Two competing reactions of tetrabutylammonium alginate in organic solvents: Amidation versus γ-lactone synthesis

Article abstract of DOI:10.1016/j.carbpol.2015.11.070

Biocompatibility and thickening properties predetermine alginates as ingredients in food, cosmetic and pharmaceutical products. Further chemical modifications are often desired for a product optimization. The introduction of hydrophobic groups can be realized by employing organic tetrabutylammonium alginate (TBA-Alg) solutions. The synthesis of alginic acid alkyl amides from TBA-Alg with 2-chloro-1-methylpyridinium iodide (CMPI) as a coupling agent, however, has so far not resulted in a high degree of amidation. The analysis of the coupling reaction revealed the formation of mannuronic acid γ-lactone structures, which required a conformation change from ¹C₄ to ⁴C₁. The opening of the γ-lactone required a high excess of butylamine. In the case of CMPI, triethylamine had to be added prior to the coupling agent in order to suppress the assumed alginic acid formation. The degrees of amidation achieved were up to 0.8, and for propylphosphonic anhydride as the coupling agent up to 1. The molecular weights of the alginic acid butyl amide were ≥35 kDa.

Full text of DOI:10.1016/j.carbpol.2015.11.070

Carbohydrate Polymers 138 (2016) 244–251
Contents lists available at ScienceDirect
Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol
Two competing reactions of tetrabutylammonium alginate in organic
solvents: Amidation versus ␥-lactone synthesis
a,∗
a
b
Thomas Schleeh , Mathieu Madau , Dierk Roessner
a
Environmental Research and Innovation Department, Luxembourg Institute of Science and Technology, 41 rue du Brill, L-4422 Belvaux, Luxembourg
Wyatt Technology Europe GmbH, Hochstrasse 12a, D-56307 Dernbach, Germany
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 August 2015
Received in revised form
Biocompatibility and thickening properties predetermine alginates as ingredients in food, cosmetic and
pharmaceutical products. Further chemical modifications are often desired for a product optimization.
The introduction of hydrophobic groups can be realized by employing organic tetrabutylammonium
alginate (TBA-Alg) solutions. The synthesis of alginic acid alkyl amides from TBA-Alg with 2-chloro-1-
methylpyridinium iodide (CMPI) as a coupling agent, however, has so far not resulted in a high degree
of amidation. The analysis of the coupling reaction revealed the formation of mannuronic acid ␥-lactone
1
8 November 2015
Accepted 26 November 2015
Available online 29 November 2015
1
4
structures, which required a conformation change from C4 to C1. The opening of the ␥-lactone required
a high excess of butylamine. In the case of CMPI, triethylamine had to be added prior to the coupling
agent in order to suppress the assumed alginic acid formation. The degrees of amidation achieved were
up to 0.8, and for propylphosphonic anhydride as the coupling agent up to 1. The molecular weights of
the alginic acid butyl amide were ≥35 kDa.
Keywords:
Mannuronic acid ␥-lactone
Conformation change
2
-Chloro-1-methylpyridinium iodide
Propylphosphonic anhydride
Alginic acid alkyl amide
© 2015 Elsevier Ltd. All rights reserved.
1
. Introduction
& Edgar, 2013; Pelletier, Hubert, Lapicque, Payan, & Dellacherie,
2
000). In contrast, the amidation of TBA-Alg in organic solvents did
Alginates are unbranched polysaccharides made of (1 → 4)
not always lead to satisfying products, especially for the attachment
of longer alkylamines such as dodecylamine (Barbucci, Consumi, &
Magnani, 2002; Barbucci et al., 2006; Leone, Torricelli, Chiumiento,
Facchini, & Barbucci, 2008; Vallée et al., 2009).
linked ␤-D-mannuronic acid (M) and its C5 epimer ␣-L-guluronic
acid (G) blocks. Excellent gelling properties and biocompatibility
resulted in numerous applications in the food and pharmaceutical
industry, such as thickening or gelling agents (Saha & Bhattacharya,
In comparison to the esterification, the amidation required
the activation of TBA-Alg with the coupling agent 2-chloro-1-
methylpyridinium iodide (CMPI) in a first step, followed by the
amidation in the presence of the hydrogen scavenger triethylamine.
The drawback of this reaction is that an intra- and intermolecular
ester formation has to be taken into account, due to the reaction of
CMPI activated alginate with itself (Saigo, Usui, Kikuchi, Shimada,
& Mukaiyama, 1977; Vallée et al., 2009). The intramolecular ester-
ification would result in lactone structures, which were already
synthesized from sodium alginate under dehydrating conditions
(Anson, Novikova, & Iozep, 2009). The intermolecular esterifi-
cation would lead to a three dimensional network formed by
the crosslinking reaction between carboxylic acids and hydroxyl
groups of different alginate moieties. These interesting hypothe-
ses led to the question of whether a better understanding of these
side reactions could be exploited for the development of a more
efficient amidation of alginates and for the evaluation of the cou-
pling agent performance. Therefore, the reactions of TBA-Alg with
CMPI and propylphosphonic anhydride (T3P) (Nagendra, Madhu,
Vishwanatha, & Sureshbabu, 2012; Wissmann & Kleiner, 1980) as
2
010), wound dressing materials (Schultz et al., 2003) and drug
delivery systems (Anal & Singh, 2007; Tonnesen & Karlsen, 2002).
Chemically modified materials are often of special interest for
further product optimization, especially for pharmaceutical appli-
cations. The chemical modification of alginates (Alg) is mostly
achieved under aqueous conditions by exploiting the excellent
water solubility of sodium alginate (Na-Alg) (Pawar & Edgar, 2012;
Yang, Xie, & He, 2011). This approach is excellent for introduc-
ing hydrophilic groups, but has disadvantages for introducing
hydrophobic ones. This obstacle can be resolved by converting
Na-Alg into tetrabutylammonium alginate (TBA-Alg), which is sol-
uble in polar aprotic organic solvents (Schleeh, Madau, & Roessner,
2
014). Recently, the esterification of TBA-Alg in organic solvents
with alkyl halides was reported to be efficient in terms of yield
and degree of substitution (Della Valle & Aurelio, 1992; Pawar
∗ Corresponding author.
E-mail address: thomas.schleeh@list.lu (T. Schleeh).
http://dx.doi.org/10.1016/j.carbpol.2015.11.070
0
144-8617/© 2015 Elsevier Ltd. All rights reserved.

T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
245
coupling agents were first studied in the absence of amines. Then,
the amidation of TBA-Alg with n-butylamine and CMPI or T3P as
coupling agent was investigated and optimized by considering the
findings from the reactions of the TBA-Alg with the coupling agents.
3.27%. Degree of substitution with TBA (DSTBA) 0.98. Yield: 2.39 g
(91.3%).
2.2.3. Ester synthesis (side product)
The studied parameters of this reaction were the coupling agent
(
1
none, CMPI and T3P), the TBA-Alg concentration (0.5 wt.-% to
0 wt.-% for CMPI, to 9 wt.-% for T3P), the ratio of TBA-Alg:coupling
2
. Materials and methods
agent (1:0.25 to 1:2 for CMPI and 1:0.25 to 1:10 for T3P) and
the reaction duration (0.5 h to 24 h, ambient temperature). For
the reaction with CMPI, TBA-Alg (6.4 mmol) and coupling agent
2
.1. Chemicals
H-Alg from brown algae was purchased from Sigma (St. Louis,
(6.4 mmol) were separately dissolved in DMF (TBA-Alg: 220 mL;
MO, USA; Na-Alg residue: 11% and M/G ratio: between 0.61/0.39
and 0.54/0.46 (Schleeh et al., 2014). The Na-Algs Manucol DM (M/G
ratio: 0.61/0.39) and Protanal GP 1740 (M/G ratio: 0.37/0.63) from
FMC BioPolymer were received as a gift from IMCD Benelux N.V.
coupling agent: 50 mL) at ambient temperature. The TBA-Alg solu-
◦
tion was cooled down to 0 C before the coupling agent solution
was added drop-by-drop under stirring. The reaction mixture was
◦
kept at 0 C for 1 h before the ice bath was removed. Then, the
(
Mechelen, Belgium). Potassium bromide (KBr) and T3P in DMF
reaction mixture was allowed to warm up to ambient temperature
under continuous agitation overnight. The products were isolated
either by hydrolysis of the remaining activated alginate moieties
and replacing the remaining TBA+ by Na+ ions (procedure (b)) or
without (procedure (a)). Procedure (a): The reaction mixture was
precipitated directly in heavily stirred absolute ethyl acetate and
then allowed to settle for 30 min. The product was then filtrated,
purified in a soxhlet with acetone and vacuum dried. All samples
were kept in a desiccator until further use. Procedure (b): A 2.5 M
NaCl solution (80 mL) was added to the reaction mixture drop-by-
drop. Then, the product was precipitated in heavily stirred aqueous
ethanol (1100 mL, ethanol/water (7/1)), followed by a period with-
out agitation of 30 min, in which the polymer was allowed to settle
down. The product was filtrated, purified by a soxhlet extraction
with acetone and finally vacuum dried. A few selected samples were
additionally purified to remove the coupling agent and solvent
traces. This was done in a stirred cell equipped with a membrane
(
c(T3P): ∼0.586 mol/L) were obtained from ACROS (Geel, Belgium),
∼
40% tetrabutylammonium hydroxide (TBAOH) solution from
Fluka (Buchs, Switzerland), CMPI, butylamine, dimethylformamide
(
DMF; water ≤0.005%), sodium chloride (NaCl) and triethylamine
from Sigma–Aldrich, acetone, DMF (water ≤0.01%), ethanol (99%),
hydrochloric acid (HCl; 1 M), ethyl acetate (water < 0.1%) and formic
acid from VWR (Radnor, PA, USA), sodium hydroxide solution
(
0.1 N, Titrisol) from Merck (Darmstadt, Germany), ethylenedi-
aminetetraacetic acid (EDTA) standard from the Leco Corporation
St. Joseph, MI, USA) and nitrogen (␣1) from Air Liquide (Paris,
(
France). Milli-Q water from Merck was used in general. All materials
were used without any further purification.
2
.2. Synthesis
In general, commercially available H-Alg was used for the inves-
tigation of the reaction of TBA-Alg and the coupling agent. Changing
the M/G ratio required the synthesis of TBA-Alg from specific Na-
Algs (Manucol and Protanol), which was done according to the
recently reported procedure (Schleeh et al., 2014). The amidation
of TBA-Alg was studied on highly DMF soluble TBA-Alg, which was
also prepared according to the same procedure.
(molecular cut-off: 10 kDa) with Milli-Q water (at least 10 times
the stirred cell volume). These samples were finally freeze-dried
and kept in the desiccator, as with all other samples.
In case where the coupling agent was absent (negative con-
trol), TBA-Alg was directly dissolved in the complete amount of
DMF (270 mL). Working with T3P as a coupling agent occasionally
required the solvent volume for dissolving TBA-Alg to be adjusted
to keep the desired TBA-Alg concentration, because T3P was deliv-
ered as a DMF solution.
2
.2.1. H-Alg synthesis in brief
Na-Alg (4 g, 20 mmol) was dispersed in aqueous formic acid
◦
(
200 mL, water content 30% (v)) at 4 C. The dispersion was stirred
◦
at 4 C for 1 h before the ice bath was removed. The reaction mixture
was then allowed to warm up by stirring for 6 h or overnight. The
product was filtrated, washed with aqueous ethanol (3 × 100 mL,
ethanol/water: 70% (v)/30% (v)) and then with acetone (100 mL),
dried under vacuum and kept in a desiccator over a drying agent
until further use. FTIR (KBr pellet): ꢀ(OH) 3900–3064 cm , ꢀ(CH)
3
ꢀ
1
2
3
ꢀ
.2.3.1. Negative
control. FTIR
(KBr
pellet)):
ꢀ(OH)
−1
−1
−1
717–3079 cm , ꢀ(CH) 2978–2819 cm , (sat COOH) 1738 cm
,
,
−
−1
−
−1
−1
as(COO ) 1614 cm , ꢀs(COO ) 1416 cm , (C O) 1258 cm
−1
−1
(C
O
C) 1094 cm , (C OH) 1026 cm . Elemental Analysis: C
3
2.02%, H 4.43%. Yield: 1.4 g.
−
1
−
1
−1
−1
−
−1
064–2787 cm , (sat COOH) 1744 cm , ꢀas(COO ) 1636 cm ,
2
.2.3.2. Ester synthesis with procedure (b) and CMPI as a cou-
−
−
−1
−1
s(COO ) 1412 cm , (C O) 1323 cm , (C
O
C) 1180 cm and
101 cm , (C OH) 1032 cm . Elemental Analysis: C: 41.00%; H:
.92%. Degree of acidification (DSH): 0.94 (Titration with 0.1 N
−1
pling agent. FTIR (KBr pellet): ꢀ(OH) 3726–3070 cm , ꢀ(CH)
2
ꢀ
1
1
−1
−1
−1
−1
991–2828 cm , (␥-lactone) 1780 cm , (sat COOH) 1742 cm ,
4
− −1 − −1
as(COO ) 1618 cm , ꢀs(COO ) 1418 cm , (C O) 1248 and
NaOH). Yield: 2.91 g (93.8%).
−1 −1 −1
207 cm , (C O C) 1096 cm , (C OH) 1036 cm . Elemental
Analysis: This could not be meaningfully performed due to limited
purity. Yield: 1.07 g.
2.2.2. TBA-Alg synthesis
H-Alg (1.12 g, 6.4 mmol) was dispersed in 100 mL Milli-Q
water, followed by a pH adjustment to between 7 and 9 with
a TBAOH solution (approximately 0.5 M). The solution was then
immediately frozen to avoid any carbon dioxide absorption,
lyophilized and kept in a desiccator over a drying agent until
2.2.3.3. T3P as
a
coupling agent. FTIR (KBr pellet): ꢀ(OH)
−
1
−1
−1
−1
3717–3070 cm , ꢀ(CH) 2988–2840 cm , (␥-lactone) 1782 cm
,
,
(sat COOH and sat. COOR) 1742 cm−1, ꢀas(COO ) 1622 cm
−
−
−1
−1
−1
ꢀ
s(COO ) 1416 cm
,
(C O) 1246 and 1205 cm
,
(C
O
C)
−
1
−1
further use. FTIR (casted film): ꢀ(OH) 3685–3041 cm (with max-
1092 cm , (C OH) 1030 cm . Elemental Analysis: This could not
be meaningfully performed due to limited purity. Yield: 0.97 g.
−
1
−1
−1
−1
ima at: 3501 cm , 3383 cm and 3200 cm ), ꢀ(CH) 2961 cm ,
−
1
−1
−
−1
2
1
1
937 cm and 2879 cm , ꢀas(COO ) 1611 cm , ı(CH2 and CH )
3
−
−
1
1
−1
−1
−1
485 cm
285 cm , (C
and 1468 cm , ıs(CH ) 1385, (C O) 1317 cm
and
and
2.2.4. Amidation of TBA-Alg in organic polar aprotic solvents
A three-neck flask was purged with nitrogen for 15 min. TBA-
Alg (7.2 mmol) was then placed in this flask, dissolved in DMF
3
−1
−1
and 1101 cm
O
C) 1173 cm , 1148 cm
−
1
(
C
OH) 1036 cm . Elemental analysis: C 61.92%, H 10.35%, N

2
46
T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
(
45 mL for the reaction with CMPI or 55 mL for the reaction with
with ˛ = N/C (determined by elemental analysis) and I = number
of carbon atoms of the substituent (TBA: I = 16; CMPI: I = 6; buty-
lamine: I = 4).
◦
T3P) and cooled down to 0 C. The dissolved coupling agent (CMPI:
7
.2 mmol in 11 mL DMF or T3P: 14.4 mmol in 4.8 mL DMF) was
added drop-by-drop under agitation and left to react by stirring
gently at this temperature for 1 h. Butylamine (CMPI: 14.4 mmol
or T3P: 72 mmol) and triethylamine (7.2 mmol) were added with
a micropipette immediately one after another. The reaction mix-
ture was stirred again gently at 0 C for 1 h, before the ice bath
was removed. The reaction mixture was agitated continuously for
2.4. FTIR
256 spectra were recorded for each sample with a Vertex 70
FTIR from Bruker (Billerica, MA, USA) in the wavenumber range
◦
−
1
−1
between 4000 and 400 cm with a resolution of 4 cm in trans-
mission mode, using a sample holder with an aperture diameter
of 5 mm. Humidity and carbon dioxide compensation, rubber band
base line correction and a normalization of the spectra were carried
out before signal evaluation. Some signals were identified on the
basis of the analysis of the second derivative of the FTIR spectrum,
because they were not fully resolved or were hidden in the spec-
trum. The signal assignment of FTIR averaged spectra was carried
out according to Hesse, Meier and Zeh (Hesse, Meier, & Zeh, 2008).
The ␥-lactone content was estimated from the lactone absorbance
(A) at 1782 cm by assuming that the sum of the absorbance sig-
nals of the lactone at 1782 cm , the carboxylic acid at 1744 cm
and carboxylate groups at 1611 cm (formula (2)) were represent-
ing all carbonyl carbons (equivalent to all alginate moieties) of the
polymer.
2
0 h to warm up to ambient temperature. Then next day, a sodium
ion rich phosphate buffer solution (NaCl, Na HPO , NaH PO ,
2
4
2
4
+
2−
c(Na ) ≥ 2.5 M, c(PO
) = 0.5 M, pH: 7.5; 29 mL) was slowly added.
The polymer was precipitated in heavily stirred cold aqueous
ethanol (280 mL, 70% ethanol), followed by a centrifugation step
4
− ◦
2
(
g-force: 3488 ms , 5 min, 5 C). The solvent fractions were col-
lected and unified and the solvent was removed. The remainder
was transferred into acetone (500 mL) to precipitate the polymer,
centrifuged, washed with acetone four times and dried under vac-
uum. The dry sample was dispersed in a sodium phosphate buffer
−
1
3
−
Na HPO , NaH PO , c(PO
4
−1
−1
(
) = 0.1 M, pH: 7–8) and let to dis-
2
4
2
4
−
1
solve overnight. The solution was then transferred to a stirred cell
equipped with a membrane with a 10 kDa molecular cut-off and
washed with Milli-Q water (at least 10 times the stirred cell vol-
ume). Finally, the product was lyophilized and kept in a desiccator
until further use.
A(carbonylcarbons) = A(ꢁ-lactone) + A(carboxylicacid)
+
A(carboxylate) = 1
(2)
2
.2.4.1. Alginic acid butyl amide (AlgNC4 synthesized with CMPI
1
as coupling agent. H NMR (D O): ı 5.66–4.34 ppm (H1-5), ı
2
3
.84 ppm (H7), ı 1.96 ppm (H8), ı 1.93 ppm (H9), ı 1.51 ppm (H10),
Degree of substitution with amide groups (DSamide): 0.43. 13C NMR
D O): ı 175.56 ppm (carboxylate C6), ı 170.38 ppm (amide C6),
2.5. NMR spectroscopy
2.5.1. 13C solid-state
The spectra were acquired with an Avance I, 400 MHz spectrom-
eter from Bruker, equipped with a magic angle spinning (MAS)
probe. The instrument was operated at a resonance frequency of
(
2
ı 100.7 ppm (C1), ı 80.71–66.97 ppm (C2–C5), ı 39.91 ppm (C7),
ı 31.16 ppm (C8), ı 20.13 (C9), ı 13.60 ppm (C10). FTIR (KBr pel-
−
1
1
−1
−1
−1
let): ꢀ(OH) 3710–3008 cm , ꢀ( CO NH ) shoulder at 3110 cm
−
2nd derivative 3095 cm ), ꢀ(CH (of CH2 and CH )) 2961 cm
3
(
−
1
−1
100.62 MHz (B0 = 9.4 T) and with 4 mm rotors at a rotation speed of
and 2876 cm , ꢀ(CH (of Alg)) 2930 cm , ( CO NH ) 1657 cm
and 1558 cm , ꢀas(COO ) 1616 cm , ı( CH2 ) 2nd derivative at
−
1
−
−1
12.5 kHz. The spectra were acquired via cross polarization (1H–13C)
−
1
−1
−1
O C) 1092 cm . Elemental
and total decoupling of 1H at 85 kHz. The contact time was 1 ms and
1
466 cm , ıs( CH3 ) 1377 cm , (C
Analysis: C 40.98%, H 5.52%, N 2.56%, DSamide: 0.41. Yield: 1.38 g
88.6%).
the relaxation time between the pulses 5 s. The spectrum of the
material received from the reaction with CMPI was averaged from
(
2
048 scans and that of the material received from the reaction with
2
(
2
.2.4.2. AlgNC4 synthesized with T3P as coupling agent: 1H NMR
DMSO-d6). ı 4.86–3.35 ppm (H1-5), ı 3.11 ppm (NH and H7), ı
.5 ppm (OH), ı 1.45 ppm (H8), ı 1.31 ppm (H9), ı 0.90 ppm (H10),
T3P from 1024 scans.
2
.5.2. 1H and 13C NMR
1
3
Deuterium oxide and DMSO-d6 were chosen as solvents in
DSamide: 0.98. C NMR (DMSO-d6): ı 168.12 ppm and 167.39 ppm
dependence of the solubility for preparing 3–4% (w/v) solutions.
The NMR spectra were acquired with a DRX-500 NMR spectrome-
(
(
C6), ı 99.86 ppm (C1), ı 77.62–66.11 ppm (C2-C5), ı 37.95 ppm
C7), ı 30.56 ppm (C8), ı 19.09 ppm (C9), ı 13.02 ppm (C10). FTIR
1
13
− −1
1
ter from Bruker at 500 MHz ( H) and at 125 MHz ( C) at ambient
temperature. The spectrum of the AlgNC4 synthesized with CMPI
(
(
KBr pellet): ꢀ(OH) 3700–3017 cm , ꢀ( CO NH ) 3109 cm
− −1
2nd derivative 3090 cm ), ꢀ(CH (of CH and CH )) 2961 cm and
2 3
1
1
13
−
1
−1
−1
as coupling agent was averaged from 16 ( H) and 7770 scans ( C)
2
1
1
0
874 cm , ꢀ(CH (of Alg)) 2936 cm
,
CO NH ) 1665 cm and
1
−
1
−1
−1
and that of the reaction with T3P as coupling agent from 16 ( H)
551 cm , ı( CH2 ) 1466 cm , ıs( CH3 ) 1379 cm , (C
O C)
20,480 scans (13C). The DS was calculated on the basis of 1H NMR
spectra by evaluating the ratio of the signal integrals of a represen-
−
1
097 cm . Elemental analysis: C 49.01%, H 7.31%, N 5.58%. DSamide
.96. Yield: 1.28 g (77.3%).
:
tative group (I ) for the substituent and for the complete polymer
S
(IR) according to formula (3).
2.3. Elemental analyses
x ∗ I
S
DSamide = (
.
(3)
The samples were vacuum dried and the analyses were per-
z ∗ IR − y ∗ I )
S
formed on a True Spec instrument from the Leco Corporation. The
◦
The signal of methyl group of the substituent aliphatic
chain (D2O at 1.51 ppm or DMSO-d6 at 0.90 ppm) representing
furnace and afterburner temperatures were set at 950 C and at
◦
8
50 C, respectively and the instrument was calibrated with EDTA.
3
× DS
protons (z = 3) was integrated for the determination
The degree of substitution (DS) of nitrogen containing alginates
TBA-Alg, CMPI activated Alg and Alg amide) was calculated based
on the N/C ratio according to the formula (1) (Hu et al., 2013).
amide
of I Where D O was a solvent, the spectral range of the alginate
S.
2
(
backbone (∼6.0 and 4.1 ppm) was integrated for IR. It represented
5
protons (x = 5, y = 0), since the polymer was dissolved in D O
2
(
˛ ∗ 6M )
C
and a complete exchange of the hydroxyl and amide protons by
deuterium was assumed. In the case of DMSO-d6, the selected
DS = (
(1)
MN − ˛ ∗ I ∗ M )
C

T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
247
Fig. 1. On the left side the 1C4 chair conformation of the guluronic ring is drawn with carboxylate and hydroxyl groups on opposite sides of the ring plane. In the middle the
4
conformations of the mannuronic acid ring with carboxylic and two hydroxyl groups are on the same side of the ring plane. The C1 chair conformation (left) is the preferred
1
one, with an equatorial positioning of the carboxylate group (C6). The carboxylate group and the hydroxyl group are in an axial position in the C4 chair conformation (right),
which enables the formation of a ␥-lactone, as displayed on the right hand side.
spectral range (∼5.4 to 2.7 ppm) for IR comprised the integrals of
the alginate backbone, the amide proton and the C7 methylene
group (3.11 ppm) due to signal overlaps, but not the protons of the
hydroxyl groups attached to the backbone (2.5 ppm). Five protons
signal could also be interpreted as resulting from ester groups, but
this reaction was considered too unlikely under these conditions.
3
3
.3. Activated intermediates
contributed to the alginate backbone signals (x = 5) and 3 × DS
amide
protons to the substituent signals (y = 3).
.3.1. Reaction and isolation under non-hydrolytic conditions
The analysis of the activated alginate intermediate required
2
.6. Size exclusion chromatography – multi angle light scattering
isolation and purification in the absence of water to avoid any
hydrolysis, which was achieved by the following purification pro-
cedure (a). The 13C solid state NMR spectra of both intermediates
Fig. 3) contained the expected signals of coupling agents, but did
not provide any information about potential side product.
The corresponding FTIR spectra confirmed the findings of the
(
SEC-MALS)
The chromatographic system consisted of a pump (Agilent
(
G1311B, Santa Clara, CA, USA) with an integrated online degasser,
and a temperature controlled autosampler (Agilent G1329B) kept at
2
◦
2 C. One 7.8 mm OH pack SEC column (Shodex SB-806 M, SHOWA
13
C solid state NMR spectra with signals from the aromatic ring in
DENKO EUROPE GmbH, Munich, Germany) was employed. A DAWN
HELEOS II multi-angle light scattering detector (Wyatt Technology
Corporation, Santa Barbara, CA, USA) and an Optilab T-rEX (Wyatt
Technology Corporation) differential refractive index detector were
connected sequentially to the SEC column. The aqueous mobile
phase consisted of 100 mM NaCl with 0.02% sodium azide, at a
flow rate of 0.5 mL/min, using the isocratic mode at room tem-
perature. All reagents for SEC were of HPLC grade and the mobile
phase was filtered through Durapore VVPP 0.1 ␮m membrane fil-
ters (Millipore). The injection volume was 20 ␮L. Data collection
and processing were performed using the ASTRA software, Version
the case of CMPI and signals from the propyl groups in the case of
T3P. The spectra further supported the hypothesis of a side reaction,
−1
as strong signals of ␥-lactone structures were found at 1782 cm
−1
in the case of the CMPI intermediate and at 1792 cm for the T3P
intermediate. Deriving any quantitative information about the effi-
ciency of activation and lactone formation was impossible, as the
FTIR spectra quality was poor due to the limited purification in
procedure (a).
3.3.2. Isolation under hydrolytic conditions
Samples isolated under hydrolytic conditions according to pro-
cedure (b) resulted in better FTIR spectra. Samples made by using
6
.1.1.17 (Wyatt Technology Corporation). Zimm formalism with
first order regression was used for the processing of light scattering
data.
3
. Results and discussions
3.1. Ring conformations
The formation of a lactone requires the positioning of the car-
boxylic group and a hydroxyl group in axial position on the same
side of the ring plane (Fig. 1). Therefore, the formation of a guluronic
lactone is impossible due to the positioning of these groups on
opposite sides of the ring plane. In the case of the mannuronic acid
ring the groups are located on the same side of the ring plane, but
the carboxylic group is in an equatorial position for the assumed
4
thermodynamic preferred conformation C1. Bringing the man-
nuronic acid ring into the 1C4 conformation moves the carboxylic
(
C6) and the C3 hydroxyl groups into axial positions, which enables
the formation of the ␥-lactone.
3.2. Negative control
TBA-Alg did not react in the absence of coupling agents under
Fig. 2. FTIR spectra of the products, which were obtained from the negative control
(
lower) and from the reaction in the presence of the coupling agents 2-chloro-
-methylpyridinium iodide (middle) or propylphosphonic anhydride (upper). All
samples were isolated according to method (a). Wavenumbers correspond at
the given conditions, as the expected lack of a lactone signal in
1
−
1
the FTIR spectra (Fig. 2) approved. The small signal at 1739 cm
was assigned to carboxylic acid structures, which resulted from the
treatment with aqueous solutions at the end of the reaction. This
−1
−1
1
782 cm
to ␥-lactones, at 1744 cm to carboxylic acids and/or esters and at
−
1
1611 cm to carboxylate ions.

2
48
T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
signal and the absence of coupling agent signals in these spectra
were correlated with an unexpected resistance of the ␥-lactone
to hydrolysis and to a fast and efficient scission of the coupling
agent complex. Furthermore, the reproducibility of the normalized
FTIR absorbance for the ␥-lactone, carboxylic acid or ester and car-
boxylate ion was also good from batch to batch. This allowed a
comparison of the different materials synthesized by applying the
method described above (formula (2)). A variation of the TBA-Alg
source resulted in a change of the normalized absorbance values
(Table 1) due to the different M/G block structures and differ-
ent M/G contents in these samples. The synthesis with TBA-Alg
made from an Alg with a low content of mannuronic acid moieties
resulted in a lower ␥-lactone absorbance than expected.
Optimizing the ␥-lactone synthesis revealed highest yields for
CMPI with a TBA-Alg concentration of 5% (w/v), a TBA-Alg:CMPI
ratio between 1:0.5 and 1:1, a reaction duration of at least 30 min
(test conditions and FTIR results are given in Table S1 of the
supplementary data). These conditions led only to a marginal
improvement of the ␥-lactone content of about 0.377 (61.8% yield)
in comparison to 0.349 (57.1% yield) for the standard conditions.
Best conditions for the synthesis of the ␥-lactone with T3P as
coupling agent were a TBA-Alg concentration of 1 or 2% (w/v), a
TBA-Alg:T3P ratio of 1:2 with a reaction duration of at least 1 h.
The lactone content increased from 41.6% (± 1.2%) under standard
conditions to 50.7% (84% of mannuronic acid moieties).
Fig. 3. 13C solid state NMR Spectra of 2-chloro-1-methylpyridinium iodide (upper
spectrum) and propylphosphonic anhydride activated alginates (lower spectrum)
isolated according to method (b). Carbon atoms of the tetrabutylammonium ion
ꢀ
ꢀ
were marked as C-1 to C-4 in the lower spectrum.
one or the other coupling agent were giving very similar spectra
−
1
−
1
At a first glance, the signal at 1744 cm in the received FTIR
spectra seemed to support the proposed formation of an inter-
molecular alginate–alginate ester (Vallée et al., 2009), because the
(
(
Fig. 3). Signals of ␥-lactone (1782 cm ), carboxylic acid or ester
both at 1744 cm 1) and carboxylate ions moieties (1611 cm−1
−
)
were found for all materials. The presence of a clear ␥-lactone
Table 1
Normalized FTIR absorbance values of the ␥-lactone (1782 cm 1), carboxylic acid respectively ester (1744 cm−1) and carboxylate (1611 cm ) groups and the corresponding
−
−1
␥
-lactone yields of mannuronic ␥-lactones obtained from the reaction between 2-chloro-1-methylpyridinium iodide (CMPI) or propylphosphonic anhydride (T3P) as coupling
a
agents and tetrabutylammonium alginates (TBA-Alg) of different origins. Coupling agents were removed by a short treatment with a sodium chloride solution.
TBA-Alg
Source
Intramolecular reaction
M/G ratiob
Coupling agent
FTIR: Normalized absorbance values
Mannuronic ␥-lactone yieldc
␥
-Lactone
Carboxylic acid or ester
Carboxylate
Commercially available
alginic acid
Protanal GP 1740
0.61/0.39
0.37/0.63
0.61/0.39
CMPI
T3P
CMPI
T3P
CMPI
T3P
0.349
0.416
0.284
0.344
0.384
0.510
0.328
0.325
0.356
0.449
0.308
0.295
0.323
0.259
0.360
0.208
0.309
0.196
57.1%
68.2%
76.9%
92.9%
62.9%
83.5%
Manucol DM
a
c(TBA-Alg) = 1% (w/v) in DMF.
M: content of mannuronic acid moieties, G: content of guluronic moieties.
b
c
␥
-lactone yield = (normalized ␥-lactone absorbance) × 100/M.
Table 2
Degree of amidation and molecular weight analysis for alginic acid butyl amide obtained from reactions with 2-chloro-1-methylpyridinium iodide (CMPI) or propylphosphonic
anhydride (T3P) as a coupling agent under various reaction conditions.
Coupling agent
TBA-Alg in DMF
% (w/v))
Ratio: TBA-Alg to coupling agent
Ratio: TBA-Alg to n-butylamine
DSamidea
Molecular weight analysis
(
Mn (kDa)
Mw (kDa)
Polydispersity
CMPI
CMPI
CMPI
CMPI
CMPIb
T3P
T3P
T3P
T3P
T3P
0.5
0.5
5
5
5
0.5
0.5
5
5
5
1:1
1:1
1:1
1:1
1:1
1:2
1:2
1:0.5
1:1
1:2
1:2
1:1
1:2
1:2
1:10
1:1
1:1
1:2
1:10
1:2
1:2
0.275
0.259
0.409
0.403
0.68
0.307
0.472
0.391
0.587
0.528
0.961
–
–
–
–
–
–
–
–
–
20.4 ± 1.1
35.5 ± 0.1
1.7 ± 0.1
–
–
–
–
–
–
–
–
–
25.1 ± 2.1
44.1 ± 1.1
1.8 ± 0.1
–
–
–
–
–
–
c
c
c
T3P
5
1:10
46.4 ± 3.5
109.3 ± 9.8
2.4 ± 0.3
d
d
d
40.4
54.7
1.4
a
DSamide calculation based on data from the elemental analysis.
Proton scavenger was added twice, before adding CMPI and before adding butylamine.
Obtained results for the assumed micelle structure.
b
c
d
Estimated result for the single alginic acid butyl amide chain.

T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
249
Fig. 4. 1H NMR (left) and 13C NMR (right) spectra of alginic acid butyl amide obtained from the reactions of tetrabutylammonium alginate with the coupling agent 2-
chloro-1-methylpyridinium iodide (DSamide = 0.41; upper spectra; solvent: D2O) and the coupling agent propylphosphonic anhydride (DSamide = 0.96; lower spectra; solvent:
DMSO-D6).
signal correlates with the carboxylic ester group. The activated
alginate esters moieties did not contribute, because they were no
longer present, due to the hydrolysis with a salt solution. How-
ever, the latter reaction led to the formation of carboxylic acid
moieties, which gave a signal in the same spectral range as esters.
This was also confirmed by putting a sample purified with proce-
dure (a) into water, resulting in an acidic dispersion. Therefore, it
was not possible at this point to confirm or disprove the presence
of alginate–alginate ester structures.
3.5. Amidation
The optimized reaction conditions for the synthesis of the man-
nuronic acid ␥-lactone were assumed to also be the optimum for
3.4. Ester hydrolysis
The intensive purification of dispersed esterified Alg with Milli-
Q water (pH ∼ 6) via ultra-filtration in a stirred cell led to a clear
solution, which was assumed to correspond to the complete hydrol-
ysis of all ester structures. Furthermore, it was deduced that the
␥-lactone structures were sensitive to a nucleophilic attack of the
hydroxyl groups of water under gentle pH conditions as long as
sufficient quantities and time were available. The final clear solu-
tion indicates also that the predicted formation of intermolecular
esters was very unlikely, because the resulting three dimensional
networks were expected to withstand aqueous treatments under
slightly acidic conditions for a longer period of time. The FTIR spec-
tra of these materials and the corresponding second derivatives
Fig. 5. FTIR spectra of alginic acid butyl amide obtained from the reactions of tetra-
butylammonium alginate with the coupling agent 2-chloro-1-methylpyridinium
iodide (DSamide = 0.41; upper spectra) and the coupling agent propylphosphonic
−
1
showed only the typical bands of Na-Alg (1612 cm ) and H-Alg
−
1
−1
(
(
1735 cm ) moieties and no longer those of ␥-lactone structures
anhydride (DSamide = 0.96; lower spectra). Wavenumbers correspond at 1616 cm
−
1
−1
and at 1551 cm to amide groups
1782 cm−1).
to carboxylate ions and at 1665 cm
CO NH ).
(

2
50
T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
Fig. 6. Reaction schemes for the amidation of tetrabutylammonium alginate in polar aprotic organic solvents with 2-chloro-1-methylpyridinium iodide (left) and propy-
lphosphonic anhydride (right) as coupling agent, including the mannuronic acid ␥-lactone formation.
the synthesis of AlgNC4. Furthermore, it was deduced from the ␥-
lactone hydrolysis during ultrafiltration that an excess of amine was
required. Therefore, the amide synthesis was performed accord-
ingly for both coupling agents with a twofold excess of amine per
alginate moiety. This approach resulted in a DSamide value of 0.47
mannuronic acid ␥-lactone did not, in our opinion, promote the
hypothesis of the intermolecular alginate esters. We even believe
that these results rather disprove this hypothesis.
The amidation of TBA-Alg with CMPI as the coupling agent
according to the optimized conditions of the ␥-lactone synthesis
with a stoichiometric ratio of the coupling agent and a twofold
excess of amine per alginate moiety led to a DSamide value of 0.41
on the basis of elemental analysis and to 0.43 on the basis of 1H
NMR analysis (Fig. 4). The partial amidation of Alg was also proven
by the presence of two C6 carbons, carboxylate and amide carbons,
in the 13C NMR spectrum (Fig. 4), and by the signals of the amide
(Table 2) for the butyl amidation of TBA-Alg with T3P as a cou-
pling agent. The FTIR spectra of this partially amidic alginate, taken
before the ultrafiltration step, again displayed the presence of man-
nuronic acid ␥-lactone structures in the material. The optimization
of the reaction resulted in an elevation of the TBA-Alg concentra-
tion to 5% (w/v) and in an amine excess of 10 times the alginate
moiety. The best DSamide was determined to be 0.96 on the basis of
the elemental analysis, and to be 0.98 according to the evaluation
of the 1H NMR spectrum (Fig. 4). The C NMR (Fig. 4) and the FTIR
spectra (Fig. 5) of this highly amidic alginate were well in line. Two
amide signals were found at 168.12 (C6) and 167.39 (C6) ppm in
13
and carboxylate groups in the FTIR spectrum (Fig. 5). The C NMR
signals of the alginate were in line with earlier reported findings
(Grasdalen, Larsen, & Smisrod, 1981) and the butyl group signals
appeared in the expected ranges. Modifications of the synthesis
conditions including the usage of an extreme excess of amine (10
units/TBA-Alg moiety) resulted at best in a similar DSamide in our
test series so far.
13
13
−1
the C NMR spectrum and at 1665 and 1551 cm in the FTIR spec-
trum. The typical carboxylate signal for partially amidic alginates
was neither present in the 13C spectrum nor in the FTIR spectrum.
Furthermore, DSamide values of about one and the presence of the
An interesting fact was that the optimized reaction conditions
for both coupling agents were found to be identical, but they

T. Schleeh et al. / Carbohydrate Polymers 138 (2016) 244–251
251
resulted in very different DSamide values. T3P led to an almost
completely amidated alginate, whereby the amidation with CMPI
remained below a DSamide of 0.5. This efficiency difference between
the two coupling agents was unexpected. Both of them were
selected for their capacity to change the alginate carboxylate moi-
eties into activated esters, and therefore, should have a similar
effect on the following amidation. This, however, did not take into
account the impact of the lactone formation. Original esterifica-
tion with CMPI (Bald, Saigo, & Mukaiyama, 1975) required proton
scavengers for the released hydroxyl proton. This additional pro-
ton from the ␥-lactone formation was not considered so far. The
presence of a free proton was believed to result in the formation of
alginic acid, which was then no longer available for the amidation.
Therefore, adding a proton scavenger before adding CMPI was
expected to reduce or to suppress the alginic acid formation. Per-
forming the reaction accordingly with triethylamine as a proton
scavenger (Fig. 6) led to a DS of up to 0.80 for the material isolated
from the solvent fraction (see Section 2.2.4). The average DS of the
complete batches were found to be 0.68. These high DSamide proved
in our opinion the amidation of some mannuronic acid ␥-lactone
units, because the mannuronic acid moiety content was 61%, which
was lower than the average DSamide. Therefore, the hypothesis of
alginic acid formation as a result of the ␥-lactone formation was
most probably correct. The amidation with CMPI as coupling agent,
however, was still not as complete as the one with T3P as coupling
agent. Further improvements of the amidation with CMPI as cou-
pling agent were assumed to be achievable by varying the proton
scavenger.
The molecular weight analysis confirmed the polymeric nature of
the resulting AlgNC4 samples despite some polymer degradation.
Acknowledgments
The authors thank Gerhard Wenz from the Saarland University
(Germany) for his support with the NMR experiments and Laurent
Delevoye from the Université Lille 1 (France) for his support on the
solid state NMR experiments.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carbpol.2015.11.070.
References
Anal, A. K., & Singh, H. (2007). Recent advances in microencapsulation of probiotics
for industrial applications and targeted delivery. Trends in Food Science &
Technology, 18, 240–251.
Anson, S. I., Novikova, E. V., & Iozep, A. A. (2009). Intramolecular esters of alginic
acid. Russian Journal of Applied Chemistry, 82, 1095–1097.
Bald, E., Saigo, K., & Mukaiyama, T. (1975). A facile synthesis of carboxamides by
using 1-methyl-2-halopyridinium iodides as coupling reagents. Chemistry
Letters, 4, 1163–1166.
Barbucci, R., Consumi, M., & Magnani, A. (2002). Dependence of water uptake and
morphology of hyaluronan- and alginate-based hydrogels on pH and degree of
crosslinking. Macromolecular Chemistry and Physics, 203, 1292–1300.
Barbucci, R., Leone, G., Chiumiento, A., Di Cocco, M. E., D’Orazio, G., Gianferri, R.,
et al. (2006). Low- and high-resolution nuclear magnetic resonance (NMR)
characterisation of hyaluronan-based native and sulfated hydrogels.
Carbohydrate Research, 341, 1848–1858.
As the amidation with T3P as coupling agent resulted, under
optimal conditions, in almost complete amidation of TBA-Alg, it
Della Valle, F., & Aurelio, R. (1992, September 15). Esters of alginic esters. Fidia, S. p.
A. US 761,511[5,147,861].
+
Grasdalen, H., Larsen, B., & Smisrod, O. (1981). 13C-N.M.R. studies of monomeric
was believed that the released phosphonic acid was binding TBA
and H , resulting from the lactone formation (Fig. 6).
+
composition and sequence in alginate. Carbohydrate Research, 89, 179–191.
Hesse, M., Meier, H., & Zeh, B. (2008). Spectroscopic methods in organic chemistry
(2nd ed.). Stuttgart: Georg Thieme Verlag.
Hu, W., Li, J., Hou, H., Yan, H., Feng, Y., Mi, X., et al. (2013). Preparation and
characterization of hydrophobic alginate derivative nanocapsules entrapping
The analysis of the molecular weights (Table 2) revealed the
still polymeric character of the samples. In case of partial amida-
tion, they were determined to decrease from Mw ∼ 77 kDA (Schleeh
et al., 2014) of TBA-Alg to 35.6 kDa where CMPI was the coupling
agent and to 44.1 where T3P was the coupling agent. The SEC MALS
measurement of the almost quantitative amidated alginate showed
a bimodal distribution curve. The evaluation of the first signal gave
a far too high molecular weight Mw. This was interpreted as the
result of micelle formation in the aqueous solution. The evaluation
of the second peak which was not baseline, subtracted from the first
peak of the micelles, resulted in a reasonable Mw of about 55 kDa.
␭
-cyhlothrin. Asian Journal of Chemistry, 25(17), 9904–9908.
Leone, G., Torricelli, P., Chiumiento, A., Facchini, A., & Barbucci, R. (2008). Amidic
alginate hydrogel for nucleus pulposus replacement. Journal of Biomedical
Materials Research Part A, 84A, 391–401.
Nagendra, G., Madhu, C., Vishwanatha, T. M., & Sureshbabu, V. V. (2012). An
expedient route for the reduction of carboxylic acids to alcohols employing
1-propanephosphonic acid cyclic anhydride as acid activator. Tetrahedron
Letters, 53, 5059–5063.
Pawar, S. N., & Edgar, K. J. (2012). Alginate derivatization: A review of chemistry,
properties and applications. Biomaterials, 33, 3279–3305.
Pawar, S. N., & Edgar, K. J. (2013). Alginate esters via chemoselective carboxyl
group modification. Carbohydrate Polymers, 98, 1288–1296.
Pelletier, S., Hubert, P., Lapicque, F., Payan, E., & Dellacherie, E. (2000). Amphiphilic
derivatives of sodium alginate and hyaluronate: Synthesis and
physico-chemical properties of aqueous dilute solutions. Carbohydrate
Polymers, 43, 343–349.
Saha, D., & Bhattacharya, S. (2010). Hydrocolloids as thickening and gelling agents
in food: A critical review. Journal of Food Science and Technology-Mysore, 47,
587–597.
4
. Summary
The confirmation of the hypothesis of lactone formation and
the understanding of the resulting impact on the reaction were
the key factors for the successful synthesis of AlgNC4 with a high
DSamide from TBA-Alg activated with CMPI (DSamide ≤ 0.8) or T3P
Saigo, K., Usui, M., Kikuchi, K., Shimada, E., & Mukaiyama, T. (1977). New method
for preparation of carboxylic esters. Bulletin of the Chemical Society of Japan, 50,
(
DSamide ≤ 1) as a coupling agent. Only the mannuronic acid moi-
1863–1866.
eties were confirmed to be able to form ␥-lactone structures,
which required a change from the preferred 4C1 to C4 con-
formation for the mannuronic acid ring. It was further deduced
that the mannuronic acid ␥-lactone was rather stable, because it
withstood short-term exposure to mild hydrolytic conditions. The
mannuronic acid ␥-lactone also expressed its stability in the nec-
essary high excess of a nucleophile required for the ring opening
reaction to result in either the mannuronic acid or a derivative such
as an amide. The proton released from the lactone formation had to
be bound by a proton scavenger for the following amidation. This
was done by the resulting phosphonic acid derivative in the case
of T3P and the specifically added triethylamine in the case of CMPI.
Schleeh, T., Madau, M., & Roessner, D. (2014). Synthesis enhancements for
generating highly soluble tetrabutylammonium alginates in organic solvents.
Carbohydrate Polymers, 114, 493–499.
1
Schultz, G. S., Sibbald, R. G., Falanga, V., Ayello, E. A., Dowsett, C., Harding, K., et al.
(
2003). Wound bed preparation: A systematic approach to wound
management. Wound Repair and Regeneration, 11, 1–28.
Tonnesen, H. H., & Karlsen, J. (2002). Alginate in drug delivery systems. Drug
Development and Industrial Pharmacy, 28, 621–630.
Vallée, F., Müller, C., Durand, A., Schimchowitsch, S., Dellacherie, E., Kelche, C., et al.
(2009). Synthesis and rheological properties of hydrogels based on
amphiphilic alginate-amide derivatives. Carbohydrate Research, 344, 223–228.
Wissmann, H., & Kleiner, H. J. (1980). New peptide synthesis. Angewandte Chemie
International Edition in English, 19, 133–134.
Yang, J. S., Xie, Y. J., & He, W. (2011). Research progress on chemical modification of
alginate: A review. Carbohydrate Polymers, 84, 33–39.