Kinetics and efficiency of chitosan reacetylation

Article abstract of DOI:10.1016/j.carbpol.2011.08.096

Chitosan reacetylation kinetics and efficiency were studied in water-methanol (MeOH) mixtures. The polymer was dissolved using acetic acid and acetic anhydride was used for reacetylation. Combining second-order kinetics and acid-base dissociation equations of chitosan, and using acetic anhydride hydrolysis rates determined by conductivity measurements, reacetylation reaction rate constants of 187, 108, 46 min^-1 M^-1 were found in 0, 50 and 80% MeOH (v/v), respectively. Contrary to previous literature, it was found that improvement in reacetylation efficiency in the presence of MeOH is mainly due to an increase of acetic acid pK_a by MeOH that limits the ionization of the polymer in the course of the reaction rather than to a decreased acetic anhydride hydrolysis rate, as previously thought. Based on these insights, the model developed in this study was able to predict the significantly reduced efficiency of the reaction for a large extent of reacetylation, without requiring any steric hindrance from the acetyl group. Conditions to maximize the reaction efficiency for a large extent of reacetylation were identified.

Full text of DOI:10.1016/j.carbpol.2011.08.096

Carbohydrate Polymers 87 (2012) 1192–1198
Contents lists available at SciVerse ScienceDirect
Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol
Kinetics and efficiency of chitosan reacetylation
Marc Lavertu ,1, Vincent Darras , Michael D. Buschmann
∗
1
Institute of Biomedical Engineering, Department of Chemical Engineering, École Polytechnique, P.O. Box 6079, Station Centre-ville, Montréal, QC, Canada H3C 3A7
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 26 July 2011
Received in revised form 29 August 2011
Accepted 30 August 2011
Available online 6 September 2011
Chitosan reacetylation kinetics and efficiency were studied in water–methanol (MeOH) mixtures. The
polymer was dissolved using acetic acid and acetic anhydride was used for reacetylation. Combin-
ing second-order kinetics and acid–base dissociation equations of chitosan, and using acetic anhydride
hydrolysis rates determined by conductivity measurements, reacetylation reaction rate constants of 187,
−1
−1
1
08, 46 min
M were found in 0, 50 and 80% MeOH (v/v), respectively. Contrary to previous literature,
it was found that improvement in reacetylation efficiency in the presence of MeOH is mainly due to an
increase of acetic acid pKa by MeOH that limits the ionization of the polymer in the course of the reac-
tion rather than to a decreased acetic anhydride hydrolysis rate, as previously thought. Based on these
insights, the model developed in this study was able to predict the significantly reduced efficiency of the
reaction for a large extent of reacetylation, without requiring any steric hindrance from the acetyl group.
Conditions to maximize the reaction efficiency for a large extent of reacetylation were identified.
© 2011 Elsevier Ltd. All rights reserved.
Keywords:
Chitosan
Polysaccharide
Biomaterial
Reacetylation
Acetic anhydride
Kinetics
Hydrolysis
Efficiency
1
. Introduction
Chitosan is a cationic polysaccharide prepared by alkaline
Reacetylation is a previously applied alternative which produces
chitosan with a more homogeneous and random distribution of
acetyl groups compared to heterogeneously deacetylated chitosan
(Aiba, 1992, 1994). Reacetylation is useful to prepare chitosan
samples with similar molecular weight but different DDA (Knaul,
Kasaai, Bui, & Creber, 1998; Maghami & Roberts, 1988) and to
obtain soluble chitosan at physiological pH values (Dal Pozzo et al.,
2000). Homogeneously reacetylated chitosans display different
enzymatic degradation properties compared to heterogeneously
prepared chitosans, and are typically less degradable (Shigemasa,
Saito, Sashiwa, & Saimoto, 1994), due to the random distribu-
tion of acetyl groups on the polymer backbone (Aiba, 1992).
Hydrogels may be prepared by reacetylating chitosan to a suffi-
cient extent so that it becomes insoluble (Hirano, Kondo, & Ohe,
1975; Moore & Roberts, 1980; Vachoud, Zydowicz, & Domard,
1997).
Reacetylation of chitosan is performed by adding acetic anhy-
dride to chitosan solutions. The reaction involves a nucleophilic
attack of an unprotonated primary amino group of glucosamine on
one carboxyl function of the acetic anhydride. Since acetic anhy-
dride is hydrolysed in water, methanol (MeOH)/water mixtures
are typically used for efficient reacetylation. It is worth noting
that chitosan is not soluble in 100% MeOH so that the reaction
cannot be performed under conditions where hydrolysis of acetic
anhydride could be completely eliminated. Acetic acid is typically
used to solubilise chitosan for reacetylation, as it is preferred to
HCl since dissociation equilibrium of the amine of chitosan can be
displaced during the course of the reaction (Vachoud et al., 1997).
deacetylation of chitin that is found in the shells of crustaceans.
It is a linear polyelectrolyte composed of glucosamine and N-
acetyl-glucosamine units linked by ␤-(1 → 4) glycosidic bonds,
where the fraction of units that are glucosamine is defined as
the degree of deacetylation (DDA). Chitosan is usually obtained
by alkaline deacetylation of chitin, where particulate chitin is sus-
pended in a hot alkaline solution (typically 40–50%, w/w, NaOH).
Although acetyl content can be targeted by alkaline deacetylation,
this heterogeneous reaction is thought to result in a block distri-
bution of acetyl groups (Aiba, 1991; Kurita, Sannan, & Iwakura,
1
1
977). Using H NMR, Varum, Anthonsen, Grasdalen, and Smidsrod
(
1991) confirmed that heterogeneously prepared chitosans have
a slightly more blockwise distribution and Ottoy, Varum, and
Smidsrod (1996) showed that such chitosans often comprise a
highly acetylated acid-insoluble fraction. On the other hand, using
nitrous acid depolymerisation and steric exclusion chromatog-
raphy, Sashiwa, Saimoto, Shigemasa, Ogawa, and Tokura (1991)
and Sashiwa, Saimoto, Shigemasa, and Tokura (1993) obtained
results suggesting a random distribution of N-acetyl group in
both heterogeneously and homogeneously prepared chitosans.
∗ Corresponding author. Tel.: +1 514 340 4711x3609; fax: +1 514 340 5129.
E-mail address: marc.lavertu@polymtl.ca (M. Lavertu).
These authors contributed equally.
1
0
144-8617/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbpol.2011.08.096

M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
1193
Previous studies report the reacetylation of chitosan using different
2.3. pKa measurements in different solvent mixtures
conditions (% MeOH, chitosan and acid concentrations, etc.) and
report variable efficiencies (Aiba, 1994; Berger et al., 2005; Knaul
et al., 1998; Qun, Ajun, & Yong, 2007; Vachoud et al., 1997). A
reduced efficiency of the reaction for larger extent of reacetyla-
tion has been observed and has been attributed, at least partly, to
increased steric hindrance at high acetyl content (Qun et al., 2007;
Vachoud et al., 1997). Although this reacetylation process has been
widely used, the kinetics of the reaction has not been examined in
detail (Vachoud et al., 1997).
A 500 mM acetic acid solution was prepared by diluting glacial
acetic in water. Sodium hydroxide was added to adjust the degree of
ionization to 0.5 so that pH ∼ pKa. 1 ml of this solution was added to
9 ml water–methanol mixtures such that final compositions were
50 mM acetic acid in 0, 25, 50, 80 and 90% methanol (v/v). The pH
of these solutions was measured using a glass body pH electrode
(PerpHecT® ROSS, model #8220BNWP) calibrated with aqueous
standard buffers. Under these conditions, the changes of pH from
one solvent composition to another equal pKa changes (Filion et al.,
2007). pKa values for various solvent compositions were calculated
from these pH changes using a reference pKa value of 4.76 for acetic
acid in water (Bacarella, Grunwald, Marshall, & Purlee, 1955) and by
correcting for the solvent effect using Eq. (15). Similarly, a 500 mM
aqueous solution of glucosamine hydrochloride with a degree of
ionization adjusted to 0.5 using sodium hydroxide was prepared.
1 ml of this solution was added to 9 ml water–methanol mixtures
and pH was measured, as described above for acetic acid solutions.
pKa values were determined using a reference pKa of 7.8 for glu-
cosamine in water (Park, Choi, & Park, 1983). These glucosamine
pKa changes were then used to estimate the values of the intrinsic
In this paper, we have characterized the kinetics and efficiency
of chitosan reacetylation in water–methanol mixtures of 0, 50 and
8
0% MeOH (v/v). A model and solution to the set of differential equa-
tions describing the kinetics of reacetylation is described, including
the calculation of the degree of ionization of glucosamine unit dur-
ing the course of the reaction and the hydrolysis kinetics of acetic
anhydride. Using appropriate dielectric constant values (Albright
&
Gosting, 1946), the ionization degree is calculated using a mean
field Poisson–Boltzmann model as described previously (Filion,
Lavertu, & Buschmann, 2007; Lavertu, Filion, & Buschmann, 2008).
The detailed model and all relevant equations are available in the
Supplementary material section.
pKa of chitosan (pK ) for various water–methanol mixtures using a
0
2
. Materials and methods
reference pK0 value of 6.7 for chitosan in water (Filion et al., 2007).
All pH measurements were done in the intersolvental pHs scale
w
2.1. Materials
(Inczédy, Lengyel, Ure, and International Union of Pure and Applied
Chemistry (1998) and March (1992)) and were converted to the pHs
s
Acetic anhydride (cat # 242845), sodium hydroxide standard
scale as described in Canals, Oumada, Roses, & Bosch, 2001. Note
that all pKa values in this study are reported on the spH scale.
s
solution – 1.0 N in H O (cat # 319511), sodium azide (cat #S2002),
2
hydrochloric acid standard solution – 1.0 N in H O (cat #318949),
deuterium oxide (cat # 151882), deuterium chloride solution –
2
2.4. Chitosan characterization
3
5 wt.% in D O (cat # 543047) and D-(+)-glucosamine hydrochlo-
2
Chitosan DDA was determined by 1H NMR as previously
described (Lavertu et al., 2003). 1H NMR measurements were
performed using a Varian Inova 500 MHz spectrometer operated
with VNMRJ 2.3A software and equipped with a 5 mm SW PFG
ride (cat # G1514) were purchased from Sigma–Aldrich. Glacial
acetic acid (cat # 00598-320) and methanol HPLC grade (cat #
9
093-03) were obtained from Anachemia Science and from J.T.
Baker, respectively. Chitosan (Ultrasan) with a degree of deacety-
lation (DDA) of 98.6% was provided by Piramal Healthcare Canada.
Dextran 500 (cat # 17-0320-01) was purchased from Amersham
Biosciences.
1
19
15
31
H/ F/X[ N- P] probe. Molecular weight of starting chitosan was
determined by size-exclusion chromatography (SEC) as previously
described (Nguyen, Winnik, & Buschmann, 2009). Measurements
were performed on a Gel Permeation Chromatography (GPC) sys-
tem consisting of an LC-20AD isocratic pump (Shimadzu), an
autosampler SIL-20AC HT (Shimadzu), an oven CTO-20AC (Shi-
madzu) coupled with a Dawn HELEOS II multiangle laser light
scattering detector (Wyatt Technology Co.), a Viscostar II (Wyatt
Technology Co.), an Optilab rEX interferometric refractometer
(Wyatt Technology Co.), and two Shodex OHpak columns (SB-806M
HQ and SB-805 HQ) connected in series. A chitosan dn/dc value of
0.214 was used (laser’s wavelength of 658 nm) and the number and
weight average molecular weights (Mn and Mw) were found to be
equal to 132 kDa and 187 kDa, respectively. The water content of
chitosan was determined using a heated centrifugal vacuum con-
centrator (Savant Speedvac, model SS11) by measuring the loss on
2.2. Hydrolysis kinetics of acetic anhydride
Acetic anhydride concentrations of 5, 15, 45, 200 mM were
tested in 0, 50, 80 and 90% methanol–water mixtures (v/v), respec-
tively. The extent of dissociation of acetic acid decreases as the
proportion of methanol increases (i.e. its pKa increases) and pro-
gressively higher concentrations of acetic anhydride are required
to accurately measure conductivity as the proportion of methanol
increases. Acetic anhydride was added to methanol–water mix-
tures while stirring and conductivity was monitored as a function of
®
time using an Accumet conductivity probe (Fisher Scientific, cat #
®
1
3-620-156) connected to an Accumet Model 20 pH/conductivity
◦
meter from Fisher Scientific. The conductivity probe was cali-
drying of 100 mg of chitosan after 48 h at 60 C. The percentage
brated using a 10 ␮S/cm standard solution (Fisher Scientific, cat
of water in chitosan was found to be 9.7% and the concentrations
of chitosan solutions prepared from non-dried powder throughout
this study were adjusted accordingly.
#
09-328-1). Solutions were tested in a jacketed beaker coupled
to a circulating bath (Neslab, model RT-111) to control tem-
◦
perature. Conductivity measurements were performed at 22.5 C.
The hydrolysis reaction rates were calculated from these con-
ductivity measurements as described in Asprey, Wojciechowski,
Rice, & Dorcas, 1996. Hydrolysis kinetics measurements were
also performed on an aqueous solution of dextran 500 (nomi-
nal Mw = 500 kDa) 10% (w/w). According to the literature, such
a solution has a dynamic viscosity (∼20 cP (Xu, Li, Zhang, & Qi,
2.5. Chitosan reacetylation kinetics and efficiency
Chitosan was first dissolved in water using glacial acetic acid.
The chitosan powder was dispersed in water using a magnetic
stirrer prior to glacial acetic acid addition. In order to obtain a
final concentration of 4.5 mg/ml after methanol addition (i.e. final
concentration of 27.5 mM of glucosamine units of chitosan for all
solutions in this study), the chitosan concentration in water prior
to methanol addition was adjusted to 4.5, 9.0 or 22.5 mg/ml for 0,
2
009)) close to that of the chitosan solutions (Martinez, Chornet, &
Rodrigue, 2004) tested in this study and was thus used to determine
if hydrolysis rate was affected by solution viscosity.

1
194
M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
Table 1
Hydrolysis and reacetylation reaction rate constants in water–methanol mixtures.
ꢀ
(min )a (×10 )
−1
3
−1
−1)a (×103)
−1
−1
)b
M
%
MeOH (v/v)
k
kH (min
M
kR (min
H
0
0
0
0
145 ± 1
68 ± 1
2.63 ± 0.01
2.40 ± 0.01
2.04 ± 0.01
1.53 ± 0.01
187 ± 17
108 ± 6
46 ± 2
–
5
8
9
23.2 ± 0.1
8.6 ± 0.1
a
n = 2, mean ± (max − min)/2.
b
n = 3, mean ± SD.
5
0 or 80% MeOH (v/v), respectively. Since acetic anhydride reacts
with the neutral form of glucosamine and in order to maximize
the reacetylation efficiency, the amount of glacial acetic acid was
adjusted to a minimum that ensured polymer solubility for a given
proportion of methanol. We found that for mixtures with 0 and
5
0% (v/v) of methanol, the final acetic acid concentration needed
is 25 mM while 80 mM is required for 80% of methanol (v/v). Thus,
acetic acid concentration prior to methanol addition was adjusted
to 25, 50 or 400 mM for 0, 50 or 80% MeOH (v/v), respectively.
Once chitosan was dissolved, an appropriate volume of methanol
was added in order to reach a final composition of 0, 50 or 80%
MeOH (v/v) and solutions were stirred for 2 h in order to ensure
complete homogenisation and thermal equilibrium. 30 ml chitosan
solutions were used to study reacetylation kinetics. Kinetics exper-
◦
iments were performed at room temperature (∼22.5 C) in a 50 ml
round bottom flask using 4.5 mg/ml chitosan water–methanol solu-
tions prepared as described above. Acetic anhydride was added
to stirred chitosan water–methanol mixtures at a molar ratio of
acetic anhydride to glucosamine (R) of 0.45. At various time points
(
30 s, 1, 2, 4, 8, 15 and 30 min), 2 ml aliquots of the reaction mixture
were pipetted in 3 ml of ice-cold 1 M sodium hydroxide solution in
order to precipitate the reacetylated chitosan and stop the reaction.
All samples were then repeatedly centrifuged at about 4000 × g
and washed with deionized water at least five times in order to
remove acetic acid and sodium hydroxide. Once washed, samples
were frozen and freeze-dried prior to analysis by 1H NMR. Effi-
ciency of the chitosan reacetylation as a function of the molar ratio
of acetic anhydride to glucosamine (R) was determined using 5 ml
chitosan solutions prepared as described above. Experiments were
Fig. 1. Hydrolysis kinetics of acetic anhydride for methanol–water mixtures mon-
itored by conductivity measurements. (A) Conductivity as a function of time
normalized to the fit value of the conductivity extrapolated to infinite time (cinf). The
experimental data (symbols) are fit to the model (dashed lines) using Eq. (4). Cal-
◦
performed at room temperature (∼22.5 C) in 20 ml scintillation
ꢀ
ꢀ
culated k values were obtained with n = 2(mean k ± (max − min)/2), and values
H
H
vials under magnetic stirring. Reaction times of 180 min and R val-
ues of 0.075, 0.145, 0.29 and 0.45 were used. In cases where a very
small amount of acetic anhydride was required, the reagent was
first diluted in methanol just prior to addition to the chitosan solu-
tion. The reacetylated samples were washed by centrifugation and
degree of acetylation DA was measured by 1H NMR as described
above. Note that in order to facilitate analysis of reacetylated mate-
rial, we limited the range of R so that the final product would be
soluble.
are reported in Table 1. (B) Representation of acetic anhydride hydrolysis kinetics
ꢀ
using Eq. (5). The theoretical curves (dashed lines) were obtained using the kH fit
values found in (A).
in these conditions decreases by about 10% as compared to the
aqueous solution. The density of this 10% (w/w) dextran solution
ꢀ
was found to be 1.034 g/ml so that about 60% of this kH reduc-
tion can be explained by the reduced concentration of water in
these conditions. Therefore, it appears that solution viscosity does
not significantly affect hydrolysis kinetics, and values obtained in
water–methanol mixtures (values in Table 1) can be used to model
kinetics in presence of chitosan.
3
. Results and discussion
3.1. Hydrolysis of acetic anhydride
3.2. Variation of pKa in water–methanol mixtures
Hydrolysis kinetics of acetic anhydride in water–methanol mix-
tures are shown in Fig. 1. The reaction rate constants obtained
from these fits using Eq. (4) are shown in Table 1. As expected, the
The changes in pKa of acetic acid and glucosamine as a func-
tion of methanol volume fraction in water–methanol mixtures are
shown in Fig. 2. Acetic acid pKa increases significantly as the propor-
tion of methanol increases. The dissociation of a neutral acid such
as acetic acid involves electrostatic work associated with the sep-
aration of two opposite charges. As methanol proportion increases
in the system, the dielectric constant of the solvent decreases
hydrolysis rate decreases as the proportion of methanol increases.
ꢀ
The value found for the pseudo-first order reaction in water (k
=
H
−1
0
.145 min ) is in good agreement with the reported value in the
−1
◦
literature of 0.16 min at 25 C (Cleland & Wilhelm, 1956). Inter-
estingly, the true second-order reaction rate (kH) decreases as the
proportion of methanol increases in the mixture. Measurements
◦
(the dielectric constant of methanol is ∼33 at 25 C versus 78.5
for water) and the work required to separate the two opposites
charges increases compared to water alone, thereby reducing the
ꢀ
on a dextran 500 aqueous solution (10%, w/w) revealed that kH

M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
1195
pure water which results in a more important acidification of the
solution as reacetylation and hydrolysis reactions progress. Calcu-
lated initial/final values of 1 − ˛ for these reactions are 0.23/0.03,
0
.4/0.14 and 0.39/0.21 for 0, 50 and 80% MeOH (v/v), respectively.
It is mostly the further ionization of the polymer in water in the
course of the reaction that limits the efficiency of reacetylation and
to a lesser extent the higher hydrolysis rate for these conditions.
3.4. Efficiency of the reacetylation reaction
Final DA values obtained using various amount of acetic anhy-
dride are compared to theoretical values in Fig. 4. Theoretical curves
were calculated using the reaction constants shown in Table 1 and
obtained from Fig. 3 results. For the range of acetic anhydride to
glucosamine ratios (R) tested, the reduction of the efficiency of the
reaction as R increases can be seen clearly for 0% MeOH. As dis-
cussed above, this result is mostly due to the acidification of the
solution by production of acetic acid by both reactions and not to
acetic anhydride hydrolysis or glucosamine consumption (for this
range of R). Since acetic acid pKa is higher in 50 and 80% MeOH than
in water, a reduction of the efficiency with increasing R is not clear
in the range of R tested for 50 and 80% MeOH (v/v), but reacety-
lation in 80% MeOH appears to be slightly more efficient than 50%
MeOH for the largest R values. The improved efficiency of the reac-
tion as the proportion of methanol increases has been reported in
the literature (Aiba, 1994) but without explanation. The theoreti-
cal reduction of the efficiency for larger R is shown in Fig. 4B. These
calculations were performed by assuming that gelation or precipi-
tation of the polymer occurring at high DA values would not affect
significantly reaction kinetics. The results illustrate how increasing
methanol proportion increases reacetylation efficiency for higher
R values. The curves also reveal that even in a twofold excess of
Fig. 2. Variation of the pKa of acetic acid and glucosamine as a function of methanol
volume fraction (v/v). The pKa changes were determined by measuring pH changes
in methanol–water mixtures with 50 mM of either acetic acid or glucosamine at a
degree of dissociation of ∼0.5. pKa were calculated from the pH changes added to
the known pKa values in water of 4.76 (Bacarella et al., 1955) and 7.8 (Park et al.,
1
983) for acetic acid and glucosamine, respectively (n = 2, mean ± (max − min)/2).
Our results for acetic acid are compared to the results of Shedlovsky and Kay (1956)
dotted line is a polynomial fit of these previous literature results).
(
dissociation constant. Solute–solvent interactions (solvation) also
influence dissociation, but they are generally outweighed by elec-
trostatic interactions (Bosch, Bou, Allemann, & Roses, 1996). Fig. 2
shows our results to be in agreement with acetic acid pKa values
reported in the literature (Shedlovsky & Kay, 1956). For exam-
ple, the pKa value measured at 80% MeOH (v/v) is within 0.02 pH
units of the reported value in the literature. We found that glu-
cosamine pKa varies to a much lower extent than acetic acid pKa, as
expected since dissociation of cationic acids involves no charge sep-
aration (no electrostatic contributions) and thus depends only on
solute–solvent interactions. As stated above in the Section 2, these
glucosamine pKa variations are used as an estimate for the varia-
tions of chitosan intrinsic pKa (pK ). The principal implications of
0
this progressive decrease in the pKa difference between acetic acid
and chitosan (pKa (glucosamine) − pKa (acetic acid) = ∼3.0 in water
vs. pKa (glucosamine) − pKa (acetic acid) = ∼1.5 in 80% MeOH) as the
amount of methanol increases are (1) a larger amount of acetic acid
is required to solubilise chitosan in mixtures with more methanol
(
25 mM vs. 80 mM acetic acid to dissolve 27.5 mM of chitosan (glu-
cosamine units) in 50 and 80% MeOH mixtures, respectively) and
2) the production of acetic acid in the course of the reaction by
(
either the reacetylation of chitosan or the hydrolysis of acetic anhy-
dride ionizes chitosan (or acidifies the solution) to a lesser extent
as the proportion of methanol increases and should result in higher
reaction efficiency.
3.3. Kinetics of chitosan reacetylation
The kinetics of reacetylation of chitosan in water–methanol
mixtures are shown in Fig. 3A. Most of the reacetylation occurs
within a few minutes and the reaction is completed after 30 min for
all conditions tested here. The reacetylation rate constant fit values
−
1
−1
for
from these experimental values are 183, 108 and 46 min
M
0, 50, and 80% MeOH (v/v), respectively (see Table 1). The reason
why the reacetylation rate decreases as the proportion of methanol
increases is currently unknown, but it could be related to changes
in the chitosan chain conformation or viscosity of the solution in
presence of methanol.
The theoretical evolution of the fraction of neutral amine moi-
eties as a function of time during reacetylation is shown in Fig. 3B.
The degree of ionisation of chitosan increases significantly more in
the course of the reaction in aqueous solutions versus 50 and 80%
MeOH (v/v). This can be explained by the lower pKa of acetic acid in
Fig. 3. Kinetics of reacetylation of chitosan as a function of methanol volume frac-
tion (v/v). The solution compositions are 27.5 mM glucosamine units of chitosan
and 25 or 80 mM acetic acid in 0 and 50 or 80% MeOH. (A) Experimental degrees
of acetylation DA (1 − DDA) values as a function of time and fit using the system of
differential Eqs. (6), (7) and (10). The reaction rate constant of reacetylation (kR) val-
ues obtained are given in Table 1 (n = 3, mean ± SD). (B) Theoretical evolution of the
proportion of the reactive neutral form of glucosamine in the course of the reaction,
calculated from fit kR values found in (A) for the same experimental conditions.

1
196
M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
Fig. 5. Comparison of theoretical DAinf values vs. R with the results of Vachoud
et al. (1997). The values of Vachoud et al. were obtained in a different solvent, 1,2-
propanediol at 50% (v/v) with water, using 30 mM chitosan (glucosamine units)
and 42 mM acetic acid. The pKa and reaction rates are affected by the nature of
the solvent but the prediction of our model can be compared with these results to
illustrate the behavior of the reaction as R increases (target DA increases) revealing
the difficulty to reach a DA of 1 for these conditions. Compositions of solutions for
theoretical calculations are chitosan (glucosamine units) 27.5 mM, acetic acid 39
and 80 mM, for 50 and 80% MeOH, respectively. Hydrolysis and reacetylation rate
constants shown in Table 1 were used.
Fig. 4. Experimental and theoretical final DA values (reaction time of 3 h), DAinf, for
various acetic anhydride to glucosamine ratios R (nAcOAc/nGluc) in methanol–water
mixtures (v/v). (A) Comparison of theory and experimental values for solution com-
positions of 27.5 mM chitosan (glucosamine units) and 25 or 80 mM acetic acid
for 0 and 50 or 80% MeOH, respectively and R ranging from 0.075 to 0.45 (n = 3,
mean ± SD). (B) Theoretical values of DAinf values for a larger range of R values.
These curves illustrate the inhibition of the reacetylation reaction by the progres-
sive ionization of chitosan as DA and R increase. For comparison, the final DA values
for the same bimolecular reaction combined with the hydrolysis kinetics in 50%
MeOH but without taking into account the changes in chitosan’s ionization degree
The saturation of the reaction at high R has been reported in
the literature (Vachoud et al., 1997) and it has been recognized
that ionization of the polymer was limiting the reaction and that
steric hindrance may also play a role in this efficiency reduction. In
their study, Vachoud et al. (1997) used conditions similar to those
used here. They used a concentration of chitosan of 0.5% (w/v) with
4
2 mM acetic acid in 50% (v/v) water–1,2-propanediol solvent. 1,2-
Propanediol is more viscous than methanol and has a dielectric
constant similar to methanol’s one. Despite this difference in the
nature of the alcohol used in the reaction, we applied our model
with similar conditions in the presence of 50 and 80% MeOH (v/v)
to compare experimental values of Vachoud et al. (1997) deter-
mined by 1H NMR and infrared IR with our theoretical predictions
(results are shown in Fig. 5). It is worth noting that IR determined
reacetylation levels were closer to modeled values at 50% MeOH
(v/v) than for reacetylation determined by 1H NMR which were
closer to theoretical values obtained at 80% MeOH (v/v). Keeping in
mind that these experimental results were obtained in a different
(
i.e. ˛ = constant) have been included (top curve in dark grey).
acetic anhydride vs. glucosamine (R = 2), a complete reacetylation
of chitosan cannot be attained. A theoretical curve of reacetyla-
tion (in grey) without taking ionization changes into account (i.e.
assuming 1 − ˛ is constant in the course of the reaction) has been
shown for 50% MeOH (v/v) in Fig. 4B to illustrate how progressive
ionization reduces significantly the final DA obtained for higher R
values. In this case (i.e. assuming 1 − ˛ constant), the reacetyla-
tion efficiency at high R values is limited only by acetic anhydride
hydrolysis and the reduction of the concentration of amino
moieties by the reacetylation reaction (the ratio of reaction speed
1
solvent and despite the fact that Vachoud et al. (1997) H NMR and
ꢀ
of the two bimolecular reactions vR/vH = kR[glucosamine]/kH
IR results are somewhat less consistent for high R values, compar-
ison of theoretical and experimental values indicate the model is
decreases as glucosamine is consumed).
Table 2
Theoretical efficiency of reacetylation in water–methanol mixtures.
Solvent composition
Initial fraction of neutral
glucosaminea
Initial ratio of reacetylation rate
to hydrolysis rate
Theoretical maximum
efficiency
(vR(t=0)/vR(t=0) + vH(t=0))
Predicted efficiency for
R = 1 (without/with
buffer)c
b
(
% MeOH (v/v))
1
− ˛ (t = 0)
vR(t=0)/vH(t=0)
(
= kR[NH2]t=0/kꢀ
)
H
0
0
0
0.23
0.40
0.39
8.2
17.8
21.7
89%
95%
96%
47%/74%
67%/83%
78%/87%
5
8
a
Experimental conditions are 27.5 mM glucosamine and 25 or 80 mM acetic acid for 0 and 50% or 80% MeOH (v/v), respectively.
This upper limit is reached only for small extent of reacetylation (i.e. when glucosamine amount and ionization degree are approximately constant in the course of the
b
reaction, see Eq. (8)).
c
The calculations are done with a buffer concentration of 150 mM (acid and conjugated base present in equivalent amount of 150 mM, i.e. initial pH ∼ =pKa) with pKa of
5
.8, 6.1 and 6.3 for 0, 50 and 80% MeOH mixtures, respectively. These particular choices of pKa ensure that the initial degrees of ionization is unchanged vs. the case without
buffer. A further increase of 0.5 units of the pKa’s would result in predicted efficiencies being further improved to 83, 90, 91% (0, 50, 80% MeOH, respectively).

M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
1197
solution in the course of the reaction that results in an increased
degree of ionization of the polymer. The presence of methanol
in the solution improves the reaction efficiency mostly by mod-
ulating the pKa of acetic acid and to a lesser extent by reducing
the effective hydrolysis rate of acetic anhydride. As pointed above,
despite an improved efficiency in presence of methanol or other
alcohols, a large excess of acetic anhydride is required to achieve
high degree of acetylation. One way to circumvent this acidifica-
tion problem and to better control the reacetylation process would
be to buffer the solution to minimize pH changes in the course
of the reaction. Theoretical results in the presence of a buffer
(150 mM) in 0 and 50% MeOH (v/v) are shown in Fig. 6. In Fig. 6A,
the buffer pKa used are such that the initial ionization degree is
the same as in the absence of any buffer. Results show that there
is a significant improvement in the reaction efficiency, especially
in water (improvement in 80% MeOH (v/v) is seen for higher DA
than for 0 and 50% MeOH (v/v) and has been omitted in the fig-
ure for clarity purposes). From these results, one can see that the
use of a buffer in 0% MeOH (v/v) could extend the linear region
(
∼
i.e. ∼constant efficiency) of the curve DA vs. R from DA = 0–0.2 to
0–0.6. The theoretical evolution of the fraction of neutral amino
groups as a function of DA in presence of a buffer vs. unbuffered
system is shown in Fig. 6B and illustrates how the buffer should
limit polymer ionization as the reaction evolves. As expected,
increasing the buffer’s pKa should result in further improve-
ment of the reaction, especially for high extent of reacetylation
(Fig. 6C). Theoretical efficiencies for low extent of reacetylation and
for R = 1 (with and without buffer) are shown in Table 2.
4. Conclusion
We found in this study that reacetylation using acetic anhydride
of chitosan dissolved in water–methanol mixtures using acetic acid
is a fast reaction as compared to acetic anhydride hydrolysis. The
reacetylation reaction rate decreases as the proportion of methanol
increases in the solution and the improved efficiency of the reac-
tion in presence of methanol is mostly due to an increase in the pKa
of acetic acid that limits acidification of the solution and ionization
of the polymer in the course of the reaction. By using a theoretical
model based on second-order kinetics combined with acid–base
dissociation equations of chitosan, we have been able to predict the
saturation of the reacetylation reaction reported in the literature at
high R values without relying on any steric hindrance contribution,
which appears to be unnecessary. Our findings suggest reacetyla-
tion be performed in 80% MeOH (v/v) where the efficiency of the
reaction is predicted to be close to 100% for DA < ∼0.6. Theory indi-
cates that the use of a buffer to limit ionization of the polymer
as the reaction evolves would be beneficial to increase efficiency
when a high extent of reacetylation is targeted or when the reac-
tion is performed in water. Taken together, our results permit a
better understanding and control of chitosan reacetylation.
Fig. 6. . Influence of the presence of a buffer on the kinetics of reacetylation (theoret-
ical) in water and 50% MeOH. Similar results are obtained at 80% MeOH but have not
been included here for clarity. The calculations are done with a buffer concentration
of 150 mM (acid and conjugated base present in equivalent amount of 150 mM, i.e.
initial pH ∼ =pKa). (A) Final DA values as a function of R in methanol–water mixtures
compared to the values in the presence of buffer. The curves predict a significant
improvement on the efficiency of the reaction for large extent of reacetylation. (B)
Evolution of the fraction of neutral glucosamine as a function of DA(t) in the course of
the reaction. The curves show a significantly faster ionization of chitosan in absence
of buffer, as expected (calculations done with R = 4). (C) Comparison of final DA for
different values of pKa, showing that increasing buffer pKa can further improve the
efficiency of the reaction.
Acknowledgments
We would like to acknowledge Sarah Dupuis who performed
some reacetylation experiments and chitosan solubility tests and
Isabelle Richard for acetic anhydride hydrolysis measurements.
Support from the Natural Sciences and Engineering Research Coun-
cil of Canada (NSERC)and from the Fonds Québécoisde la Recherche
sur la Nature et les Technologies (FQRNT) is acknowledged.
able to adequately predict the behavior of the reacetylation reaction
without considering steric hindrance at high DA as a limiting fac-
tor of the reaction that was proposed previously (Qun et al., 2007;
Vachoud et al., 1997).
3.5. Addition of a buffer could increase reaction efficiency
Appendix A. Supplementary data
From the analysis presented above, the main factor limiting the
efficiency of the reacetylation reaction is the acidification of the
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carbpol.2011.08.096.

1
198
M. Lavertu et al. / Carbohydrate Polymers 87 (2012) 1192–1198
References
Lavertu, M., Filion, D. & Buschmann, M. D. (2008). Heat-induced transfer of pro-
tons from chitosan to glycerol phosphate produces chitosan precipitation and
gelation. Biomacromolecules, 9(2), 640–650.
Lavertu, M., Xia, Z., Serreqi, A. N., Berrada, M., Rodrigues, A., Wang, D., et al. (2003).
A validated H-1 NMR method for the determination of the degree of deacety-
lation of chitosan. Journal of Pharmaceutical and Biomedical Analysis, 32(6),
Aiba, S. (1991). Studies on chitosan. 3. Evidence for the presence of random and block
copolymer structures in partially N-acetylated chitosans. International Journal of
Biological Macromolecules, 13(1), 40–44.
Aiba, S. (1992). Studies on chitosan. 4. Lysozymic hydrolysis of partially N-acetylated
chitosans. International Journal of Biological Macromolecules, 14(4), 225–228.
Aiba, S. (1994). Preparation of N-acetylchito-oligosaccharides from lysozymic
hydrolyzates of partially N-acetylated chitosans. Carbohydrate Research, 261(2),
1149–1158.
Maghami, G. G. & Roberts, G. A. F. (1988). Evaluation of the viscometric constants
for chitosan. Makromolekulare Chemie, 189(1), 195–200.
March, J. (1992). Advanced organic chemistry: Reactions, mechanisms and structure
2
97–306.
(
Fourth Edition).
Albright, P. S. & Gosting, L. J. (1946). Dielectric constants of the methanol–water
system from 5 to 55 Deg. Journal of the American Chemical Society, 68, 1061–1063.
Asprey, S. P., Wojciechowski, B. W., Rice, N. M. & Dorcas, A. (1996). Applications of
temperature scanning in kinetic investigations: The hydrolysis of acetic anhy-
dride. Chemical Engineering Science, 51(20), 4681–4692.
Bacarella, A. L., Grunwald, E., Marshall, H. P. & Purlee, E. L. (1955). The poten-
tiometric measurement of acid dissociation constants and pH in the system
methanol–water. pKa values for carboxylic acids and anilinium ions. Journal of
Organic Chemistry, 20, 747–762.
Berger, J., Reist, A., Chenite, A., Felt-Baeyens, O., Mayer, J. M. & Gurny, R. (2005).
Pseudo-thermosetting chitosan hydrogels for biomedical application. Interna-
tional Journal of Pharmaceutics, 288(2), 197–206.
Bosch, E., Bou, P., Allemann, H. & Roses, M. (1996). Retention of ionizable compounds
on HPLC. pH scale in methanol–water and the pK and pH values of buffers.
Analytical Chemistry, 68(20), 3651–3657.
Martinez, A., Chornet, E. & Rodrigue, D. (2004). Steady-shear rheology of concen-
trated chitosan solutions. Journal of Texture Studies, 35(1), 53–74.
Moore, G. K. & Roberts, G. A. F. (1980). Chitosan gels. 2. Mechanism of gelation.
International Journal of Biological Macromolecules, 2(2), 78–80.
Nguyen, S., Winnik, F. M. & Buschmann, M. D. (2009). Improved reproducibility in the
determination of the molecular weight of chitosan by analytical size exclusion
chromatography. Carbohydrate Polymers, 75(3), 528–533.
Ottoy, M. H., Varum, K. M. & Smidsrod, O. (1996). Compositional heterogene-
ity of heterogeneously deacetylated chitosans. Carbohydrate Polymers, 29(1),
1
7–24.
Park, J. W., Choi, K. H.
related properties of chitosan. Bulletin of the Korean Chemical Society, 4(2),
8–72.
&
Park, K. K. (1983). Acid–base equilibriums and
6
Qun, G., Ajun, W. & Yong, Z. (2007). Effect of reacetylation and degradation on the
chemical and crystal structures of chitosan. Journal of Applied Polymer Science,
Canals, I., Oumada, F. Z., Roses, M.
& Bosch, E. (2001). Retention of ioniz-
1
04(4), 2720–2728.
able compounds on HPLC. 6. pH measurements with the glass electrode in
methanol–water mixtures. Journal of Chromatography A, 911(2), 191–202.
Cleland, F. A. & Wilhelm, R. H. (1956). Diffusion and reaction in viscous-flow tubular
reactor. AIChE Journal, 2, 489–497.
Dal Pozzo, A., Vanini, L., Fagnoni, M., Guerrini, M., De Benedittis, A. & Muzzarelli, R. A.
A. (2000). Preparation and characterization of poly(ethylene glycol)-crosslinked
reacetylated chitosans. Carbohydrate Polymers, 42(2), 201–206.
Filion, D., Lavertu, M. & Buschmann, M. D. (2007). Ionization and solubility of
chitosan solutions related to thermosensitive chitosan/glycerol–phosphate sys-
tems. Biomacromolecules, 8(10), 3224–3234.
Hirano, S., Kondo, S. & Ohe, Y. (1975). Chitosan gel. A novel polysaccharide gel.
Polymer, 16(8), 622.
Inczédy, J., Lengyel, T., Ure, A. M., & International Union of Pure and Applied Chem-
istry. (1998). Compendium of analytical nomenclature: Definitive rules 1997. Osney
Mead, Oxford; Malden, MA: Blackwell Science.
Knaul, J. Z., Kasaai, M. R., Bui, V. T. & Creber, K. A. M. (1998). Characterization of
deacetylated chitosan and chitosan molecular weight review. Canadian Journal
of Chemistry-Revue Canadienne De Chimie, 76(11), 1699–1706.
Kurita, K., Sannan, T. & Iwakura, Y. (1977). Studies on chitin. 4. Evidence for formation
of block and random copolymers of N-acetyl-d-glucosamine and d-glucosamine
by hetero- and homogeneous hydrolyses. Makromolekulare Chemie, 178(12),
Sashiwa, H., Saimoto, H., Shigemasa, Y., Ogawa, R. & Tokura, S. (1991). Distribution
of the acetamide group in partially deacetylated chitins. Carbohydrate Polymers,
1
6(3), 291–296.
Sashiwa, H., Saimoto, H., Shigemasa, Y. & Tokura, S. (1993). N-acetyl group distribu-
tion in partially deacetylated chitins prepared under homogeneous conditions.
Carbohydrate Research, 242, 167–172.
Shedlovsky, T.
& Kay, R. L. (1956). The ionization constant of acetic acid in
water–methanol mixtures at 25 Deg from conductance measurements. Journal
of Physical Chemistry, 60, 151–155.
Shigemasa, Y., Saito, K., Sashiwa, H. & Saimoto, H. (1994). Enzymatic degradation
of chitins and partially deacetylated chitins. International Journal of Biological
Macromolecules, 16(1), 43–49.
Vachoud, L., Zydowicz, N. & Domard, A. (1997). Formation and characterisation of a
physical chitin gel. Carbohydrate Research, 302(3–4), 169–177.
Varum, K. M., Anthonsen, M. W., Grasdalen, H. & Smidsrod, O. (1991). Determina-
tion of the degree of N-acetylation and the distribution of N-acetyl groups in
partially N-deacetylated chitins (chitosans) by high-field N.M.R. spectroscopy.
Carbohydrate Research, 211(1), 17–23.
Xu, X. L., Li, H. Q., Zhang, Z. D. & Qi, X. (2009). Hydrodynamic properties of aqueous
dextrin solutions. Journal of Applied Polymer Science, 111(3), 1523–1529.
3
197–3202.