A bio-generated Fe(iii)-binding exopolysaccharide used as new catalyst for phenol hydroxylation

Article abstract of DOI:10.1039/c004967k

The hydroxylation of phenol with aqueous H₂O₂ to afford catechol and hydroquinone was studied in a biphasic reaction medium, as well as in pure water, in the presence of a bio-generated iron(iii) catalyst. This catalyst was purified

Full text of DOI:10.1039/c004967k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
A bio-generated Fe(III)-binding exopolysaccharide used as new catalyst for
phenol hydroxylation
a
a
a
b
b
Franco Baldi,* Davide Marchetto, Davide Zanchettin, Elisabetta Sartorato, Stefano Paganelli and
b,c
Oreste Piccolo*
Received 8th April 2010, Accepted 14th June 2010
First published as an Advance Article on the web 13th July 2010
DOI: 10.1039/c004967k
The hydroxylation of phenol with aqueous H
2
2
O to afford catechol and hydroquinone was studied
in a biphasic reaction medium, as well as in pure water, in the presence of a bio-generated iron(III)
catalyst. This catalyst was purified from a culture of Klebsiella oxytoca BAS-10 growing under
anaerobic conditions, with Fe(III)-citrate as the energy and carbon source. The overproduction of
3
+
an exopolymer (EPS) encrusted bacterial cells. The EPS, binding Fe , (Fe-EPS), was extracted
and studied before and after a reaction with phenol. Some of the reaction’s parameters, such as
temperature, pH, and molar ratio between reagents and catalyst, were investigated to identify the
best compromise between conversion and selectivity. The results could be useful either from a
synthetic point of view or to support the biodegradation of aromatic substrates.
Introduction
grows in 50 mM ferric citrate and produces a thick Fe(III)
hydrogel, CO
Fe(II) ions during the fermentation.
2
and a solution containing acetic acid and soluble
The transition metal-catalyzed oxidation of organic substrates
remains an important topic in synthetic, industrial and bi-
ological chemistry. One fundamental target is the selective
hydroxylation of aromatic compounds to obtain useful multi-
ton commodities. Furthermore, hydroxylation is often the first
step to permit the degradation of toxic aromatic compounds
when remediation treatments are applied.
There are numerous papers related to the iron-catalyzed hy-
droxylation of aromatic substrates, such as benzene and phenol,
using iron compounds that are free or bound to inorganic
15,17
Recently, a structural
analysis was undertaken and the heptameric repeating unit of
this Fe(III)-binding exopolysaccharide (Fe-EPS) characterized
18
by spectroscopic methods. It was found that the stability
3+
formation constant of Fe with one repeating unit (heptamer)
of EPS was about one or two orders of magnitude larger than
12
13
11
that of the Fe(III)-citrate complex (K ª 10 –10 vs. K ª 10 ) but
24 17
much less than that of the Fe-EDTA complex (K ª 10 ). In
our opinion, the use of this bio-generated iron complex could be
worthy of investigation as a catalyst; therefore, research to verify
its catalytic ability in a model oxidation reaction, such as the
1
–14
supports, or are present in homogeneous complexes.
The comparison of heme and non-heme iron enzymes that are
able to catalyze a number of fascinating biological oxidations is
a source of inspiration for researchers in their efforts to discover
new and efficient catalytic systems. These proposed synthetic
hydroxylation of phenol with H
2
2
O (Scheme 1), was carried out.
2,11–14
iron catalysts often contain nitrogen ligands.
If we exclude
12
natural ferritin, these nitrogen ligands have to be prepared,
and may be expensive and/or not suitable for treating waste
water in an environmental friendly way. In order to find new
ligands for similar catalytic reactions, interest could turn to
bacteria as a source of useful molecules. Several years ago,
a strain of Klebsiella oxytoca BAS-10 was isolated from the
Scheme 1 The oxidation of phenol with H
oxytoca strain BAS-10 as a catalyst
2
O
2
using Fe-EPS from K.
15
acid mine drainage of pyrite mines. This strain produces an
exopolysaccharide (EPS) during citrate fermentation, and this
tricarboxylic acid is transported into the cell through a sodium-
dependent carrier coupled to proton efflux by producing acetate
Experimental
Cultivation of K. oxytoca BAS-10
16
at the end of fermentation. K. oxytoca strain BAS-10 also
◦
The strain was retrieved from cryovials kept at -80 C in 25%
glycerol in nutrient broth (Difco). An aliquot of 1 mL of
overnight culture was transferred under anaerobic conditions
into two media containing, per L: 2.5 g NaHCO , 1.5 g NH Cl,
a
Department of Environmental Sciences, Ca¢ Foscari University, Calle
larga Santa Marta 2137, 30123, Venice, Italy. E-mail: baldi@unive.it;
Tel: +39 041-2348901
3
4
0
.6 g NaH
2
PO , 0.1 g KCl, and iron or sodium citrate species.
4
b
Department of Chemistry, Ca¢ Foscari University, Calle larga Santa
-
1
The first medium, containing 50 mM Fe(III)-citrate (13.5 g L ),
hereafter referred to as FeC medium, was buffered at pH 7.6 with
NaOH. The second medium, containing 50 mM (14.7 g L )
Marta 2137, 30123, Venice, Italy
c
SCSOP, Via Born o` 5, 23896, Sirtori, Italy. E-mail: orestepiccolo@
-
1
tin.it; Tel: +39 335-6828222
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1405–1409 | 1405

View Article Online
Na-citrate, hereafter referred to as NaC medium, was also
buffered at pH 7.6 with NaOH.
A TEM analysis was also undertaken directly without a
staining preparation, mounting sample on copper grids to
determine the Fe(III)-EPS structure after the precipitates formed
in the reaction vessel of the phenol transformation.
An aliquot of 10 mL of ferric hydrogel was deposited and
smeared on an aluminium stub in order to analyze the iron
encrusted cells of the BAS-10 culture using an environmental
scanning electron microscope (ESEM). Specimens were visual-
ized by a scanning electron microscope (FEI Quanta mod. 200,
Eindhoven, NL) (Fig. 1B).
The cultivation of strain BAS-10 was performed separately
either in 1 L of FeC medium or in 1 L of NaC medium, incubated
◦
at 30 C under anaerobic conditions. After 10 d, the suspension
was first centrifuged to eliminate bacterial cells, and then the
supernatant was treated with 800 mL of cooled ethyl alcohol
(
95%) to precipitate the polysaccharides. The purification was
repeated twice. Both colloidal materials from the FeC and NaC
media were dried out under vacuum to obtain either Fe-EPS or
Na-EPS as solid materials.
General procedure of the oxidation reaction
In a 250 mL jacketed glass reactor, 16.9 mg of Fe-EPS
Microscope observations
+3 17
(
equivalent to about 6 mg of Fe ), 0 or 0.4 mmol of RCOOH
[
trifluoroacetic acid (TFA) or acetic acid (AcOH)], 0 or 12 mL
To observe the iron in EPS and cell envelope conditions,
of acetonitrile (AN) and 0 or 10 mL of water were stirred 2 h
at room temperature. Next, 1.69 g (18 mmol) of phenol was
added, and the mixture heated and maintained at 37 C. Finally,
K. oxytoca BAS-10 cells were grown in FeC medium after
◦
1
0 d of incubation at 30 C. During cell growth, after hydrogel
◦
precipitation, several specimens were prepared for transmission
electron microscopy (TEM) according to the usual procedure,
cells being harvested by centrifugation at 8100 g. The bacterial
an aqueous solution of 35% (w/v) H
2
O
2
was slowly added
0.1 mmol min until reaching 0.3 mL) to the mixture; a
second identical quantity of H was added after 2 h (in some
-
1
(
◦
2
O
2
pellet was fixed for 1 h at 4 C with 2.5% glutaraldehyde,
experiments, further quantities of this reagent were added). It
was decided to stop and work-up each test after 24 h to be
sure that all the H
after with 0.1 M lysine in a 0.066 M cacodylate buffer at
pH 7.2 for 30 min at room temperature. The cells were washed
five times with the same buffer and post-fixed for 1 h at room
temperature in 1% osmium tetroxide, rinsed with distilled water
and embedded in Spurr resin. Ultrathin sections were prepared
using a LKB II Nova Ultramicrotome device with a diamond
knife. Sections were stained with 3.0% uranyl acetate solution
for 15 min, washed once with distilled water and incubated in
lead citrate for 10 min. TEM observations were performed using
a JEOL JEM 100b (Tokyo, Japan) instrument operating under
standard conditions (Fig. 1A).
2
2
O had been consumed. The mixture was
acidified (to about pH 1), filtered on a sintered glass, extracted
with 2 ¥ 10 mL of 2-methyl-tetrahydrofuran (MTF) and the
organic phase analyzed by gas chromatography [GC system:
Agilent 6850; HP-1 column (30 m ¥ 0.32 mm ¥ 0.25 mm); T =
◦
◦
◦
-1
8
0 C (5 min) then up to 250 C at 10 min ]. The results
of the different runs under each set of reaction conditions are
reported in Table 1. The conversion of phenol and the amount of
the dihydroxylated products (catechol and hydroquinone) was
determined by calibration vs. cyclooctane, taken as an internal
standard for the GC analyses.
SEM micrographs
SEM micrographs were taken to investigate the structural prop-
erties of the precipitate obtained from the treatment with phenol
and H
2
2
O in pure water, and the images were analyzed to verify,
in particular, the emergence of recurrent patterns resembling
ordered particles by the Fourier decomposition of selected
lines of the raster grayscale. Practically, each line provided a
sequence of pixels, each with a value ranging from 0 (black:
electron dense iron) to 255 (white: space between particles),
which were then processed analogously to a time series. It is
notable that the three sections cut the image with the same angle
and preserved the image’s original dimensions (i.e., an isomeric
rotation was performed). Spectral densities were evaluated by
smoothing the periodograms with a w = 5 Hamming window.
As for temporal series, data autocorrelation was used to evaluate
the background noise spectrum, i.e., the energy that is likely
to be produced at a given scale by a random noise process.
Accordingly, only spectral peaks that were significantly above
the background noise spectrum were assumed to be a true feature
of the investigated process. In particular, sample s1 was used
to deduce the properties of the background noise spectrum by
treating the grayscale sequence as a first order autoregressive
Fig. 1 A: A TEM micrograph of a BAS-10 strain producing an
exopolysaccharide rich in iron after fermentation in the presence of
Fe(III)-citrate under anaerobic conditions (25 000¥). B: An ESEM
micrograph of a BAS-10 strain that is entirely encrusted with iron
deposits.
19
process, i.e., according to the formula
1
406 | Green Chem., 2010, 12, 1405–1409
This journal is © The Royal Society of Chemistry 2010

View Article Online
2 2
Table 1 Experimental conditions and results of phenol oxidation by H O in the presence of a catalytic amount of Fe-ESP
a
a
Run
Solvent(s) (/mL)
AN (12)
RCOOH (/mmol)
H
2
O
2
/mL [/mmol]
Conversion (%)
Catechol (II) (%)
Hydroquinone (III) (%)
1
2
3
4
5
TFA (0.4)
TFA (0.4)
TFA (0.4)
AcOH (0.4)
AcOH (0.4)
AcOH (0.8)
—
0.6 [6.8]
1.3 [14.7]
1.3 [14.7]
1 [11.3]
0.6 [6.8]
1.2 [13.6]
0.6 [6.8]
1.2 [13.6]
0.6 [6.8]
4
3
7
7
17
15
9.5
14
8.0
0
1
3
3
8
7
4.9
6
4.8
0
H
2
O (10)
10
10
25
22
18.3
20
18.3
0
AN/H
AN/H
AN/H
AN/H
2
2
2
2
O (12/10)
O (12/10)
O (12/10)
O (12/10)
b
6
7
8
H
H
2
O (10)
O (10)
b
2
—
—
9
MTF/H
2
O (12/10)
a
b
AN = acetonitrile, TFA = trifluoroacetic acid, AcOH = acetic acid, MTF = 2-methyl- tetrahydrofuran. In these experiments, the quantities of
+3
phenol [3.38 g (36 mmol)] and Fe-EPS [33.8 mg, corresponding to about 12 mg of Fe ] were doubled in comparison with the other runs.
2
2
F(w) = (1 - a )/(1 + a - 2a cos(2pw/N)),
where a is the sample s1 lag-1 autocorrelation, w = 1 ... N/2
is the frequency index in the spatial domain and N is the
number of points of the sample (pixels). The significance (set
20
at the 5% level in this study) was calculated by multiplying
the background spectrum by the 95th percentile value for a
2
chi-square distribution c .
Three sequences were obtained from the images of SEM
micrographs (Fig. 2A), representing, respectively, an outer
unorganized state (series s1), an amorphous state (series s2) and
an ordered state (series s3) (Fig. 2B).
FT-IR analysis
During the reaction of phenol with H
2
2
O in pure water, as
reported above, a precipitate was observed. To exclude the
formation of ferric oxides, this precipitate, as well as Na-EPS
(
produced in NaC medium) and Fe-EPS (produced in FeC
medium and used as a catalyst in this reaction), were analyzed
by FT-IR. The IR spectra (KBr pellets) were recorded on an
FT-IR Nicolet Magna 750 instrument (Fig. 3).
Results and discussion
8,9
In previous reports it was demonstrated that strain BAS-10
produces, under anaerobic conditions, a polysaccharide with a
heptameric repeating structure with four a-rhamnose, two b-
glucuronic acids and one b-galactose as follows:
3
+
Each heptameric unit is able to bind Fe up to 36 ± 4.5%. The
-
1
production of total Fe-EPS is about 6.65 g L (on dry weight)
-
1
in FeC medium and that of Na-EPS is only 0.7 g L in NaC
medium. BAS-10 cells produce this Fe(III) complex very late in
the stationary phase when the medium is spent. Cells, observed
by TEM, are surrounded by a dense EPS structure with spherical
electrodense particles in the row. As seen by ESEM observations,
a cell is completely encrusted by ultradense particles containing
iron (Fig. 1B).
Fig. 2 A: SEM micrograph of purified precipitated Fe polysaccharide
obtained in water during the oxidation reaction. B: Spectral densities
of samples s2 (amorphous state) and s3 (ordered state) compared to
the 95% confidence level of the theoretical background noise spectrum
estimated from sample s1 (outer unorganized state).
When Fe-EPS was used as a catalyst in pure water,
after 3 h of contact with phenol and H
2
2
O , a precipitate was
observed. This material was analyzed to find out if iron was freed
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1405–1409 | 1407

View Article Online
back to the possible properties of the ordered structures in iron
polysaccharide after precipitation.
The more relevant results of using Fe-EPS as a catalyst in
the hydroxylation of phenol are reported in Table 1. Each test,
except for 6 and 8, was duplicated, and the results represent an
average of two runs. Using an accurate calibration method vs. a
standard for GC analyses, the mass balance of compounds (I),
(
II) and (III) in experiments 6 and 8 corresponded to 95–96%,
so indicating that no polyhydroxylated products were present in
a relevant amount.
To single out the more important factors affecting the con-
version and selectivity, experiments were carried out to change
the amount and nature of the solvents and/or of RCOOH, and
the ratio between phenol and H
2
2
O . It was also observed that a
dried, aged Fe-EPS sample showed lower or no catalytic activity
in this reaction, but the activity was recovered after pulverization
of the solid and its rehydration in water. In the absence of water,
3
+
the polymeric structure probably reorganizes itself so that Fe
becomes unavailable to the other reagents. On the basis of these
results, it is possible to see that the reaction occurs at a neutral pH
as well as when an organic acid of moderate acidity is present,
with a slight better performance in the latter case. The use of
TFA, a preferentially added acid according to the literature, is,
on the contrary, deleterious. No reaction occurs when an organic
solvent that is not miscible with water is present, probably
because, in this case, phenol prefers to stay in the organic phase.
For this aromatic substrate, a solvent such as acetonitrile is
probably useless, but for other less polar aromatic substrates,
the presence of a water-miscible organic solvent might help
contact be made between the substrate and the catalyst. The
molar ratio between the dihydroxylated products, catechol (II)
and hydroquinone (III), is about 2, but there is a slight difference
when working in a more or less concentrated mixture (compare
experiments 7 and 8, where a ratio of 2.3 vs. 1.7 was measured). A
small quantity of catechol might remain preferentially entrapped
into the catalyst as a new metal ligand, so decreasing the value
of this ratio when the mixture is less dilute. As a matter of fact,
the analytical results of these experiments before acidification of
the reaction mixture up to pH 1 showed, in general, a lower total
yield of dihydroxylated products and a worse mass balance.
Fig. 3 FTIR spectra of A: purified Na-ESP from NaC medium, B: pu-
rified Fe-ESP from FeC medium and C: precipitated Fe polysaccharide.
from the EPS and transformed in oxides during the reaction.
An X-ray diffractometry determination did not seem to show
the presence of iron crystals. Also, FT-IR confirmed that
iron in this precipitate was still bound to EPS, but contained
additional bands, very probably due to aromatic compounds,
such as phenol and its hydroxylation products (Fig. 3). The
-
1
pronounced shoulder at 3100 cm is assigned to the aromatic
-
1
C–H stretching, while the bands at 1500–1450 cm may be
due to C=C stretching vibrations. The peaks in the region
1
1
300–1200 cm- are supposed to correspond to the aromatic
C–O stretching and to the ring hydrogen rocking vibration,
respectively. Moreover, aromatic substitution bands are evident
-
1
in the region between 850 and 650 cm . Therefore, the
precipitate might be a different, less soluble Fe-EPS conformer
that includes some molecules of the phenol derivatives or a
modified iron polysaccharide complex, where some molecules
of the phenol derivatives are present as new ligands of the
metal; the latter would exhibit a lower solubility in the aqueous
medium in the absence of an organic solvent. Actually, it is not
possible to distinguish between these two possibilities. Analysis
of the SEM image of this precipitate (Fig. 2B) shows the density
spectrum of s2 (amorphous state) and s3 (ordered state), and
a comparison of them with the background noise spectrum
deduced by s1 (unorganized state). Spectral peaks above the 95%
confidence level for the background noise separate out the true
features of the series, i.e., in this case, the structural properties of
the observed material that do not originate by chance. The two
series show similar properties at very small scales, i.e., <1 nm,
also with comparable signals at 2 and 9–15 nm, therefore
pointing to nanostructures of such scales being present in both
the amorphous and ordered material. Conversely, an easily
identifiable structural organization emerges in series s3 around
the 2.5–5 nm band, as well as around 6.5 nm, while series s2
shows a low structural organization at 5.5 and 7.2 nm scales.
The relevance of these findings emerges as far as these are drawn
On the basis of the added quantity of H
2
O
2
, it is possible to
calculate that the amount of H actually used for the reaction
2
O
2
is 48–58%, while the remaining amount is decomposed. On the
contrary, the reacted phenol is nearly quantitatively converted
into catechol and hydroquinone (i.e. the selectivity of phenol
is ≥95%). These results are encouraging and might be, in our
opinion, further improved by scaling it up and fine tuning the
addition rate of H
2
O
2
. Finally, in one experiment, the reaction
◦
temperature was increased to 55 C, but without any change
in the results. In a blank experiment where Na-EPS was used
instead of Fe-EPS, no reaction was observed, as expected.
Conclusions
Although detailed mechanistic information is not currently
available, the results suggest that this oxidation reaction of
phenol in the presence of Fe-EPS could proceed by two parallel
mechanisms, and not by a pure radical mechanism, where
∑
∑
OOH and/or OH radicals are present and act as oxidizing
1
408 | Green Chem., 2010, 12, 1405–1409
This journal is © The Royal Society of Chemistry 2010

View Article Online
agents, as in the Fenton reaction. In fact, the high selectivity
of phenol suggests the presence of transient species that are
more sterically demanding than the simple hydroxyl radical, as
6 J-S. Choi, S-S. Yoon, S-H. Jang and W-S. Ahn, Catal. Today, 2006,
11, 280.
Kurian and S. Sugunan, Chem. Eng. J., 2006, 115,
1
7
M
39.
1
2,13
suggested previously.
8 J. Peng, F. Shi, Y. Gu and Y. Deng,, Green Chem., 2003, 5,
This new bio-generated Fe(III) catalyst, working under opti-
mal reaction conditions, compares quite well, in our opinion,
to other Fe-based catalytic systems considered state-of-the-
224.
9
Y. Li, H. Xia, F. Fan, Z. Feng, R. A. van Santen, E. J. M. Hensen
and C. Li, Chem. Commun., 2008, 774.
1
1
1
1
1
1
1
1
0 H Xin, A. Koekkoek, Q. Yang, R. A. van Santen, C. Li and E. J. M.
Hensen, Chem. Commun., 2009, 7590.
3,5,6,11,12,14,21
art,
with some advantages in terms of cost and low
1 C Liu, X. Ye, R. Zhan and Y. Wu, J. Mol. Catal. A: Chem., 1996,
environmental impact.
1
12, 15.
2 N. Zhang, F. Li, Q. J. Fu and S. C. Tsang, React. Kinet. Catal. Lett.,
000, 71, 393.
Acknowledgements
2
3 D Bianchi, M. Bertoli, R. Tassinari, M. Ricci and R. Vignola, J. Mol.
Catal. A: Chem., 2003, 204–205, 419.
4 H. S. Abbo, S. J. J. Titinchi, R. Prasad and S. Chand, J. Mol. Catal.
A: Chem., 2005, 225, 225.
The authors wish to thank Mr Mattia Amadio for certain
catalytic experiments, Dr Patrizia Canton for the SEM analyses
and Prof. Paola Zuddas of the University of Cagliari for ESEM
analyses of the bacteria cells.
5 F Baldi, A. Minacci, M. Pepi and A. Scozzafava, FEMS Microbiol.
Ecol., 2001, 36, 169.
6 J. S. Lolkema, H. Enequist and M. E. Van Der Rest, Eur. J. Biochem.,
1
994, 220, 469.
References
7 F. Baldi, D. Marchetto, D. Battistel, S. Daniele, C. Faleri, C. De Catro
1
2
3
M. B. Hocking and D. J. Intihar, J. Chem. Technol. Biotechnol. A,
985, 35, 365.
J. P. Hage, A. Llobet and D. T. Sawyer, Bioorg. Med. Chem., 1995, 3,
383.
D Wang, Z. Liu, F. Liu, X. Ai, X. Zhang, Y. Cao, J. Yu, T. Wu, Y.
Bai, T. Li and X. Tang, Appl. Catal., A, 1998, 174, 25.
J Guo and M. Al-Dahhan, Ind. Eng. Chem. Res., 2003, 42, 2450.
A. L. Villa, C. A. Caro and C. Montes de Correa, J. Mol. Catal. A:
Chem., 2005, 228, 233.
and R. Lanzetta, J. Appl. Microbiol., 2009, 107, 1241.
18 S. Leone, C. De Castro, M. Parrilli, F. Baldi and R. Lanzetta,
Eur. J. Org. Chem., 2007, 5183.
19 D. B. Percival and A. T. Walden, Spectral Analysis for Physical
Applications, 1993 Cambridge University Press, Cambridge.
20 D. Zanchettin, A. Rubino, P. Traverso and M. Tomasino, J. Geophys.
Res., 2008, 113, D12102.
21 K. M. Parida, S. Singha and P. C. Sahoo, Catal. Lett., 2010, 136, 155
(published after the completion of our work).
1
1
4
5
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1405–1409 | 1409