Hydrolysis of chitosan to yield levulinic acid and 5-hydroxymethylfurfural in water under microwave irradiation

Article abstract of DOI:10.1039/c2gc35048c

A novel route has been developed that yields levulinic acid (4-oxopentanoic acid, LA) and 5-hydroxymethylfurfural (5-HMF) from chitosan. Hydrolysis of chitosan was performed in the presence of a range of Lewis acids with SnCl ₄·5H₂O providing the best results. All reactions were performed in sealed vessels under microwave irradiation at 200 °C for 30 min. Typical pressures achieved were 17 to 19 bar. 23.9 wt% LA was produced from 100 mg chitosan using 0.24 mmol SnCl₄·5H₂O and 4 mL water. Under more dilute conditions, 10.0 wt% 5-HMF was obtained using 0.12 mmol SnCl₄·5H₂O and 15 mL water. We propose that under more concentrated reaction conditions the 5-HMF formed reacts further to produce LA. When chitin is treated similarly, no 5-HMF is produced but up to 12.7 wt% LA can be obtained. For comparison, 32.0 wt% LA was produced from 100 mg glucosamine hydrochloride using 0.26 mmol SnCl₄·5H ₂O and 20 mL water. This corresponds to a yield of 59.4%. The SnCl₄ forms SnO₂ and HCl in solution and under similar conditions using SnO₂ and HCl, chitosan formed 27.4 wt% LA.

Full text of DOI:10.1039/c2gc35048c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2012, 14, 1480
www.rsc.org/greenchem
PAPER
Hydrolysis of chitosan to yield levulinic acid and 5-hydroxymethylfurfural
in water under microwave irradiation†
Khaled W. Omari, Jessica E. Besaw and Francesca M. Kerton*
Received 11th January 2012, Accepted 9th March 2012
DOI: 10.1039/c2gc35048c
A novel route has been developed that yields levulinic acid (4-oxopentanoic acid, LA) and
5
-hydroxymethylfurfural (5-HMF) from chitosan. Hydrolysis of chitosan was performed in the presence
of a range of Lewis acids with SnCl ·5H O providing the best results. All reactions were performed in
4
2
sealed vessels under microwave irradiation at 200 °C for 30 min. Typical pressures achieved were 17 to
9 bar. 23.9 wt% LAwas produced from 100 mg chitosan using 0.24 mmol SnCl ·5H O and 4 mL water.
1
4
2
Under more dilute conditions, 10.0 wt% 5-HMF was obtained using 0.12 mmol SnCl ·5H O and 15 mL
4
2
water. We propose that under more concentrated reaction conditions the 5-HMF formed reacts further to
produce LA. When chitin is treated similarly, no 5-HMF is produced but up to 12.7 wt% LA can be
obtained. For comparison, 32.0 wt% LAwas produced from 100 mg glucosamine hydrochloride using
0
.26 mmol SnCl ·5H O and 20 mL water. This corresponds to a yield of 59.4%. The SnCl forms SnO
4 2 4 2
and HCl in solution and under similar conditions using SnO and HCl, chitosan formed 27.4 wt% LA.
2
Introduction
such it is an industrial waste material of fisheries and a renewable
feedstock with much potential. Three steps are used to purify
chitin from crustacean’ waste. These are (i) deproteinization
using a strong base such as NaOH, (ii) demineralization using an
acid such as HCl, and (iii) decolouration using a bleaching agent
Chitin is an important biopolymer that can be sourced from the
ocean and is the second most abundant biopolymer on Earth
after cellulose, Fig. 1. The estimated annual production of
chitin worldwide is about 1.5 × 10 tons. It is mainly available
from crustaceans’ shells such as crab, lobster and shrimp, and as
1
,2
5
2
2
such as H O . New methods are being developed that use green
2
2
chemistry techniques in this field, for example, the use of ionic
3
liquids.
Chitin is deacetylated under alkaline conditions to yield
chitosan. 100% deacetylation cannot be achieved and therefore,
chitosan is a copolymer of glucosamine and n-acetylglucosa-
4
mine. It is readily available in a range of molecular weights and
4
6
degrees of deacetylation, 5 × 10 to 2 × 10 Da and 40 to 98%,
respectively.
In recent years, many useful chemicals have been produced
5,6
from renewable feedstocks. For example, catalytic conversions
7
–10
11–15
16–21
of cellulose,
fructose
and glucose
into 5-HMF and
LA have been reported but amino-sugars and carbohydrates have
been overlooked. The chemistry of chitin and chitosan should be
investigated in concert with current studies on cellulose in order
to achieve maximum benefits from the most abundant bio-feed-
stocks available. The processing and usage of these N-containing
polysaccharides has been somewhat restricted over the years
because they contain many hydroxyl groups that are able to form
strong intra- and inter-molecular hydrogen bonds. Chitosan is
Fig. 1 Comparison of chitin, chitosan and cellulose formulae.
22,23
insoluble in basic media and water.
Likewise, chitin is in-
soluble in water, most organic solvents, and dilute acidic or basic
solutions.
Despite the currently limited industrial applications of these
biopolymers, their hydrolysis has been quite widely studied.
They can be hydrolyzed using enzymes such as cellulase,
hemicellulase, lysozyme, papain, pectinases and lipases to
Centre for Green Chemistry and Catalysis, Department of Chemistry,
Memorial University of Newfoundland, St. John’s, NL, Canada,
A1B 3X7. E-mail: fkerton@mun.ca
3
†
Electronic supplementary information (ESI) available: Graphs concern-
ing typical temperature, pressure and power conditions for microwave-
assisted reactions. Typical GC traces and mass spectra. SEMs. See DOI:
1
0.1039/c2gc35048c
1480 | Green Chem., 2012, 14, 1480–1487
This journal is © The Royal Society of Chemistry 2012

View Article Online
2
4,25
produce glucosamine, n-acetylglucosamine and oligomers.
dihydroxymethylfuran, 2,5-bis(hydroxymethyl)tetrahydrofuran
and 2,5-dimethylfuran. The latter is a promising liquid transpor-
Chitinases depolymerize chitin to produce chito-oligosaccharides
consisting of one to six N-acetylglucosamine (GlcNAc) sub-
units. High yields of diacetylchitobiose (GlcNAc)2 can be
obtained through hydrolysis of colloidal chitin in dimethylsulf-
oxide and lithium chloride using Vibrio furnissii chitinase.
Chitooligosaccharides can also be produced by chitin and chito-
san cleavage using mineral acids. Depolymerization of chitosan
in nitrous acid produces chitooligosaccharides and 2,5-anhydro-
D-mannose (M). Cleavage of chitin in sulfuric acid in the pres-
ence of acetic anhydride produces N-acetylchitooligosaccharide
17,38
tation fuel.
place of terephthalic acid.
FDCA can be used in polyester production in the
25
39
26
Results and discussion
Catalyst screening
25
Two series of reactions were performed initially to identify
whether a catalyst was required to hydrolyze chitosan in super-
heated water under microwave conditions and whether an acidic
or basic catalyst would give superior results. For each set of
reactions, a control reaction (no catalyst) was performed and 21
catalysts were screened and the amounts of LA and 5-HMF pro-
duced from medium molecular weight chitosan were determined.
The amount of chitosan processed under the two conditions was
fixed at 100 mg. The first condition was 0.24 mmol catalyst
and 4 mL deionized water. The second was 0.12 mmol catalyst
and 20 mL deionized water. The potential catalysts that were
assessed were anhydrous lanthanum trifluoromethanesulfonate
La(CF SO ) (1), gadolinium trifluoromethanesulfonate hydrate
25,27
peracetates.
However, aside from the production of simple sugars and
oligosaccharides, there have been few reports on the production
of chemicals from chitosan or chitin. One published example is
the conversion of chitin to 5-(chloromethyl)furfural and LA
using aqueous HCl and large amounts of 1,2-dichloroethylene.
The trimer (GlcNAc) -M has been prepared by treating chitosan
fraction of N-acetylated units = 0.59) in HNO2. This trimer
was then reported to produce 1% 5-HMF upon further exposure
to concentrated HNO . Conversion of gluocosamine (GlcN) to
-HMF has been briefly reported by other researchers using
organic acids in the presence of DMSO. The current main use
of glucosamine is as a dietary supplement and therefore some
studies concerning its stability have been performed. Pyrolysis of
glucosamine at 200 °C in the solid-state yields a mixture of
furans, pyridines, pyrroles and pyrazines. The most abundant
products 2-acetylfuran and 2-(2-furyl)-6-methylpyrazine were
present at levels of only 0.063 mg per g of glucoasmine pyro-
1
2
2
8
(
2
5
3
3 3
2
9
Gd(CF SO ) ·xH O (2), ytterbium trifluoromethanesulfonate
3 3 3 2
hydrate Yb(CF SO ) ·xH O (3), zinc perchlorate hexahydrate
3
3 3
2
Zn(ClO ) ·6H O (4), amberlyst 15 hydrogen form (5) which is
4
2
2
an acidic resin, anhydrous indium(III) chloride InCl (6), hydro-
3
chloric acid HCl (7), iron perchlorate hydrate Fe(ClO ) ·xH O
4
3
2
(8), nickel perchlorate hexahydrate Ni(ClO ) ·6H O (9), zirconyl
4
2
2
chloride octahydrate ZrOCl ·8H O (10), copper perchlorate
2
2
30
lyzed. In aqueous solution, when glucosamine was heated to
00 °C, a similar mixture of products formed and around 20 mg
hexahydrate Cu(ClO ) ·6H O (11), bismuth chloride BiCl (12),
4 3 2 3
chromium perchlorate hexahydrate Cr(ClO ) ·6H O (13), zirco-
4 3 2
nium tetrachloride ZrCl (14), tin chloride pentahydrate SnCl ·
4 4
1
31
of furan products were produced from 1 g of the sugar. These
results show that glucosamine has the potential to be a renewable
feedstock for furans and N-containing heterocycles and that
clearly, there is a need to further investigate the conversion of
chitin and chitosan into useful renewable chemical building
blocks.
5H O (15), manganese perchlorate hydrate Mn(ClO ) ·xH O
2
4 2
2
(16), yttrium trifluoromethanesulfonate Y(CF SO ) (17), acetic
3
3 3
acid CH COOH (18), ammonia NH (19), sodium hydroxide
3
3
NaOH (20), and basic alumina Al O (21). These were chosen
2
3
40
because the metal complexes are generally stable in water and
We present here our attempts to hydrolyze chitosan in the
absence of concentrated acids by generating superheated water
under microwave conditions and enhancing the hydrolysis pro-
cesses that occur by using a Lewis acid. Water is a clean, non-
corrosive, non-flammable, renewable, readily available, cheap
some of the other species, e.g., amberlyst, have given good
18
results for glucose/cellulose transformations. HCl was studied
1
because it is known to yield LA from chitin. SnCl ·5H O has
4
2
17
been reported to yield 5-HMF from glucose in an ionic liquid.
Both the concentrated and dilute reaction mixtures were heated
under microwave irradiation at 200 °C for 30 min. In the control
reactions and in the presence of 16 to 21 neither LA nor 5-HMF
were produced (details concerning the limit of detection and
limit of quantitation levels for GC-MS determination of LA and
5-HMF are provided in ESI†). For the concentrated reaction mix-
tures (0.24 mmol catalyst, 4 mL water), Fig. 2 shows the weight
percentages of LA and 5-HMF produced using catalysts 1 to 15.
25 wt% LA corresponds to a mol% yield of 33.8 for 80% de-
acetylated chitosan.
Generally, the metal-containing catalysts gave superior conver-
sions than the simple acids and bases. Acetic acid and the bases
studied, 18–21, yield no products, and HCl and amberlyst pro-
duced only a small amount of 5-HMF (<2.3%) and no LA. The
reason for the poor performance of amberlyst may be insufficient
swelling of the resin in aqueous solvents and also deposition of
biopolymer on the surface of the beads, which was evident
through SEM studies (Fig. S1 in the ESI†).
3
2,33
and environmentally friendly solvent.
Some of the advan-
tages of using microwave heating are that it reduces reaction
times and can increase product yields compared with conven-
3
4–36
tional heating methods.
Microwave heating is particularly
efficient for chemical transformations in water as it can be super-
3
5
heated in sealed vessels. In our study, different parameters
were varied, such as temperature, concentration, the Lewis acid
used and reaction time. Two compounds were identified as the
primary products from the hydrolysis of chitosan described
herein: levulinic acid (LA) and 5-hydroxymethylfurfural (5-
HMF). In addition, chitin under similar conditions produced LA.
LA can be used to produce many compounds such as ethyl levu-
linate and 2-methyltetrahydrofuran that can be used as miscible
diesel biofuel additives, δ-aminolevulinic acid a herbicide and
β-acetylacrylic acid, diphenolic acid and 1,4-pentanediol, which
3
7
are polymer building blocks. 5-HMF can yield other renewable
building blocks such as 2,5-furandicarboxylic acid (FDCA), 2,5-
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 1480–1487 | 1481

View Article Online
Fig. 3 Comparison of weight percentages LA and 5-HMF produced
Fig. 2 Comparison of weight percentages LA and 5-HMF produced
under dilute reaction conditions for a range of catalysts.
under concentrated reaction conditions for a range of catalysts.
these conditions, a smaller number of catalysts were effective,
namely, 3, 6–8, and 10 to 15.
Of the metal salts, the triflate catalysts were also ineffective
for this process, 1 to 3 produced small amounts of 5-HMF
Interestingly, under these dilute conditions, no catalyst was
ever able to produce LA indicating that LA production is
strongly dependent on reaction concentration. Catalysts 1, 2, 4, 5
and 9 although able to hydrolyze chitosan to 5-HMF or LA (in
the case of 9) under concentrated conditions, did not yield any
(<0.7%) and 17 was completely inactive. This may be due to
due to the non-coordinating nature of these anions compared
with chloride. The metal perchlorate Lewis acid catalysts were
generally a little more effective and in most cases produced LA
and/or 5-HMF, however, the Mn(II) salt (16) produced neither
LA nor 5-HMF. The Fe(III) salt (8) produced only 5-HMF
5
-HMF under these more dilute conditions. HCl (7) was the only
conventional (Brønsted) acid or base that produced 5-HMF
1.3 wt%) but this was a significantly smaller amount than that
generated by some of the Lewis acids studied. Also, Yb
CF SO ) ·xH O (3) was the only metal triflate that produced 5-
(
(3.16 wt%), whereas the Ni(II) salt (9) and the Cr(III) salt (13)
produced only LA (9.9 wt% and 13.0 wt%, respectively). The
Cu(II) salt (11) was the only metal perchlorate that produced LA
and 5-HMF simultaneously (10.9 wt% and 1.8 wt%, respect-
ively). These results indicate that the choice of metal ion is
important in determining the selectivity of the hydrolysis reac-
tion. However, in general, the overall yields of LA were strongly
(
3
3 3
2
HMF (1.7 wt%), whereas under more concentrated conditions
La(CF SO ) (1) and Gd(CF SO ) (2) were also active. Fe(III)
3
3 3
3
3 3
(
(
8), Cu(II) (11) and Cr(III) (13) perchlorates produced 5-HMF
3.3 wt%, 3.6 wt% and 5.1 wt%, respectively) in similar quan-
−
tities to those afforded under concentrated conditions. However,
no LAwas produced. Metal chloride Lewis acid catalysts yielded
larger amounts of 5-HMF under dilute conditions compared with
the more concentrated reaction mixtures. It is worth noting that
dependent on the anion and follow the reactivity trend: CF SO
<
3
3
−
−
ClO4 < Cl .
Under concentrated conditions, the metal chloride Lewis acid
catalysts generally produced the largest amounts of LA and 5-
InCl (6), in particular, was a much more effective catalyst under
3
HMF. InCl (6) produced a small amount of 5-HMF (2.2 wt%)
3
dilute reaction conditions, yielding 7.7 wt% 5-HMF compared
with 2.2 wt% under concentrated conditions. Under both con-
centrated and dilute reaction conditions SnCl ·5H O (15)
but no LA was produced. BiCl (12) produced LA (12.4 wt%)
3
and the largest amount of 5-HMF (4.0 wt%) under these
conditions amongst the catalysts screened. ZrCl4 (14) and
SnCl ·5H O (15) produced the largest amounts of LA (20.4 wt%
4
2
afforded the highest conversions, to 23.9 wt% LA (concentrated
conditions) and 10.0 wt% 5-HMF (dilute conditions). At this
point it is worth noting that SnCl ·5H O is a relatively cheap
4
2
and 23.9 wt%, respectively). ZrCl is known to react with water
4
4
1
4
2
to produce ZrOCl ·8H O (10) and HCl (7). As a result, we
2
2
and easily handled reagent compared to some of the other
species studied.
examined catalyst 10, which produced a moderate amount of LA
and 5-HMF (10.8 wt% and 3.4 wt%, respectively). However,
neither ZrOCl ·8H O nor HCl gave the largest amounts of LA
2
2
and 5-HMF under the conditions studied. Thus indicating that
this process is catalyzed more effectively by a suitable Lewis
acid rather than a Brønsted acid. The largest amounts of LA
4
2
Full factorial designs
In general, several factors are varied in order to optimize exper-
imental conditions. Chemists can use factorial design (FD) to
understand the interactions between factors and thereby obtain a
full picture of information and optimize a chemical reaction. If
each factor is studied at two different levels (high and low), FD
(23.9 wt%) and 5-HMF (3.95 wt%) produced in this set of reac-
tions came from those using SnCl ·5H O (15) and BiCl (12),
4
2
3
respectively.
For the more dilute reaction conditions screened (0.12 mmol
catalyst, 20 mL water), Fig. 3 summarizes the weight percentages
of LA and 5-HMF produced using the catalysts screened. Under
k
42
is expressed as 2 , where k is the number of factors. FD has
4
3–46
been practiced in many catalytic research applications.
This journal is © The Royal Society of Chemistry 2012
1482 | Green Chem., 2012, 14, 1480–1487

View Article Online
4
Table 1 Weight percentage of LA and 5-HMF produced from 100 mg chitosan in the presence of SnCl
4
·5H
2
O (15) using 2 factorial designs
Water volume (mL)
4 2
SnCl ·5H O (mmol)
Temperature (150 °C)
Time (15 min)
Temperature (200 °C)
Time (30 min)
Time (15 min)
Time (30 min)
a
Yield LA (wt%)
4
0.12
ND
ND
ND
ND
ND
ND
ND
ND
10.2 (6.2)
18.0 (1.0)
0 (8.8)
10.9 (4.9)
23.9 (0.6)
0 (10.0)
0
.24
0.12
.24
2
0
0
16.8 (3.3)
21.6 (3.6)
a
ND = not detected. Value in parentheses is the wt% 5-HMF produced concomitantly.
Fig. 4 Normal plot of the effects on the weight percentage of LA.
In order to optimize the reaction conditions using SnCl ·5H O
15), we applied 2 FD by varying the following four factors:
4
2
Fig. 5 Effect of temperature on the weight percent LA produced from
00 mg chitosan in 4 mL water with 0.24 mmol of SnCl ·5H O.
4
(
1
4
2
volume of water, catalyst loading, temperature and reaction time
to study all effects and their interplay at once. Table 1 shows
weight percentages of LA and 5-HMF produced at different
levels for each variable. The low and high levels of each factor
varied between 150 and 230 °C while the catalyst loading was
kept constant at 0.24 mmol. Fig. 5 shows that no LA was pro-
duced at lower temperatures (150–170 °C). However, the amount
produced increased almost linearly from 170 °C to 210 °C and
then levelled out.
Additional experiments were then performed in order to deter-
mine if any of the other variables would have a significant effect
on the amount of LA obtained. Doubling the amount of catalyst
to 0.48 mmol and performing the reaction at 200 °C yielded
4
were determined from preliminary experiments. 2 FDs were
studied using Minitab software. We used Lenth’s analysis to
analyze our FD (an alternative procedure to a normal plot). We
studied 15 effects. In order to determine the significant effects
out of the 15 effects, pseudo standard error (PSE) should be cal-
culated to determine the margin of error (ME). After calculating
the median (m) of the absolute values of the effects, let s = 1.5 ×
2
6.6 wt% LA compared with 23.9 wt% at tin levels of
m. The median (m ) of absolute values that are less than 2.5 × s
ο
0.24 mmol. Performing the hydrolysis reaction at 210 °C at high
tin levels (0.48 mmol), yielded 28.6 wt% LA. Actually, increas-
ing the time did not have a positive effect on the wt% of LA. For
example, the optimum condition at 30 min reaction time pro-
duced 23.9 wt% LA, which increased to reach 26.0 wt% at
were calculated. PSE is then equal to 1.5 × m . ME is when an
ο
effect lies within 95% confidence interval. Therefore, ME is
given by t(
1−alpha/2), df
× PSE. Where t is t-distribution, alpha =
1
–95/100 and df is the degree of freedom that is equal to number
47
of effects/3.
6
0 min. Therefore, although close to 30 wt% LA could be
obtained by increasing reaction time and the amount of tin used,
the effects of these changes are not significant when compared
with the effect of temperature as shown in Fig. 5.
Effects on LA production
The significant effects on LA production as shown in Fig. 4,
when alpha= 0.05 and ME = 5.23, are the catalyst amount (B)
and the temperature (C) factors.
Based on the data generated by the 2 FD experiment, we
changed the significant factors keeping the water volume con-
Effects on 5-HMF production
4
A similar factorial design experiment was performed in order to
optimize the reactions under dilute conditions and the production
of 5-HMF from chitosan. The significant effects, when alpha =
stant at 4 mL and the time constant at 30 min. Temperature was
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 1480–1487 | 1483

View Article Online
Fig. 6 Normal plot of the effects of factors on the weight percentage
of 5-HMF produced during microwave-assisted tin-catalyzed hydrolysis
of chitosan.
Fig. 8 Effect of dilution on weight percent 5-HMF produced (100 mg
chitosan, 0.12 mmol SnCl ·5H O).
4 2
produced increases from 4 mL until 15.0 mL as conditions
become more dilute and then levels out. When the volume of
water used was 15.0 mL or 20.0 mL, the amount of 5-HMF pro-
duced was similar. Further dilution of reactions mixtures, e.g.
3
0 mL water, was detrimental in terms of wt% 5-HMF obtained.
In order to examine the significant effects proposed from FD
analysis for 5-HMF production (i.e. AC, volume and tempera-
ture, and BC, [Sn] and temperature, Fig. 6), we compared the wt
%
5-HMF produced from simultaneously varying two factors.
Increasing the volume of water and temperature, whilst keeping
Sn] constant at 0.12 mmol, produced moderately less 5-HMF:
[
1
2
5 mL water at 200 °C yielded 10.0 wt% 5-HMF, Fig. 8 and
0 mL water at 210 °C yielded 9.0 wt% 5-HMF, Fig. 7. In con-
trast, simultaneously increasing the temperature from 200 to
2
0
10 °C and decreasing the catalyst loading from 0.12 to
.06 mmol SnCl ·5H O (in a constant volume of 15 mL H O)
4
2
2
Fig. 7 Effect of temperature on weight percent 5-HMF produced under
dilute reaction conditions (100 mg chitosan, 20 mL water and
caused a significant decrease in the amount of 5-HMF observed
(4.6 wt%) in line with changes predicted by FD.
0
4 2
.12 mmol SnCl ·5H O).
In summary, the optimum conditions for 5-HMF production
9.9 wt%) were microwave irradiation of 100 mg chitosan in
(
0
.05 and ME = 0.89, are volume of water (A), catalyst amount
1
5 mL water with 0.12 mmol SnCl ·5H O at 200 °C for 30 min.
4 2
(B) and temperature (C) factors as shown in Fig. 6.
Based on the data from the factorial design experiment, we
changed the significant factors keeping the time constant at
0 min. The effect of temperature on the production of 5-HMF
Conventional heating compared with microwave irradiation
3
was investigated between 150 and 210 °C while the volume of
water was kept constant at 20 mL and the catalyst amount main-
tained at 0.12 mmol. Fig. 7 shows that no 5-HMF was detected
for reactions conducted at 170 °C or lower. Between 170 and
As microwave heating, at present, is less widely used than con-
ventional heating methods, efforts were made to study the reac-
tions using conventional heating (stirrer-hotplate) with the same
optimised catalyst loadings and volumes of water as described
above. The reaction times and temperatures studied were chosen
using the Biotage Prediction Chart and Time Converter in
2
00 °C, 5-HMF production increased linearly but decreased
above this temperature. This decrease is likely due to more rehy-
dration reactions occurring at higher temperatures to yield LA
from the 5-HMF initially generated.
48
reverse, and as such similar yields of products would be pre-
dicted. The two mixtures (concentrated and dilute) were heated
at 100 °C in a round-bottomed flask equipped with a reflux con-
denser. This conventional heating process produced 12.08 wt%
LA after 9 days from 100 mg chitosan, 4 mL water, 0.24 mmol
SnCl ·5H O. This amount is almost half of that obtained via
Catalyst loading was studied at 0.06 and 0.24 mmol of Sn in
2
0 mL water at 200 °C and the percentages of 5-HMF were 6.3
and 3.6, respectively. This shows that the highest 5-HMF pro-
duction was at 0.12 mmol, which yielded 10 wt% 5-HMF.
The effect of dilution on production of 5-HMF was studied by
performing reactions in 4.0 to 30.0 mL water at a constant temp-
erature of 200 °C and a catalyst loading of 0.12 mmol
SnCl ·5H O. Fig. 8 shows that the weight percent of 5-HMF
4
2
microwave irradiation. In attempts to yield 5-HMF using the
optimum conditions from the microwave experiments, no
product was obtained after 9 days of conventional heating. Reac-
tions of chitosan at atmospheric pressure using microwave
4
2
1484 | Green Chem., 2012, 14, 1480–1487
This journal is © The Royal Society of Chemistry 2012

View Article Online
heating under both concentrated and dilute conditions were per-
formed (100 °C, 2 h, CEM Discover SP system). Unfortunately,
LA and 5-HMF were not produced. These results suggest a criti-
cal role for pressure in the reactions and they may be occurring
under near-critical water conditions.
weights (average 9.8 wt%). In contrast, chitin failed to produce
5-HMF, or in fact LA or any small EtOAc soluble organic com-
pounds under the conditions studied (Table 2). In the absence of
the catalyst, under the optimized conditions, chitosan produced
neither LA nor 5-HMF, Table 2. Also, reduced catalyst loadings
failed to yield the desired products under the conditions studied.
Studies using chitin as the feedstock
Proposed mechanism
In our studies, we have found that chitin can also be used to
produce LA. Different sources of chitosan and chitin were tested
under the optimum conditions of LA production that were deter-
mined using medium molecular weight chitosan. Table 2 sum-
marizes these results. Chitin produced an average of 12.3 wt%
LA and no differences were seen between chitin from crab or
from shrimp. Similarly, the amount of LA produced from differ-
ent sources and molecular weights of chitosan resemble each
other (average 24.2 wt%). This was somewhat surprising given
the tough appearance of high molecular weight chitosan. Gener-
ally, from our studies it can be seen that chitosan produces two
times more LA than chitin.
Our proposed mechanism for this process is shown in Fig. 9.
In the first step, chitosan is hydrolysed to yield glucosamine
via cleavage of the glycosidic bond (C1 → C4 linkage in
the polymer) in the presence of SnCl ·5H O and microwave
4
2
irradiation. Microwave-assisted acid-catalysed biopolymer
hydrolysis has previously been performed on starch to yield
49
glucose. This process occurs in a stepwise fashion and oligo-
mers are formed initially. Evidence for the formation of chito-
oligosaccharides has been obtained through mass spectrometric
studies, see below. We assume that as we see no product in the
absence of catalysts (Brønsted or Lewis acids) that coordination
of the amine functional group to either a proton or a metal centre
facilitates this bond-breaking process by weakening the proximal
As shown in Table 2, 5-HMF was produced in similar
amounts from chitosan of different sources and molecular
C –O bond and making it more amenable to hydrolysis. Further-
more by performing the reactions in water, both hydronium
and hydroxide ions are readily available to react at this site.
1
Table 2 Summary of results for different types of chitin and chitosan
treated using the optimized procedures for generation of LA and
a
5
-HMF
However, it is well known that an aqueous solution of SnCl is
4
Biopolymer
LA (wt%)
5-HMF (wt%)
mostly hydrolyzed and exists in equilibrium with colloidal tin
50
(
IV) oxide and HCl(aq). Therefore, reaction mixtures prepared
Low M.Wt chitosan
Medium M.Wt chitosan
High M.Wt chitosan
Chitosan, without catalyst
S-189
S-190
Chitin
24.7
23.9
25.1
ND
24.1
23.0
11.5
12.7
12.6
10.3
10.0
10.3
ND
9.3
for this process are acidic. If colloidal tin oxide (Alfa Aesar) and
HCl(aq) were used together rather than SnCl ·5H O, 100 mg of
4
2
chitosan yielded 27.4 wt% LA under the optimized conditions
described above. This is slightly greater than the yield under the
same conditions using SnCl ·5H O, and is significantly greater
8.8
ND
ND
ND
4
2
T-187
T-188
than the yields obtained using either HCl(aq) or tin oxide alone.
For comparison and to provide some evidence for the last
steps in the proposed mechanism, the reaction of glucosamine
was studied under similar conditions. 32.0 wt% LA was pro-
duced from 100 mg glucosamine hydrochloride using
a
ND = not detected. Chitosan and chitin purchased from Aldrich and
used as received. S-189 = shrimp chitosan, S-190 = crab chitosan, T-187
=
crab chitin and T-188 = shrimp chitin from Newfoundland’ fishery
waste provided by ChitinWorks America LLC.
0
.26 mmol SnCl ·5H O and 20 mL water heated to 200 °C for
4 2
Fig. 9 Proposed mechanism for 5-HMF and LA production from glucosamine. The glucosamine forms in situ via acid-catalysed hydrolysis of the
biopolymer.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 1480–1487 | 1485

View Article Online
3
0 min. This corresponds to a yield of 59.4%. Unfortunately,
degraded using green chemistry principles to give useful chemi-
cal building blocks in a similar way to ongoing research in the
area of cellulosic feedstocks. The volume of water used and
the loading of SnCl ·5H O can be varied to produce either LA
attempts to date to detect the ammonia or any reducing sugars in
the aqueous phase after extraction of the products from the reac-
tion mixtures have been unsuccessful. Therefore, we cannot
provide unequivocal evidence for all steps in the proposed
mechanism.
In recent literature, glucose is known to yield 5-HMF,
the subsequent conversion of 5-HMF to give LA is also
known.
HMF through loss of three water molecules – this process is also
facilitated by the presence of a Lewis acid. Furthermore, in the
presence of a high concentration of the Lewis acidic catalyst, as
in the optimum conditions for generation of LA from chitosan,
4
2
or 5-HMF with good selectivity. Factorial design was success-
fully employed to optimize the reaction conditions for this
process. Microwave irradiation proved to be a more effective
heating method for the generation of these small molecules com-
pared with conventional heating, as 5-HMF could not be gener-
ated effectively under the conditions studied using conventional
heating. A mechanism for this process has been proposed based
on the known chemistry of cellulose and glucose, and some
studies using glucosamine. Furthermore, evidence for oligosac-
charide intermediates has been obtained using MALDI-TOF
mass spectrometry. We were also able to perform the last step in
the mechanism under microwave irradiation to convert 5-HMF
to LA in good yields in water. The method developed for chito-
san could be extended to the more robust parent carbohydrate
chitin from which LA can also be obtained, albeit in smaller
quantities. The results of this study will open the possibility of
chemical and thermochemical transformations of the non-toxic
and cheap biopolymers, chitosan and chitin, to yield useful, sus-
tainable chemicals with possible industrial applications. Studies
in our group are ongoing using chitosan, chitin and amino-
sugars as feedstocks in different solvents e.g. ionic liquids (ILs)
and employing a wide range of catalysts.
1
7–21
and
18,21
In such processes, glucose dehydrates to yield 5-
5
-HMF is rehydrated into LA and formic acid through addition
of two water molecules. In order to test the last step in the mech-
anism, namely rehydration of 5-HMF into LA, 5-HMF
(26.0 mg) was processed under the optimum conditions for LA
production from chitosan (i.e., 4 mL water, 0.24 mmol
SnCl ·5H O, T = 200 °C and t = 30 min). The weight percent
yield of LA was 85.4 (or 92.7 mol%) thus providing evidence
for this last step. In the absence of the Lewis acid, no LA was
detected thus indicating the essential role of a metal centre in
this last step.
4
2
It should be noted that two volatile compounds (furan and
-methylfuran) were also identified via head-space analysis of
2
the gas phase inside the sealed reaction vessel. These have pre-
5
1
viously been observed when sugars are roasted. These com-
pounds have a lower degree of solubility in aqueous solution
than LA and 5-HMF and therefore, they were not found upon
extraction of organics from the aqueous reaction mixture. Other
hydrophilic or otherwise insoluble compounds would also be
overlooked using the GC-MS method employed. MALDI-TOF
mass spectra were obtained from aliquots of reaction mixtures in
an attempt to observe intermediate depolymerization products
and any compounds remaining in the aqueous phase. 2,5-dihy-
droxybenzoic acid (DHB) was used as the matrix, as it had been
used previously in the study of chitooligosaccharides. For ali-
quots taken 15 min into the reaction, for both dilute and concen-
trated reaction mixtures, some interesting peaks were observed in
the mass spectra. For concentrated reaction samples (4 mL
water), peaks were observed at m/z 1188 and 543. These could
Acknowledgements
We thank Memorial University, NSERC of Canada (Discovery
and RTI Grants–FMK, Undergraduate Student Research Award–
JEB), the Provincial Government of Newfoundland and
Labrador (IRIF grant–FMK), RDC-Newfoundland (Ocean
Industries Research Award – KWO) and Canada Foundation for
Innovation (Leaders Opportunity Fund Award–FMK) for gener-
ous funding. Thanks also to ChitinWorks America LLC for pro-
vision of chitosan and chitin samples sourced from the
Newfoundland fishery.
5
2
Notes and references
be assigned to the oligosaccharides (glucosamine) -n-acetylglu-
6
1
2
M. Mascal and E. B. Nikitin, ChemSusChem, 2009, 2, 859–861.
G. A. F. Roberts, Chitin Chemistry, Macmillan Press Ltd., London, UK,
cosamine and (glucosamine) -n-acetylglucosamine, respectively.
2
These peaks were not present in spectra from aliquots taken after
1
992.
3 Y. Qin, X. Lu, N. Sun and R. D. Rogers, Green Chem., 2010, 12, 968–
71.
3
0 min reaction time. This suggests that the oligosaccharides are
9
true intermediates, as they are formed at an early stage in the
reaction and then consumed. For dilute reaction samples (15 mL
water), a peak was observed at m/z 501. This could be assigned
4
M. F. A. Goosen, Applications of Chitin and Chitosan, Technomic Pub-
lishing Company, Pennsylvania, USA, 1997.
5 D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem., 2010, 12,
493–1513.
1
to the trisaccharide, (glucosamine) . This peak was also present
3
6
7
J. J. Bozell and G. R. Petersen, Green Chem., 2010, 12, 539–554.
J. H. Clark and F. E. I. Deswarte, Introduction to Chemicals from
Biomass, John Wiley & Sons, Ltd, West Sussex, UK, 2008.
in aliquots analyzed after 30 min reaction time. This indicates
that incomplete degradation/hydrolysis of the biopolymer occurs
under the dilute conditions studied.
8 R. Palkovits, K. Tajvidi, J. Procelewska, R. Rinaldi and A. Ruppert,
Green Chem., 2010, 12, 972–978.
9
M. Stocker, Angew. Chem., Int. Ed., 2008, 47, 9200–9211.
0 J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew. Chem., Int. Ed.,
007, 46, 7164–7183.
1 S. Hu, Z. Zhang, Y. Zhou, B. Han, H. Fan, W. Li, J. Song and Y. Xie,
1
2
Conclusions
1
Green Chem., 2008, 10, 1280–1283.
2 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr, ChemSusChem,
This study demonstrates that chitosan can be used to produce LA
and 5-HMF in water under microwave conditions. This work is
a proof-of-principle that N-containing biopolymers can be
1
2009, 2, 944–946.
13 J. Y. G. Chan and Y. Zhang, ChemSusChem, 2009, 2, 731–734.
1486 | Green Chem., 2012, 14, 1480–1487
This journal is © The Royal Society of Chemistry 2012

View Article Online
1
4 X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr, Green Chem., 2009,
1, 1327–1331.
5 Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic, Science, 2006, 312,
933–1937.
35 D. Dallinger and C. O. Kappe, Chem. Rev., 2007, 107, 2563–
2591.
36 J. D. Moseley and C. O. Kappe, Green Chem., 2011, 13, 794–
806.
1
1
1
1
1
6 M. Chidambaram and A. T. Bell, Green Chem., 2010, 12, 1253–1262.
7 S. Hu, Z. Zhang, J. Song, Y. Zhou and B. Han, Green Chem., 2009, 11,
37 M. Mascal and E. B. Nikitin, Green Chem., 2010, 12, 370–373.
38 F. Ilgen, D. Ott, D. Kralisch, C. Reil, A. Palmberger and B. Koenig,
Green Chem., 2009, 11, 1948–1954.
39 J. N. Chheda, Y. Roman-Leshkov and J. A. Dumesic, Green Chem.,
2007, 9, 342–350.
40 S. Kobayashi, S. Nagayama and T. Busujima, J. Am. Chem. Soc., 1998,
120, 8287–8288.
41 V. V. Guzeev and A. N. D’yachenko, Glass Ceram., 2002, 59, 49–52.
42 R. D. Moen, T. W. Nolan and L. P. Provost, Quality Improvement
Through Planned Experimentation, 2nd edn, McGraw-Hill, New York,
USA, 1999.
1
746–1749.
8 A. Takagaki, M. Ohara, S. Nishimura and K. Ebitani, Chem. Commun.,
009, 6276–6278.
9 T. Staahlberg, M. G. Sorensen and A. Riisager, Green Chem., 2010, 12,
21–325.
0 G. Yong, Y. Zhang and J. Ying, Angew. Chem., Int. Ed., 2008, 47, 9345–
348.
1 H. Zhao, J. E. Holladay, H. Brown and Z. C. Zhang, Science, 2007, 316,
597–1600.
1
1
2
2
2
2
3
9
1
2 G. Ma, D. Yang, D. Su, X. Mu, J. F. Kennedy and J. Nie, Polym. Adv.
Technol., 2010, 21, 189–195.
43 G. Epane, J. C. Laguerre, A. Wadouachi and D. Marek, Green Chem.,
2010, 12, 502–506.
2
2
3 B. Krajewska, Sep. Purif. Technol., 2005, 41, 305–312.
4 L. Ramirez-Coutino, M. d. C. Marin-Cervantes, S. Huerta, S. Revah and
K. Shirai, Process Biochem., 2006, 41, 1106–1110.
44 L. Veum, S. R. M. Pereira, J. C. van der Waal and U. Hanefeld,
Eur. J. Org. Chem., 2006, 1664–1671.
45 F. Stazi, G. Palmisano, M. Turconi and M. Santagostino, Tetrahedron
Lett., 2005, 46, 1815–1818.
46 J. R. Ochoa-Gomez, O. Gomez-Jimenez-Aberasturi, B. Maestro-
Madurga, A. Pesquera-Rodriguez, C. Ramirez-Lopez, L. Lorenzo-
Ibarreta, J. Torrecilla-Soria and M. C. Villaran-Velasco, Appl. Catal., A,
2009, 366, 315–324.
2
5 R. A. A. Muzzarelli and M. G. Peter, Chitin Handbook, European Chitin
Society, Grottammare, Italy, 1997.
2
2
6 J. H. Yoon, Enzyme Microb. Technol., 2005, 37, 663–668.
7 K. Kurita, K. Tomita, S. Ishii, S. Nishimura and K. Shimoda, J. Polym.
Sci., Part A: Polym. Chem., 1993, 31, 2393–2395.
2
2
8 K. Tømmeraas, K. M. Varum, B. E. Christensen and O. Smidsrød,
Carbohydr. Res., 2001, 333, 137–144.
47 G. E. P. Box, J. S. Hunter and W. G. Hunter, Statistics for Experimenters:
Design, Innovation, and Discovery, John Wiley & Sons, New Jersey,
USA, 2005.
48 Biotage Pathfinder, Prediction chart and time converter, http://www.
biotagepathfinder.com/util.jsp, accessed March 9th 2012.
49 H.-M. Yu, S.-T. Chen, P. Suree, R. Nuansri and K.-T. Wang, J. Org.
Chem., 1996, 61, 9608–9609.
9 T. Wu and S. Zivanovic, Abstracts of Papers, 234th ACS National
Meeting, Boston, MA, United States, August 19–23, 2007, IEC-015.
0 J. Chen and C.-T. Ho, J. Agric. Food Chem., 1998, 46, 1971–1974.
1 M. Jun, Y. Shao, C.-T. Ho, U. Koetter and S. Lech, J. Agric. Food
Chem., 2003, 51, 6340–6346.
3
3
3
3
3
2 F. M. Kerton, Alternative Solvents for Green Chemistry, Royal Society of
Chemistry Publishing, Cambridge, UK, 2009.
3 U. M. Lindstrom, Organic Reactions in Water: Principles, Strategies and
Applications, Blackwell Publishing, Oxford, UK, 2007.
4 C. O. Kappe, D. Dallinger and S. S. Murphree, Practical Microwave Syn-
thesis for Organic Chemistry: Strategies, Instruments, and Protocols,
WILEY-VCH, Weinheim, Germany, 2009.
50 N. Wiberg, A. F. Holleman and E. Wiberg, Holleman-Wiberg Inorganic
Chemistry, 34th edn (101st printing), Academic Press, San Diego, 2001,
p. 908.
51 A. Limacher, J. Kerler, T. Davidek, F. Schmalzried and I. Blank, J. Agric.
Food Chem., 2008, 56, 3639–3647.
52 S. Trombotto, C. Ladavière, F. Delolme and A. Domard, Biomacromole-
cules, 2008, 9, 1731–1738.
This journal is © The Royal Society of Chemistry 2012
Green Chem., 2012, 14, 1480–1487 | 1487