
ate temperature increase during ball milling might play a role
in breaking down the interchain hydrogen-bond network in
chitin.
The FTIR spectra are shown in Figure 1b. Overall, the chemi-
cal backbones of the samples were maintained after all treat-
ments. The detailed assignments of peaks for chitin have been
illustrated previously.[12] The bands that range from 3200 to
3500 cmÀ1 are ascribed to the inter- and intrahydrogen bonds
between the hydroxyl groups. Meanwhile, the peaks located at
1600–1700 cmÀ1 are assigned to the amide I band (two types
of hydrogen bonds in a C=O group with the NH group of the
adjacent chain and the OH group of the interchain). The bands
assigned to these types of hydrogen bonding became appa-
rently broader and smoother for the BM-1 sample, which sug-
gests that the hydrogen-bond networks among chitin polymer
molecules have been remarkably impaired.[13] Such changes
have also been found in the Acid and Base samples but to
a lesser extent. Negligible changes were observed for the rest
of the samples. The FTIR results are in perfect agreement with
the XRD data. Ball milling is by far the most effective method
to destroy the crystalline regions and the networks of the poly-
mer chains in chitin. As the BM-1 sample showed the most
prominent decrease in XRD signals, it has been further com-
pared to native chitin by means of solid-state 13C NMR spectro-
scopic techniques (Figure 1c). Peak broadening was observed
in the BM-1 chitin sample, thus indicating a significant de-
crease in crystallinity, which is in agreement with the XRD and
FTIR results. Furthermore, the peak for C-1 shifted from d=104
to 102 ppm in the ball-mill-treated chitin, whereas there is no
appreciable chemical shift for other peaks. The C-1 position for
N-acetyl-d-glucosamine (NAG, chitin monomer) and chitin was
at d=95 and 104 ppm, respectively, in the literature.[14] The
Analysis of structure changes using GPC, SEM, and BET
The molecular weights (MWs) were analyzed by using GPC (see
Figure S5 in the Supporting Information) and calculated (see
Table S1 in the Supporting Information). Both the native and
processed samples showed wide MW distributions (high poly-
dispersity). Since all the samples were dissolved and detected
under identical conditions, the changes in MW after treatment
could be observed and compared. The Base and IL samples
had MWs comparable to that of native chitin; SE and BM-2
samples exhibited moderately decreased MW, whereas BM-1
and Acid samples displayed considerably decreased MW. The
Base and IL treatments led to samples with the least reduced
MW owing to the mild process conditions. The decrease in
MW is not unexpected for SE and Acid samples and is consis-
tent with previous studies on cellulose.[15] Steam explosion was
conducted with water under elevated temperature and pres-
sure, which caused a moderate decrease. The large decrease in
that of the Acid sample was plausible owing to the fact that
chitin polymers could be chemically attacked by acid mole-
cules during the process. Ball milling also seems to be effective
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1567
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
in cleavage of chitin polymer chain, plausibly owing to the
“hot-point” effect induced by frictions and impacts.[16]
samples, only crowded granules were observed without long-
chain fibers. The results showed that the processes modified
the chitin particle size and the chitin fiber morphology. BET
analysis was conducted to confirm whether pore structure was
generated after treatment. The nitrogen sorption isotherms of
the samples are shown in Figure S8 in the Supporting Informa-
tion. The BET surface areas and pore volumes were small for all
samples. Unexpectedly, these values of the treated samples
showed a general decrease relative to the native chitin. There-
fore, the treatment did not induce the generation of a porous
structure in chitin.
SEM images were first taken at a low magnification of 25”
(see Figure S6 in the Supporting Information). It was observed
that the particle size of the BM-1 and Acid samples decreased
remarkably, a phenomenon that was not observed in other
samples.
At high magnification, the treated samples displayed a dis-
tinctive morphology: Magnification images of the samples at
3000 and 50000” are shown in Figure S7 in the Supporting In-
formation and Figure 2. At 3000”, SE and IL samples showed
an uneven surface similar to native chitin as bulky flakes. Nev-
ertheless, the BM-1, Acid, and Base samples appeared as much
smaller flakes. In addition, it seems that pores were generated
on the surface of flakes of BM-2 chitin. With the higher mag-
nification of 50000, the fibers and adherent granules could be
observed. BM-1 and BM-2 chitin samples showed the existence
of chitin fibers but fewer granules. However, IL chitin exhibited
distinguished swollen and wider chitin fibers, which was
caused by the dissolution process. The SE sample showed
highly arranged thinner fibers and granules. For Acid and Base
The overall structure changes of different samples are sum-
marized in Table 1. The BM-1 sample exhibited the most promi-
nent decrease in crystallinity. At the same time, the hydrogen-
bond networks and the MW of the BM-1 sample were remarka-
bly reduced as well, whereas the BM-2 sample was less affect-
ed. Acid and Base samples showed decreased crystallinity, hy-
drogen-bond networks, and molecular weights. The particle
size of the Acid sample was also reduced. Relative to these
samples, less change in the structure was observed for the SE
and IL samples.
Transformation performance: Dehydration of native chitin
and treated samples
A dehydration reaction was conducted that employed all sam-
ples as substrates. Both N-methyl-2-pyrrolidone (NMP) and
[BMim]Cl were used as the solvent for chitin depolymerization-
dehydration into 3A5AF (shown in Figure 3a). Overall, the IL
solvent was superior to the NMP solvent for the transforma-
tion. In NMP, the 3A5AF yield of native chitin was 4.6% under
the employed conditions. The BM-1 and Acid samples afforded
a higher yield of 3A5AF (10.6 and 5.3%, respectively). The SE
sample showed a comparable yield (4.8%), whereas the rest
showed decreased yields. However, all treated samples exhibit-
ed enhanced or at least comparable product yields in [BMim]Cl
solvent. Acid and Base samples provided a more than twofold
increase in 3A5AF yield; BM-1 chitin achieved the highest yield
of 20.2%. Thus, the BM-1 sample was further studied under
different conditions (Figure 3b). Rapid initial reaction rates
were observed and peak values were reached within just ten
minutes. Increasing the temperature to 1808C while maintain-
ing the reaction time at ten minutes provided the highest
3A5AF yield. Even higher temperatures were not investigated
because the decomposition of [BMim]Cl solvent could occur.
By means of dry ball milling, the yield of the target product in
chitin dehydration was remarkably improved from the previous
record of 7.5 to 28.5% (the highest value obtained), with a de-
creased reaction time from 60 to 10 minutes. Column chroma-
tography was used to separate the product from IL solvent
with an isolated yield of around 20%. For scalable production,
other feasible separation methods should be considered. It has
been reported that 3A5AF can be extracted from IL by using
ethyl acetate.[4c] This could be a first separation step.
Figure 2. SEM images (”50000) of (a) native chitin; (b) BM-1; (c) BM-2;
(d) SE; (e)Acid; (f) Base; and (g) IL.
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1568
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Figure 4. The kinetic profile of liquefaction of BM-1 chitin. Reaction condi-
tions: 1 g EG, 0.08 g sulfuric acid (8% w/w EG), and 0.15 g substrate (15%
w/w EG) at 1508C for 0, 5, 10, 20, 30, and 40 min.
samples afforded more target products of hydroxyethyl-2-
amino-2-deoxyhexopyranoside (HADP) and hydroxyethyl-2-
acetamido-2-deoxyhexopyranoside (HAADP) than the rest of
the samples. The BM-1 sample still afforded the highest total
yield of HADP and HAADP (32.3%) with 88.7% liquefaction
yield, under much milder conditions than previously repor-
ted.[6b] The yields of formed target products follow this order:
BM-1>Base>Acid>IL>BM-2>native chitinꢀSE.
Figure 3. (a) Dehydration of chitin and treated samples. Blue columns in
NMP solvent: 2158C, NMP (3 mL), substrate (100 mg), boric acid (400 mol%),
LiCl (200 mol%), HCl (100 mol%), 1 h; red columns in IL solvent: 1808C,
[BMim]Cl (1 g), substrate (80 mg), boric acid (400 mol%), HCl (100 mol%),
1 h. (b) Studies of BM-1 sample in IL solvent under different conditions:
[BMim]Cl (1 g), substrate (80 mg), boric acid (400 mol%), HCl (100 mol%), re-
action for 5, 10, 30, and 60 min at 160, 170, and 1808C, respectively.
Since the BM-1 sample had the best performance, the kinet-
ic profile was investigated further (see Figure 4). The liquefac-
tion reaction quickly reached about 50% chitin conversion
within five minutes, and then gradually increased to 88% at
20 minutes. A slight decrease (about 6%) in liquefaction effi-
ciency was observed at 30 and 40 minutes, which might indi-
cate the formation of insoluble char. The chitin monomer was
detected during the liquefaction, thus suggesting that the hy-
drolysis was the first step along the reaction pathway. The
yield was about 7% at five minutes and then declined con-
stantly to a yield of 0.2% at 40 minutes. Similarly, the liquefied
product HAADP peaked at five minutes (9.6% yield) and then
decreased to 4% yield. The major product HADP had a steady
increase in the initial 20 minutes and reached a plateau after-
wards. A similar trend was also observed for acetic acid (HAc),
the yield of which decreased after 20 minutes. Relative to
native chitin, the BM-1 sample showed a faster reaction rate,
and the peak values were achieved within 20 minutes instead
of 60 minutes under lower temperature.
Table 2. Liquefaction results of chitin and treated samples.
Entry[a]
HADP [%]
HAADP [%]
Conversion [%]
native chitin
BM-1
BM-2
SE
Acid
Base
6.6
28.0
7.9
0.9
4.3
0.9
0.7
3.2
4.2
2.6
28.0
88.7
13.5
27.9
48.2
67.2
43.5
6.7
10.2
18.5
7.8
IL
[a] Reaction conditions: EG (1 g), sulfuric acid (0.08 g; 8% w/w EG), and
substrate (0.15 g; 15% w/w EG) at 1508C for 20 min.
In both dehydration and liquefaction reactions, the BM-1
sample showed the best performance. The Acid and Base sam-
ples also exhibited an appreciable increase in product yields.
These results, together with the structure characterizations,
provide the opportunity to identify the major factors responsi-
ble for enhanced reactivity. Considering the apparent decrease
in crystal size and hydrogen-bond networks of BM-1, Acid, and
Transformation performance: Liquefaction of native chitin
and treated samples
The native and treated chitin samples were also tested in the
liquefaction reaction in ethylene glycol (EG) over sulfuric acid
at 1508C. The results were mostly congruous with those in the
dehydration reaction (see Table 2). The BM-1, Base, and Acid
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1569
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
10 mm were used. BM-2 was ground with a PL11 PFI Beater (PFI
Mill) at a speed of 6000 rpm in the presence of 10% water (wet
mode). Further optimization of dry ball-milling parameters was
conducted with a Pulverisette P7 Premium Line (Fritsch) with
a chamber (ZrO2) volume of 45 mL with 10 min milling and 5 min
rest as a cycle. The balls (ZrO2) had a diameter of 5 mm. The tem-
perature after the ball-milling process was recorded for selected
optimization experiments. A thermometer was inserted into the
solids and detected the temperature after milling.
Base samples, these two properties are highly likely to be the
essential factors in improving reaction activity. Surface area,
however, was not a crucial factor. Molecular weight and parti-
cle size might be related to the reaction performance, but this
is not conclusive. Interestingly, IL chitin exhibited moderately
increased product yields in some cases, despite the only struc-
ture change of IL chitin being the morphology change. The
SEM image suggested that the chitin fibers swelled after IL
treatment, thus indicating that the swelling of chitin polymer
is beneficial, to some extent, for the chemical transformation.
Steam explosion was carried out in a 5 L batch reactor (Weihai Au-
tomatic Control Reactor Ltd., China). About 200 g of the dried
sample were fed into the reactor. Steam was charged into the reac-
tor at 1.5 MPa at 2008C for 3 min and then released. The resulting
chitin contained 50–70% water content because the steam con-
densed. After the process, the material was dried in a forced-air
oven at 608C for 24 hours (denoted as the SE sample).
Conclusion
This study describes a systematic comparison study of the in-
fluences of five treatment methods on chitin structure and
chemical reactivity. Different methods exhibited varied effects
in changing the structural parameters. We observed a strong
correlation between the crystal size/hydrogen-bonding net-
work intensity and the reactivity of chitin. A treatment method
that is effective in decreasing the crystal size and hydrogen-
bonding network will consequently lead to a considerable in-
crease of chitin reactivity in the subsequent transformations.
The dehydration yield was increased from the previous 7.5 to
28.5%. A feasible separation method is needed to separate the
product from IL solvent, for example, first by extraction and
then further purification. Our study demonstrated that chitin is
indeed a promising material for the direct production of N-
containing chemicals with high yields after damaging its crys-
tallinity by a proper treatment method.
Protocol for acid, base, and IL treatment
H3PO4 dissolution/reprecipitation was conducted in the following
manner. Chitin (35 mg) was added into an H3PO4 solution (1 g,
68 wt%) in a thick-walled glass tube and heated at 608C for 1 h
with a stirring speed of 700 rpm. Then 10 wt% NaOH was used to
neutralize the solution, during which chitin precipitated. The white
solid was washed three times with deionized water and dried in
oven at 708C overnight (denoted as the Acid sample).
The dissolution of chitin in base solution was reported in a previous
study,[17] and this process was used for alkaline treatment of chitin.
Briefly, chitin (600 mg) was dispersed in an aqueous solution (total
30 g) of NaOH (8 wt%) and urea (4 wt%) in a plastic centrifuge
tube. The suspension was kept at À208C and stirred twice over
48 h. The solution was centrifuged, and the liquid supernatant was
collected. Then the supernatant was neutralized using H2SO4
(10 wt%) aqueous solution. Regenerated chitin was precipitated,
washed with deionized water until the pH reached 7, and then
dried in an oven at 708C overnight (denoted as the Base sample).
Experimental Section
Materials and chemicals
Chitin was purchased from Wako Pure Chemical Industry. Boric
acid (ACS grade) was purchased from Amresco. Sodium chloride
(NaCl, AR grade) was purchased from Schedelco. Lithium chloride
(LiCl, 98%) and dimethylacetamide (DMAc, anhydrous, 99.8%)
were purchased from Alfa Aesar. Phosphoric acid (H3PO4, 85%) and
ethylene glycol (EG, >99%) were purchased from VWR Singapore.
Sodium hydroxide (NaOH, 99%), acetic acid (HAc, 99%), and hydro-
chloric acid (HCl, 37%) were purchased from Merck. Urea (ACS
grade), hexamethyldisilazane (HMDS, 99%), trifluoreacetic acid
(99%), pyridine (AR), N-acetyl-d-glucosamine (NAG, 99%), and N-
methyl-2-pyrrolidone (NMP, anhydrous, 99.5%) were purchased
from Sigma–Aldrich. Sulfuric acid (H2SO4, 97%) was purchased
from J.T. Baker. 1-Butyl-3-methylimidazolium chloride ([BMim]Cl,
99%) and 1-allyl-3-methylimidazolium bromide ([AMim]Br, 99%)
were purchased from the Lanzhou Institute of Chemical Physics. All
chemicals were used as received.
IL dissolution/reprecipitation was conducted in a similar manner.
Chitin (35 mg) was dissolved in [AMim]Br (1.5 g) at 110 8C for 1 h in
a thick-walled glass tube with a stirring speed of 600 rpm. After-
wards it was cooled to room temperature, and deionized water
(20 mL) was added to precipitate the chitin. The solid was collect-
ed by centrifugation, washed three times with deionized water,
and then dried in an oven at 708C overnight (denoted as the IL
sample).
XRD and solid-state 13C NMR spectroscopic analysis
XRD analysis was performed with a Bruker D8 Advance diffractom-
eter with CuKa radiation at 40 kV. The scan range was from 5 to 408
without rotation. The equation for the CI calculation is shown
below [Eq. (1)].[18]
Protocol for ball milling and steam explosion treatment
CI ½%Š ¼ ðI110ÀIamÞ=I110  100 %
ð1Þ
Ball-milling samples were obtained under different grinding condi-
tions, including dry and wet modes (denoted as BM-1 and BM-2,
respectively). BM-1 chitin was ground in dry mode with a planetary
ball mill (PM 100 CM; Retsch) with a chamber volume of 125 mL at
650 rpm and 4 h total mill time (20 min mill and 10 min rest as
a cycle). The material of the chamber and balls was zirconium
oxide (ZrO2). The feed quantity was 1 g, and 45 balls with diameter
in which I110 is the maximum intensity of the diffraction for the
(110) plane at approximately 2q=19.28 and Iam is the intensity of
the amorphous diffraction at approximately 2q=12.78.
Native chitin and BM dry chitin samples were sent for solid-state
13C NMR spectroscopic analysis for further comparison. Solid-state
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1570
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
13C NMR spectroscopy was conducted with a Bruker Advance 400
(DRX400) with cross-polarization/magic-angle spinning (CP/MAS).
fluoreacetic acid (TFA, 60 mL), and a small stirring bar were added.
The closed vial was put in a water bath at 608C for 1 h. After silyla-
tion, a portion of the reaction mixture (250 mL) and diethyl ether
(750 mL) were mixed together and analyzed.
FTIR and GPC analysis
FTIR was conducted with a Bio-Rad FTS-3500 ARX instrument.
Transmission mode was used and the results were collected under
N2 flow. GPC analysis was carried out with a system equipped with
a Waters 2410 refractive index detector, a Waters 515 high-per-
formance liquid chromatography (HPLC) pump, and two Waters
styragel columns (HT3 and HT4) using 5 wt% LiCl/DMAc as eluent
at a flow rate of 0.5 mLminÀ1 at room temperature. The samples
were dissolved in the mobile phase solution at 808C with rigorous
stirring for 2 days under nitrogen atmosphere prior to analysis. The
raw data were processed using narrow polydispersity polystyrene
standards and calibration using the software Breeze (the calibra-
tion curve is shown in Figure S6 of the Supporting Information).
Note that the treated samples must be washed thoroughly to
remove any acid residues before dissolution. Otherwise, the acid
residue would cause decomposition during dissolution and result
in much reduced MW values.
Dehydration of chitin into 3A5AF
Two different solvent systems were attempted for the reaction, in-
cluding NMP as solvent and IL as solvent. The procedures for chitin
dehydration in IL solvent were carried out:[6c] Substrate (80 mg,
0.4 mmol based on the NAG monomer) was placed in a thick-
walled glass tube (35 mL). Following that, a magnetic stir bar, boric
acid (98 mg, 1.6 mmol), HCl (0.4 mmol), and [BMim]Cl (1 g) were
added. In the IL solvent system, HCl instead of NaCl was used to-
gether with boric acid as additive because it gave the best results
in additive screening. The tube was sealed by a Teflon stopper
(with an O ring) and placed into a preheated oil bath at a desired
temperature for a certain time at a stirring speed of 700 rpm. The
experiments in NMP solvent were conducted in a similar way but
with different additives. For product identification and quantifica-
tion, procedures are provided elsewhere.[6a] The yield of 3A5AF was
calculated as follows: product [mol]/starting chitin [mol]”100%.
Since the yield was enhanced significantly in IL solvent after ball-
milling treatment, column chromatography was conducted for the
isolated yield of the 3A5AF product. Silica gel with particle sizes
that ranged from 40 to 63 mm was used as the stationary phase,
and 5% methanol and 95% dichloromethane were used as the
mobile phase; ball-milled chitin (153 mg) was used as the substrate
in IL (1.5 g). After reaction, the solution was reheated to 808C to
melt the IL and transfer it to the column, whereas the solution was
still too viscous to be fully transferred and some of it adhered to
the tube wall. After column separation, a yellow-brown solid
(25 mg) was obtained, and the solid (ꢀ3 mg) was dissolved in
[D6]DMSO for NMR spectroscopic analysis (see Figure S10 in the
Supporting Information). The isolated yield was about 20%.
SEM and BET analysis
SEM images were obtained with a JEM-6700F scanning electron
microscope (JEOL). The sample was immobilized on a copper sub-
strate by conductive adhesives and was coated by platinum before
analysis by means of high vacuum evaporation. Gas sorption iso-
therms were measured using a Micrometrics ASAP 2020 surface
area and pore-size analyzer. Prior to the measurement, the samples
were outgassed at 1008C for at least 6 h. The BET specific surface
areas were calculated using the adsorption data in the relative
pressure (P/P0) range of 0.05–0.25. The total pore volumes were es-
timated from the amount adsorbed at a relative pressure P/P0 of
0.98.
Liquefaction of chitin into monosaccharides
HPLC analysis
In a general procedure for chitin liquefaction, chitin or the treated
sample (0.15 g, 0.75 mmol) was added with EG solvent (1 g) and
8 wt% H2SO4 acid catalyst (80 mg). The mixtures were stirred at
420 rpm and heated for the desired time at 1508C. After the reac-
tion, the mixtures were cooled and neutralized to pH 7 by using
10 wt% NaOH. Next, the reaction mixture was separated into three
(denoted as A, B, and C) fractions, and the major products were
separated by HPLC and confirmed by NMR spectroscopy.[6b] Fractio-
n A contained unreacted chitin and was used to determine the
conversion. Fractions B and C contained products and were silylat-
ed for GC analysis. The yield obtained is based on mole percent-
age: product [mol]/starting chitin [mol]”100%.
HPLC analysis of dehydration product was performed with an Agi-
lent 1200 Series (Agilent Technologies, Germany) LC system by
using an Agilent ZORBAX Eclipse carbon-18 column. The mobile
phase was 83% water and 17% acetonitrile. The flow rate was
kept at 0.5 mLminÀ1 with a run time of 20 min. A UV/Vis detector
setting at 230 nm was used to analyze the product. Other products
such as HAc and NAG were quantified by HPLC by plotting exter-
nal calibration curves with standard chemical solutions. HAc was
analyzed with an Agilent Hi-Plex H column with 100% H2SO4 solu-
tion (0.005m) as the mobile phase, and NAG was analyzed with an
Agilent Hi-Plex Ca column with 100% water as the mobile phase.
The flow rates were at 0.6 mLminÀ1, and the UV/Vis detector was
set at 210 nm for HAc and 195 nm for NAG.
Acknowledgements
Gas chromatography (GC) analysis
We thank Prof. Xindong Mu (Qingdao Institute of Bioenergy and
Bioprocess Technology, Chinese Academy of Sciences) for support-
ing this study. This study was financially supported by A*Star, PSF
funding with WBS: R-279-000-403-305.
The major liquefied products were silylated and quantified by
using dodecane as the internal standard. The GC analysis was per-
formed with an Agilent 7890 A gas chromatograph with a flame
ionization detector (FID) equipped with HP-5 capillary column
(30 m”250 mm). The yield was calculated by using the effective
carbon numbers.[19] A general procedure for silylation was as fol-
lows.[20] In an analytical vial (1 mL), the product mixture to be ana-
lyzed (40 mg), pyridine (700 mL), hexamethyldisilazane (700 mL), tri-
Keywords: biomass
· chitin · homogeneous catalysis ·
structure–activity relationships · sustainable chemistry
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1571
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[1] a) D. R. Dodds, R. A. Gross, Science 2007, 318, 1250–1251; b) H. Kobaya-
shi, A. Fukuoka, Green Chem. 2013, 15, 1740–1763; c) E. L. S. Ngee, Y.
Gao, X. Chen, T. M. Lee, Z. Hu, D. Zhao, N. Yan, Ind. Eng. Chem. Res.
2014, 53, 14225–14233; d) H. Ben, A. J. Ragauskas, ACS Sustainable
Chem. Eng. 2013, 1, 316–324; e) F. Jin, H. Enomoto, Energy Environ. Sci.
2011, 4, 382–397; f) S. Dutta, S. De, B. Saha, ChemPlusChem 2012, 77,
259–272.
[2] a) J. Zhang, J. Teo, X. Chen, H. Asakura, T. Tanaka, K. Teramura, N. Yan,
ACS Catal. 2014, 4, 1574–1583; b) S. Siankevich, G. Savoglidis, Z. Fei, G.
Laurenczy, D. T. L. Alexander, N. Yan, P. J. Dyson, J. Catal. 2014, 315, 67–
74; c) J. Zhang, H. Asakura, J. v. Rijn, J. Yang, P. Duchesne, B. Zhang, X.
Chen, P. Zhang, M. Saeys, N. Yan, Green Chem. 2014, 16, 2432–2437;
d) S. Siankevich, Z. Fei, R. Scopelliti, G. Laurenczy, S. Katsyuba, N. Yan,
P. J. Dyson, ChemSusChem 2014, 7, 1647–1654; e) Y. Gao, J. Zhang, X.
Chen, D. Ma, N. Yan, ChemPlusChem 2014, 79, 825–834.
[8] a) G. Cheng, P. Varanasi, C. Li, H. Liu, Y. B. Melnichenko, B. A. Simmons,
M. S. Kent, S. Singh, Biomacromolecules 2011, 12, 933–941; b) T. Kobaya-
shi, B. Kohn, L. Holmes, R. Faulkner, M. Davis, G. E. Maciel, Energ. Fuel.
2011, 25, 1790–1797.
[9] J. Zhang, Y. Wang, L. Zhang, R. Zhang, G. Liu, G. Cheng, Bioresour. Tech-
nol. 2014, 151, 402–405.
[10] Y. Wu, T. Sasaki, S. Irie, K. Sakurai, Polymer 2008, 49, 2321–2327.
[11] a) H. Xie, S. Zhang, S. Li, Green Chem. 2006, 8, 630–633; b) K. Prasad,
M.-A. Murakami, Y. Kaneko, A. Takada, Y. Nakamura, J.-i. Kadokawa, Int. J.
Biol. Macromol. 2009, 45, 221–225.
[12] F. G. Pearson, R. H. Marchessault, C. Y. Liang, J. Polym. Sci. 1960, 43,
101–116.
[13] a) M. Osada, C. Miura, Y. S. Nakagawa, M. Kaihara, M. Nikaido, K. Totani,
Carbohydr. Polym. 2012, 88, 308–312; b) M. Osada, C. Miura, Y. S. Naka-
gawa, M. Kaihara, M. Nikaido, K. Totani, Carbohydr. Polym. 2013, 92,
1573–1578.
[14] S. F. Tanner, H. Chanzy, M. Vincendon, J. C. Roux, F. Gaill, Macromolecules
1990, 23, 3576–3583.
[15] J. Fernµndez-BolaÇos, B. Felizón, A. Heredia, R. Rodríguez, R. GuillØn, A.
JimØnez, Bioresour. Technol. 2001, 79, 53–61.
[16] S. M. Hick, C. Griebel, D. T. Restrepo, J. H. Truitt, E. J. Buker, C. Bylda,
R. G. Blair, Green Chem. 2010, 12, 468–474.
[3] a) X. Chen, N. Yan, Catal. Surv. Asia 2014, 1–13; b) F. M. Kerton, Y. Liu,
K. W. Omari, K. Hawboldt, Green Chem. 2013, 15, 860–871.
[4] a) M. Mascal, E. B. Nikitin, ChemSusChem 2009, 2, 859–861; b) Y. Wang,
C. M. Pedersen, T. Deng, Y. Qiao, X. Hou, Bioresour. Technol. 2013, 143,
384–390; c) M. W. Drover, K. W. Omari, J. N. Murphy, F. M. Kerton, RSC
Adv. 2012, 2, 4642–4644; d) M. Osada, K. Kikuta, K. Yoshida, K. Totani,
M. Ogata, T. Usui, Green Chem. 2013, 15, 2960–2966; e) F. D. Bobbink, J.
Zhang, Y. Pierson, X. Chen, N. Yan, Green Chem. 2015, 17, 1024–1031;
f) K. W. Omari, L. Dodot, F. M. Kerton, ChemSusChem 2012, 5, 1767–
1772.
[5] K. W. Omari, J. E. Besaw, F. M. Kerton, Green Chem. 2012, 14, 1480–1487.
[6] a) X. Chen, S. L. Chew, F. M. Kerton, N. Yan, Green Chem. 2014, 16, 2204–
2212; b) Y. Pierson, X. Chen, F. D. Bobbink, J. Zhang, N. Yan, ACS Sustain-
able Chem. Eng. 2014, 2, 2081–2089; c) X. Chen, Y. Liu, F. M. Kerton, N.
Yan, RSC Adv. 2015, 5, 20073–20080.
[17] X. Hu, Y. Du, Y. Tang, Q. Wang, T. Feng, J. Yang, J. F. Kennedy, Carbohydr.
Polym. 2007, 70, 451–458.
[18] B. Focher, P. L. Beltrame, A. Naggi, G. Torri, Carbohydr. Polym. 1990, 12,
405–418.
[19] J. T. Scanlon, D. E. Willis, J. Chromatogr. Sci. 1985, 23, 333–340.
[20] S. H. Langer, S. Connell, I. Wender, J. Org. Chem. 1958, 23, 50–58.
[7] a) S. Gandolfi, G. Ottolina, R. Consonni, S. Riva, I. Patel, ChemSusChem
2014, 7, 1991–1999; b) C. K. Nitsos, K. A. Matis, K. S. Triantafyllidis,
ChemSusChem 2013, 6, 110–122; c) H. Tadesse, R. Luque, Energy Envi-
ron. Sci. 2011, 4, 3913–3929.
Manuscript received: July 20, 2015
Revised: August 21, 2015
Accepted Article published: August 25, 2015
Final Article published: September 7, 2015
ChemPlusChem 2015, 80, 1565 – 1572
www.chempluschem.org
1572
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim