Alginic acid aerogel: A heterogeneous Br?nsted acid promoter for the direct Mannich reaction

Article abstract of DOI:10.1039/c5nj00349k

We demonstrate that highly dispersed alginic acid aerogel beads can function as heterogeneous promoters for Br?nsted acid catalyzed reactions. The activity displayed by alginic acid, benchmarked in the three-component direct Mannich reaction, depended on its morphology, with the aerogel formulation providing distinct advantages over other forms.

Full text of DOI:10.1039/c5nj00349k

NJC
View Article Online
View Journal
LETTER
Alginic acid aerogel: a heterogeneous Brønsted
acid promoter for the direct Mannich reaction†
Cite this:DOI: 10.1039/c5nj00349k
Asja Pettignano,ab Luca Bernardi,*b Mariafrancesca Fochi,b Lorenzo Geraci,b
Mike Robitzer,a Nathalie Tanchouxa and Françoise Quignard*a
Received (in Montpellier, France)
9th February 2015,
Accepted 30th March 2015
DOI: 10.1039/c5nj00349k
www.rsc.org/njc
We demonstrate that highly dispersed alginic acid aerogel beads can
function as heterogeneous promoters for Brønsted acid catalyzed
reactions. The activity displayed by alginic acid, benchmarked in
the three-component direct Mannich reaction, depended on its
morphology, with the aerogel formulation providing distinct advantages
over other forms.
Alginates are polysaccharide biopolymers, available at extremely
cheap prices and in virtually unlimited amounts by extraction
mainly from brown algae. These biodegradable polymers are
commonly encountered in daily life, being used both in the food
industry and in textile printing as thickening agents, whereas their
capability to serve as enzyme supports is exploited in brewing and
laundry powders.1 In the current vision of a necessary switch from
fossil to non-depleting feed-stocks,2 additional utilization of these
biopolymers is arguably highly attractive and worth pursuing,3
even for less mass-intensive applications.
Alginates are constituted by sequences of (1 - 4) linked
b-D-mannuronate (M) and a-L-guluronate (G) residues (Fig. 1).
In contrast to cellulose and starch but in common with other more
expensive polysaccharides of marine origin (chitin/chitosan and
carrageenan), alginates feature a functional group (a carboxylate
moiety) in each monomeric unit of their backbone, and are thus
polymers with an extremely high functional group density
(5.6 mmol gÀ1). In analogy with their natural structural function,
alginates can be easily obtained as hydrogels.4 However, recent
findings have demonstrated their tremendous manipulability,
Fig. 1 Metal alginate and alginic acid structure.
well beyond the hydrogel form.5 Appropriate treatments allow
control of the shape, morphology, mechanical properties, porosity
and surface area of metal alginates and alginic acids6 (obtained by
acidification), almost at will. In particular, the aerogel formulation
renders a manageable macro-porous material with very high surface
area (350–500 m2 gÀ1), wherein all functional groups are exposed
and available.7 These features seem to be well suited for catalysis,
and have already been exploited for the preparation of alginate-
supported metallic catalysts.8 Moreover, recent studies have sug-
gested the possibility of easily obtaining highly shaped mesoporous
materials from fresh macroalgae, allowing their exploitation for
chromatography, catalyst supports and drug delivery systems.9
Possibly prompted by the increasing attention to unconventional
applications of renewable materials, the study of the capability of
biopolymers in promoting simple reactions is a topic of current
interest. Several unmodified biopolymers, such as chitosan,10
calcium alginate,11 polydopamine,12 silk fibroin,13 collagen and
gelatin,14 have been found to be active as Lewis or Brønsted base
catalysts. Besides its potential in the frame of heterogeneous
catalysis, interest in this peculiar research field is related to the
intriguing hypothesis that these structural biopolymers, widely
found in living systems and employed as passive supports/
matrices in different settings (food and pharmaceutical industries,
brewing, etc.), might not be as inert as they are thought to be.
Whereas the above examples well demonstrate Lewis or Brønsted
basic activation of substrates, the engagement of biopolymers in
Brønsted acid catalysis has not been intensively studied. A single
a
´
´
´
Institut Charles Gerhardt, Materiaux Avances pour la Catalyse et la Sante,
UMR 5253, CNRS-UM2-ENSCM-UM1, 8 Rue de l’Ecole Normale, 34296,
Montpellier, Cedex 5, France. E-mail: quignard@enscm.fr
b Department of Industrial Chemistry ‘‘Toso Montanari’’, School of Science,
Alma Mater Studiorum – University of Bologna, V. Risorgimento 4,
40136 Bologna, Italy. E-mail: luca.bernardi2@unibo.it
† Electronic supplementary information (ESI) available: Preparation of the gel
materials, additional optimization results, kinetic studies, experimental details
and product characterization, and copies of the 1H and 13C NMR spectra. See DOI:
10.1039/c5nj00349k
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Letter
NJC
Table 1 Solvent screening in the 2C Mannich reactiona
Entry
Solvent
Catalyst
4a-Yieldb (%)
4a : 6c
61 : 39
52 : 48
62 : 38
75 : 25
65 : 35
495 : 5
495 : 5
495 : 5
Scheme 1 Alginic acid aerogel promoted 3C Mannich reaction.
1
2
3
4
5
6
7
8
Toluene
CH2Cl2
EtOAc
THF
AG-1
AG-1
AG-1
AG-1
AG-1
AG-1
—
17
13
15
9
35
91d
24
27
example, appeared while this work was in progress, showed
moderate activity of alginic acid powder in a Hantzsch reaction,15
without however reporting data on the dependency of the activity
on the polymer morphology.
CH3CN
CH3CN–H2O 8 : 2
CH3CN–H2O 8 : 2
CH3CN–H2O 8 : 2
AcOH
Herein, we present the evidence that alginic acid aerogel can
act as a heterogeneous Brønsted acid promoter. In more detail,
we disclose that the carboxylic moieties of the biopolymer can
function as a mild catalyst for a three component (3C) Mannich
addition reaction (Scheme 1), demonstrating the essential
benefits of using the highly dispersed aerogel form of this
material. Furthermore, upon acidification, even a seaweed
commercialized for food purposes can be rendered catalytically
active. For our studies, we chose the Mannich reaction as a
benchmark not only for its known susceptibility to acidic (both
Brønsted and Lewis) catalysis16 but also for the typical challenges
associated with this reaction, such as competing aldol addition,
reversibility, formation of undesired a,b-unsaturated carbonyls, and
the requirement of a water compatible reaction promoter.17 These
challenges render this 3C process a reliable test for demonstrating
the efficiency of a Brønsted acid. Besides, the Mannich reaction is a
fundamental transformation in organic chemistry, its 3C direct
version provides a very straightforward access to structurally diverse
b-amino carbonyl compounds (Mannich bases).18
At the outset, the capability of alginic acid to function as a
Brønsted acid was assessed in a 2C Mannich reaction (Table 1), in
order to simplify the system. The reaction between preformed imine
5 and cyclohexanone 3a was investigated in different solvents, in the
presence of guluronic-rich alginic acid (G/M 63 : 37)19 in its aerogel
bead formulation (aerogel AG-1, 20 mol% calculated on carboxylic
functions). This preliminary solvent screening (Table 1, entries 1–5)
suggested acetonitrile as a promising medium for the reaction,
wherein a moderate conversion was observed (Table 1, entry 5).
In all the solvents tested, the desired Mannich adduct 4a was
accompanied by considerable amounts of an undesired by-product
6, derived from a retro-aza-Michael reaction also promoted by the
heterogeneous acid.
a
Conditions: imine 5 (0.15 mmol), ketone 3a (0.30 mmol), cat. AG-1
(20 mol%), RT, 18–24 h. Yields were determined by 1H NMR using
b
bibenzyl as internal standard. Determined by 1H NMR. Isolated
c
d
yield after chromatography on the silica gel.
catalyst, since higher amounts of water promoted a non-negligible
background uncatalyzed process, which would prevent the correct
evaluation of the catalyst activity. This 80 : 20 ratio was thus
employed throughout this study.20 Monitoring by 1H NMR spectro-
scopy the evolution of the yield with time (see ESI†), we were able to
assess that the reaction follows approximately pseudo-first order
kinetics and to choose 18–20 h as the best reaction time, as the
reaction considerably slows down already after 9 h and it appears
almost complete after 18 h, allowing the isolation of the pure
Mannich product 4a in 91% yield (Table 1, entry 6). Under these
conditions (18 h reaction time), the low yield obtained performing
the reaction in the absence of AG-1 (Table 1, entry 7) clearly
demonstrated the activity of alginic acid as a heterogeneous
Brønsted acid promoter in the model reaction. Its effectiveness
was further confirmed by comparison with a homogeneous
carboxylic acid, such as AcOH, which showed a negligible yield
(Table 1, entry 8), comparable with the background, uncatalysed
reaction.
After having optimized the 2C reaction, implementation of
its 3C version proved to be straightforward, allowing a short
investigation of its scope by combining different aldehydes 1,
It was then found that a consistent improvement in reaction
efficiency could be achieved by using CH3CN–H2O mixtures as
solvents instead of pure CH3CN, which allowed increasing the
productivity and completely suppressing the by-product (6)
formation. To assess the influence of the amount of water, a
series of experiments with different CH3CN–H2O mixtures was
performed (Fig. 2). Short reaction times (5 h) were used, to
better appreciate variances in the activity displayed by AG-1.
These experiments showed a steady increase in the yield of the
Mannich adduct 4a with the amount of water used. In particular,
an 80: 20 ratio between acetonitrile and water appeared to be
optimal for studying and demonstrating the efficiency of AG-1 as
Fig. 2 Influence of the amount of water on the 2C Mannich reaction
between 5 and 3a (reaction time: 5 h). Yields were determined by 1H NMR
spectroscopy using bibenzyl as internal standard.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Letter
anilines 2 and cyclic ketone donors 3 (Table 2). Besides the the macroscopic nature of the beads seemed to be unchanged,
Mannich adduct 4a derived from benzaldehyde 1a and aniline and the carboxylic functions were clearly visible by IR spectro-
2a, a few other adducts 4b–h were obtained in moderate to scopy analysis, we speculate that pore occlusion by the reac-
excellent yields from different aldehydes 1b–e and electron- tants/products was the main reason for this partial loss of
poor anilines 2b–d (Table 2, entries 1–8).21 Electron rich activity.
aldehydes 1f–g and aniline 2e could also be applied successfully
Different formulations of alginic acid (xerogel XG, solvogel
to the reaction, even if the instability of adducts 4i–k upon SG, and, for the sake of comparison, commercial alginic
chromatographic purification on silica gel prevented the deter- powder C) were then tested. Using pure acetonitrile as a
mination of the yield of isolated products (Table 2, entries solvent, the obtained results (Fig. 3) clearly indicated the
9–11). Other cyclic ketones 3b–d were tested in the AG-1 requirement of a polysaccharide material with high porosity
catalyzed reaction. While pyran-4-one 3b and cyclopentanone and surface area for the promotion of the reaction, with
3c performed well in the reaction (Table 2, entries 12 and 13), solvogel SG giving results comparable to AG-1, while xerogel
the 7-membered ring ketone 3d did not afford the expected XG and the commercial powder C hardly showing any reactivity.
product 4n (Table 2, entry 14), thus highlighting the limitation Different results were obtained using the usual CH3CN–H2O
of the present catalytic protocol to more easily enolisable 8 : 2 mixture, where all the formulations showed an increased
ketone substrates.
activity, with all gels leading to similar yields. To better under-
Given the heterogeneous nature of the catalyst, its recovery stand this different behaviour the reaction was repeated using a
and reuse were preliminarily investigated. At the end of the protic solvent (EtOH) instead of water, so as to verify if it has to
reaction between benzaldehyde 1a, aniline 2a and cyclohexa- be ascribed to a swelling or a proton transfer effect of water.
none 3a, performed using the optimized conditions, the The results clearly showed that the AG-1 and SG formulations,
catalyst was separated from the solution and carefully washed in which all the active sites are accessible, maintain almost the
with the CH3CN–H2O 8 : 2 mixture, before application in the same activity in the presence of EtOH, while XG and C are
next cycle. After the first recycle, the yield dropped to 77% of essentially unable to promote the reaction. Thus, the activity
that obtained in the first run (Y4a: 65%). In the next two cycles, featured by XG and C in the presence of water is simply due to
the yield stabilized at a moderate value (Y4a: 50%), falling the swelling of the polysaccharide structures leading to a higher
further upon the fourth recycle (Y4a: 40%). Attempts to restore amount of exposed catalytic sites.
the catalytic activity by acidic washings were unfruitful. Since
Then, the effect of natural variability in terms of the alginic
acid composition on the reaction was verified by comparing the
performance of the guluronic rich AG-1 (G/M 63 : 37) with a
mannuronic-rich alginic acid aerogel,22 AG-2 (G/M 33 : 67).19
The G/M ratio influences the mechanical properties of the gels,
their strength increases with the guluronic content.23
As expected, the use of the weaker AG-2 led to a poorer yield
(Y-4a: 66% vs. 91%), presumably due to the partial disruption
of the gel structure of AG-2 during the reaction. Finally, our
model reaction was performed using a commercial seaweed (S),
Laminaria Digitata (G/M 41 : 59),24 conveniently pre-treated to
obtain an acid seaweed aerogel (Fig. 4 and ESI†). The result of
Table 2 Scope and limitations of the 3C Mannich reaction between
different aldehydes 1, anilines 2 and ketones 3 catalyzed by AG-1a
Entry 1: Ar1
2: Ar2
3
4-Yieldb (%) anti/sync
1
2
3
4
1a: C6H5
2a: C6H5
2a: C6H5
2a: C6H5
2a: C6H5
2a: C6H5
2b: 4-ClC6H4
2c: 4-BrC6H4
2d: 3,4-Cl2C6H4 3a 4h-90
1a: C6H5
3a 4a-90
3a 4b-68
3a 4c-76
3a 4d-60
3a 4e-61
3a 4f-57
3a 4g-95
61 : 39
63 : 37
64 : 36
61 : 39
64 : 36
53 : 47
50 : 50
65 : 35
61 : 39
64 : 36
64 : 36
65 : 35
61 : 39
—
1b: 4-ClC6H4
1c: 4-NO2C6H4
1d: 2-naphth
1e: 4-FC6H4
1b: 4-ClC6H4
1b: 4-ClC6H4
1b: 4-ClC6H4
1f: 4-MeOC6H4
5
6d
7d
8d
9
3a 4i-76e
3a 4j-60e
10
11d
12
13
14
1g: 3,4-(MeO)2C6H3 2a: C6H5
1b: 4-ClC6H4
1b: 4-ClC6H4
1b: 4-ClC6H4
1b: 4-ClC6H4
2e: 4-MeOC6H4 3a 4k-72e
2a: C6H4
2b: 4-ClC6H4
2b: 4-ClC6H4
3b 4l: 92
3c 4m: 50
3d 4n: o10
a
Conditions: aldehyde 1 (0.15 mmol), aniline 2 (0.15 mmol), ketone 3
(0.30 mmol), AG-1 (20 mol%), CH3CN–H2O 8 : 2 (480 mL), RT,
18–20 h. b Isolated yield after chromatography on the silica gel. c Determined
by 1H NMR on the crude mixture. d 5 equiv. of ketone (0.75 mmol) were used.
e Product unstable upon chromatography. Yield determined by 1H NMR
spectroscopy on the crude with an internal standard (bibenzyl).
Fig. 3 Effect of the solvent on the activity displayed by the different
alginic acid formulations in the reaction (RT, 18–20 h) between preformed
imine 5 and cyclohexanone 2a (2 equiv.). Yields determined by 1H NMR
spectroscopy using bibenzyl as internal standard.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.

View Article Online
Letter
NJC
8 (a) M. Chtchigrovsky, Y. Lin, K. Ouchaou, M. Chaumontet,
M. Robitzer, F. Quignard and F. Taran, Chem. Mater., 2012,
24, 1505; (b) F. Shi, Y. Chen, L. Sun, L. Zhang and J. Hu,
Catal. Commun., 2012, 25, 102; (c) Y. Dong, W. Dong, Y. Cao,
Z. Han and Z. Ding, Catal. Today, 2011, 175, 346.
9 J. R. Dodson, V. L. Budarin, A. J. Hunt, P. S. Shuttleworth
and J. H. Clark, J. Mater. Chem. A, 2013, 1, 5203.
10 (a) R. Valentin, K. Molvigner, F. Quignard and D. Brunel,
New J. Chem., 2003, 27, 1690; (b) A. Ricci, L. Bernardi,
C. Gioia, S. Vierucci, M. Robitzer and F. Quignard, Chem.
Commun., 2010, 46, 6288; (c) C. Gioia, A. Ricci, L. Bernardi,
K. Bourahla, N. Tanchoux, M. Robitzer and F. Quignard,
Eur. J. Org. Chem., 2013, 588; (d) D. J. Macquarrie and
J. J. E. Hardy, Ind. Eng. Chem. Res., 2005, 44, 8499;
Fig. 4 Preparation of the acidified seaweed aerogel (S).
´
(e) D. Ku¨hbeck, G. Saidulu, K. Rajender Reddy and D. Dıaz
´
( f ) A. El Kadib, ChemSusChem, 2015, 8, 217.
the benchmark 2C reaction between 5 and 3a (Y-4a: 85%) is
comparable to that obtained using AG-1; thus, this result
suggests that the carboxylic acid functionalities can be made
available directly from seaweed without any extraction or
purification of the polysaccharide.
Dıaz, Green Chem., 2012, 14, 378; for a recent review;
11 (a) C. Verrier, S. Oudeyer, I. Dez and V. Levacher, Tetrahedron
¨
Lett., 2012, 53, 1958; (b) D. Ku¨hbeck, J. Mayr, M. Haring,
M. Hofmann, F. Quignard and D. Dıaz Dıaz, New J. Chem.,
´
´
In conclusion, we have demonstrated that alginic acid can
function as a mild Brønsted acid catalyst for a three component
Mannich reaction. Thus, the present report offers prospects in
the frame of heterogeneous catalysis based on renewable
resources. Besides, the capability of unmodified alginic acid
in promoting a chemical transformation provides additional
support to the provoking suggestion that biopolymers featuring
structural functions, widely found in living systems and used in
the food and other industries as additives, might not be as
innocent as they are thought to be.12–14
2015, 39, 2306.
´
´
12 R. Mrowczynski, A. Bungea and J. Liebscher, Chem. – Eur. J.,
2014, 20, 8647.
´
´
13 D. Ku¨hbeck, M. Ghosh, S. S. Gupta and D. Dıaz Dıaz, ACS
Sustainable Chem. Eng., 2014, 2, 1510.
14 (a) D. Ku¨hbeck, B. B. Dhar, E.-M. Schon, C. Cativiela,
¨
´
´
´
V. Gotor-Fernandez and D. Dıaz Dıaz, Beilstein J. Org. Chem.,
¨
2013, 9, 1111; (b) D. Ku¨hbeck, J. Bachl, E.-M. Schon,
´
´
´
V. Gotor-Fernandez and D. Dıaz Dıaz, Helv. Chim. Acta,
2014, 97, 574.
15 M. G. Dekamin, S. Ilkhanizadeh, Z. Latifidoost, H. Daemi,
Z. Karimi and M. Barikani, RSC Adv., 2014, 4, 56658.
16 S. Kobayashi, Y. Mori, J. S. Fossey and M. M. Salter, Chem.
Rev., 2011, 111, 2626.
17 L. Bernardi and A. Ricci, in Science of Synthesis: Multi-
component Reactions 1, ed. T. J. J. Mu¨ller, Thieme Verlag KG,
Stuttgart, 2014, pp. 123–164.
18 E. F. Kleinman, in Comprehensive Organic Synthesis, ed.
B. Trost and I. Fleming, Pergamon, New York, 1991, vol. 2,
pp. 893–951.
19 C. Jouannin, C. Vincent, I. Dez, A.-C. Gaumont, T. Vincent
and E. Guibal, Nanomaterials, 2012, 2, 31.
Acknowledgements
This work was co-funded through a SINCHEM Grant. SINCHEM
is a Joint Doctorate programme selected under the Erasmus
Mundus Action 1 Programme (FPA 2013-0037). We are grateful to
Prof. Alfredo Ricci and Dr Claudio Gioia for useful discussions at
the early stages of the project.
Notes and references
1 D. J. McHugh, A Guide to the Seaweed Industry, FAO, Rome,
20 The beneficial effect of water in the reaction can be inter-
preted considering its known positive influence on direct
Mannich reactions [G.-p. Lu and C. Cai, Catal. Commun.,
2010, 11, 745, and references therein] by increasing the
polarity of the medium, and/or influencing equilibria, or
assisting proton-transfer events. However, it cannot be
entirely excluded that water might have an impact, by e.g.
hydrophobic effects, on the mass transfer processes occur-
ring between the bulk solution and the catalyst surface.
Following referees suggestions, the reaction was also tested
in pure water, in the absence and in the presence of phase-
transfer co-catalysts (SDS and aliquat-336). Whereas the
reactions without PTC and with aliquat-336 did not afford
the product, presumably due to the low solubility of the
2003, ch. 4 and 5, pp. 27–50.
¨
2 R. Hofer and J. Bigorra, Green Chem., 2007, 9, 203.
3 V. L. Budarin, Y. Zhao, M. J. Gronnow, P. S. Shuttleworth,
S. W. Breeden, D. J. Macquarrie and J. H. Clark, Green
Chem., 2011, 13, 2330.
4 G. Skjåk-Braek, H. Grasdalen and O. Smidsrød, Carbohydr.
Polym., 1989, 10, 31.
5 F. Quignard, R. Valentin and F. Di Renzo, New J. Chem.,
2008, 32, 1300.
6 L. Li, Y. Fang, R. Vreeker and I. Appelqvist, Biomacromolecules,
2007, 8, 464.
7 R. Valentin, R. Horga, B. Bonelli, E. Garrone, F. Di Renzo
and F. Quignard, Biomacromolecules, 2006, 7, 877, and
references cited therein.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015

View Article Online
NJC
Letter
substrates, in the presence of SDS there was no evidence of
AG-1 catalytic activity. See ESI†.
A. Haug, Composition and Properties of Alginates, Norwegian
Inst. Seaweed Res., Trondheim, Norway, 1964, Rept. No. 30.
´
21 The poor diastereomeric ratios, favouring the anti-isomers, 23 G. Skjak-Bræk, O. Smidsrød and B. Larsen, Int. J. Biol.
possibly reflect the thermodynamic ratios of the two
diastereoisomers.
22 The two monosaccharides display a small acidity difference
(D-mannuronic acid pKa: 3.38, L-guluronic acid pKa: 3.65):
Macromol., 1986, 8, 330.
24 S. T. Moe, K. I. Dragel, G. Skjåk-Bræk and O. Smidsrød, in Food
Polysaccharides and Their Applications, ed. A. M. Stephen, Marcel
Dekker Inc., New York, 1995, pp. 245–286.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
New J. Chem.