A green approach to Fenton chemistry: Mono-hydroxylation of salicylic acid in aqueous medium by the electrogeneration of Fenton's reagent

Article abstract of DOI:10.1039/c2nj21007j

First, the electrochemical characteristics of Fenton's reagent (Fe(iii), Fe(ii) and hydrogen peroxide), including its catalytic (EC′) behaviour were investigated. Second, the electrogeneration of hydrogen peroxide by a two-electron reduction of dissolved oxygen was conducted at a carbon electrode in a divided electrolysis cell and the concentration of hydrogen peroxide obtained was determined by the titration. The two approaches were then combined to give a one-pot, relatively green approach to aromatic hydroxylation reactions, with the electro-Fenton method allowing the iron to be used catalytically and the oxygen/water employed as the hydroxylation reagent by the in situ production of hydrogen peroxide. In particular, the preparative-scale hydroxylation of salicylic acid has been studied in an oxygen-saturated 0.1 M Na₂SO₄ pH 3.0 solution containing 5 mM Fe(ii) and 5 mM salicylic acid at the controlled potential of -1.0 V vs. SCE. The study showed that after the attack of the hydroxyl radical produced in situ by the electro-Fenton process on salicylic acid, 2,3- and 2,5-dihydroxybenzoic acids were detected as primary monohydroxylated products with the highest yield of ca. 21.6 ± 2.5%. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012.

Full text of DOI:10.1039/c2nj21007j

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2012, 36, 1265–1272
www.rsc.org/njc
PAPER
A green approach to Fenton chemistry: mono-hydroxylation of salicylic
acid in aqueous medium by the electrogeneration of Fenton’s reagentw
Janjira Panchompoo,a Leigh Aldous,a Mikhail Kabeshov,b Ben S. Pilgrim,b
Timothy J. Donohoeb and Richard G. Compton*a
Received (in Montpellier, France) 2nd December 2011, Accepted 28th February 2012
DOI: 10.1039/c2nj21007j
First, the electrochemical characteristics of Fenton’s reagent (Fe(III), Fe(II) and hydrogen
peroxide), including its catalytic (EC0) behaviour were investigated. Second, the electrogeneration
of hydrogen peroxide by a two-electron reduction of dissolved oxygen was conducted at a carbon
electrode in a divided electrolysis cell and the concentration of hydrogen peroxide obtained was
determined by the titration. The two approaches were then combined to give a one-pot, relatively
green approach to aromatic hydroxylation reactions, with the electro-Fenton method allowing the
iron to be used catalytically and the oxygen/water employed as the hydroxylation reagent by the
in situ production of hydrogen peroxide. In particular, the preparative-scale hydroxylation of
salicylic acid has been studied in an oxygen-saturated 0.1 M Na2SO4 pH 3.0 solution containing
5 mM Fe(II) and 5 mM salicylic acid at the controlled potential of ꢀ1.0 V vs. SCE. The study
showed that after the attack of the hydroxyl radical produced in situ by the electro-Fenton
process on salicylic acid, 2,3- and 2,5-dihydroxybenzoic acids were detected as primary
monohydroxylated products with the highest yield of ca. 21.6 ꢁ 2.5%.
Fe2+ + H2O2 - OH + ꢀOH + Fe3+
(1)
ꢂ
1. Introduction
Advanced oxidation processes (AOPs) have been widely used
for the treatment of wastewater containing organic contami-
nants such as pesticides, surfactants, colouring matters and
pharmaceuticals that are difficult to degrade.1–8 The main
mechanism of AOPs are based on the generation of hydroxyl
radical (ꢂOH), a very powerful oxidant capable of rapidly
reacting with organic compounds yielding dehydrogenated or
hydroxylated derivatives until achieving their complete miner-
alization (i.e. conversion into carbon dioxide, water and
Although the reaction involves the consumption of one mole-
cule of Fe(II) for each ꢂOH radical produced, Fe(II) catalysts can
be regenerated in the system through the reaction of Fe(III) with
H2O2 (Fenton-like-reaction) shown in eqn (2).5–7,12–15
+
Fe3+ + H2O2 - Fe2+ + OOH + H
(2)
ꢂ
According to the mechanism above (eqn (1) and (2)), it can
be seen that the iron in Fenton and Fenton-like reactions
cycles between the Fe(II) and Fe(III) oxidation states. However,
the reduction of Fe(III) is generally slower than the oxidation
Fe(II),19 hence the amounts of Fe(II) and hydrogen peroxide
used in Fenton process to generate hydroxyl radical are
often greater than the stoichiometric quantities of aromatic
contaminants.3,15,20,21 The main drawbacks of this technique
involve the high price of commercial hydrogen peroxide and
the operational costs associated with safe handling and trans-
portation of hydrogen peroxide and maintaining the correct
concentration of this compound throughout a reaction.12,18 In
particular, the decomposition of hydrogen peroxide can occur
rapidly when it has been exposed to sunlight or certain metallic
impurities and enzymes, as shown in eqn (3).
inorganic ions).1–9 In AOPs, the OH radical can be obtained
ꢂ
from Fenton (Fe(II)/H2O2, first studied by H. J. H. Fenton10 in
1894) and/or Fenton-like (Fe(III)/H2O2) processes which have
been extensively reported in the literature.1–18 Generally, the
Fenton system consists of ferrous ions mixed with hydrogen
peroxide under acidic conditions (pH ranging from 2–4),
generating the ꢂOH radical, a powerful oxidizing intermediate
as shown in eqn (1).6–18
a Department of Chemistry, Physical and Theoretical Chemistry
Laboratory, University of Oxford, South Parks Road, Oxford,
OX1 3QZ, United Kingdom
b Department of Chemistry, Chemistry Research Laboratory,
University of Oxford, Mansfield Road, Oxford, OX1 3TA,
United Kingdom. E-mail: richard.compton@chem.ox.ac.uk;
Fax: +44 (0) 1865 275 410; Tel: +44 (0) 1865 275 413
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c2nj21007j
2H2O2 (aq) - 2H2O(l) + O2 (g)
(3)
Recently, new advanced oxidation processes induced by
electrochemistry have been studied and proposed as an alternative
approach. These are known as electrochemical advanced oxidation
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1265–1272 1265

View Online
Scheme 1 Products of the attack of ꢂOH radical on salicylic acid reported by Halliwell and Grootveld.32,33
processes (EAOPs).3,5,7,18,22–26 Electrochemical methods are
the concentration of hydrogen peroxide significantly influenced the
clean, environmentally friendly, versatile and effective techniques
for the direct production (anodic oxidation) or indirect genera-
tion via Fenton’s reagent (electro-Fenton approach) of the
hydroxyl radical.3,5,7,18,22–26 In anodic oxidation, the hydroxyl
radical is generated as an intermediate and often remains
adsorbed on a high O2-overvoltage anode in a direct water
oxidation process, as shown in eqn (4).
degradation of salicylic acid.15,20 Relatively high molar ratios of
H2O2 to salicylic acid and Fe(II) were favourable. Complete
degradation of salicylic acid was achieved within 1–2 min when
the molar ratio of H2O2 to salicylic acid was 4 : 1.20 Moreover, it
has been demonstrated that salicylic acid was rapidly mineralised
by ꢂOH radical generated from Fenton’s reaction via the main
reaction intermediates of 2,3-dihydroxybenzoic, 2,5-dihydroxy-
benzoic and 2,6-dihydroxybenzoic acids.3,21 These intermediates
H2O - OH(ads) + H+ + eꢀ
(4)
ꢂ
ꢂ
were oxidised further by OH radical to, for example, a-ketoglu-
taric acid, glycolic acid, maleic acid, fumaric acid, malic acid,
tartronic acid and finally to oxalic acid which could then form an
Fe(III)-oxalate complex, indicating the complete degradation of
salicylic acid. The proposed reaction sequence for salicylic acid
mineralisation by ꢂOH radical formed from Fenton’s process was
also reported.3
In the electro-Fenton method, hydrogen peroxide is electro-
chemically generated in situ in acidic medium via the two-
electron reduction of dissolved oxygen on the surface of a
carbonaceous cathode, shown in eqn (5). It is interesting to
note that oxygen is very cheap (less than d0.1 per 100 g),27
abundant, and essentially risk-free, compared to the high cost
(at least d10 per 100 mL)28 and the difficulty in handling and
storage of commercial hydrogen peroxide discussed earlier.
In addition, the use of Fenton chemistry for the hydroxyla-
tion of salicylic acid has previously been investigated, by the
application of stoichiometric (or greater) quantities of Fenton
reagents.31,32 For example, Tomita et al.31 reported that when
salicylic acid, hydrogen peroxide and Fe(II) were mixed together,
ꢂOH radical was generated in the system and it then attacked
salicylic acid to produce the major products of 2,3- and 2,5-
dihydroxybenzoic acids with 2,3-dihydroxybenzoic acid being
prominent. Note that various amounts of catechol, 2,4- and
2,6-dihydroxybenzoic acids were also observed when the
concentration of Fe(II) was equal to or larger than that of
hydrogen peroxide. Besides, Halliwell and Grootveld32,33 also
reported that after the attack of ꢂOH radical on salicylic acid,
three products namely 2,3-dihydroxybenzoic acid (ca. 49%),
2,5-dihydroxybenzoic acid (ca. 40%) and catechol (ca. 11%),
can be detected, as illustrated in Scheme 1.
O2 (g) + 2H+ + 2eꢀ - H2O2
(5)
The electrogenerated hydrogen peroxide can be catalytically
transformed to hydroxyl radicals by means of ferrous ions as
in eqn (1). Hydroxyl radicals formed by the Fenton process
can then act as powerful oxidising agents for the complete
mineralisation of aromatic compounds,3–7,15,20,21,23 including
pharmaceuticals such as salicylic acid in wastewater.
Salicylic acid (2-hydroxybenzoic acid) is widely used in
cosmetics and pharmaceuticals.3,15,20,21,29,30 It is a key ingredient
in many skin-care products for the treatment of acne, corns
and warts due to its distinct exfoliating, antiseptic and anti-
inflammatory properties.15,20 It is also a primary hydrolysate
of acetylsalicylic acid (aspirin) which is commonly used as
anti-inflammatory medicine.3,30
As highlighted in the literature,32,33 the attack of ꢂOH
radical on salicylic acid produces 2,3-dihydroxybenzoic acid
and 2,5-dihydroxybenzoic acid as major products, suggesting
that the formation of these two dihydroxybenzoic acids could
In previous work regarding the mineralization of salicylic
acid by chemical Fenton treatment, it has been reported that
c
1266 New J. Chem., 2012, 36, 1265–1272
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Online
be applied as a responsive assay for ꢂOH radical formation.
Indeed, aromatic hydroxylation has been previously used as a
method for measuring the production of ꢂOH radical both
in vitro and in vivo.32–34
University of Oxford, UK, in DMSO-d6 and referenced to the
residual solvent peak and integrated relative to cyclohexene as
an internal standard. Chemical shifts are quoted in ppm (parts
per million) to the nearest 0.01 ppm with signal splittings
recorded as singlet (s), doublet (d) and triplet (t).1H NMR
spectra were recorded at room temperature.
In this work, the electrochemical characteristics of Fe(III) in
the presence of either hydrogen peroxide (Fenton’s reagent) or
salicylic acid are investigated. Furthermore, the electrogenera-
tion of hydrogen peroxide from oxygen reduction is performed
in aqueous media by controlled-potential coulometry, and the
concentration of in situ hydrogen peroxide produced is then
determined by the titration. Finally, the application of the
electro-Fenton method with simultaneous hydrogen peroxide
electrogenerated in situ for the mono-hydroxylation of salicylic
acid, representative of soluble aromatic compounds, yielding
the products of 2,3- and 2,5-dihydroxybenzoic acids, is first
reported and discussed here.
Electrochemical measurements were recorded using an
Autolab PGSTAT 20 computer-controlled potentiostat
(EcoChemie, Utrecht, The Netherlands) with a standard
three-electrode configuration.
A glassy carbon electrode
(GC, 3.0 mm diameter, BAS Technical, UK) was used as the
working electrode. A saturated calomel electrode (SCE) and a
carbon rod acted as the reference and counter electrodes
respectively. The GC was polished using diamond pastes
of decreasing sizes (Kemet, UK). Cyclic voltammetry was
recorded at a scan rate of 50 mV sꢀ1, except where stated.
Bulk electrolysis was performed in a divided electrolysis cell
that consisted of a reticulated vitreous carbon (RVC) working
electrode, a platinum wire (Pt) counter electrode and a reference
electrode, SCE. All electrochemical experiments were carried
out at room temperature in 0.1 M sodium sulphate solution
(pH 3.0).
2. Experimental
2.1 Reagents and equipment
Iron(III) sulphate pentahydrate (Fe2(SO4)3ꢃ5H2O, 97%), and
iron(II) sulphate heptahydrate (FeSO4ꢃ7H2O, 98%) were purchased
from Sigma-Aldrich (Gillingham, UK). Salicylic acid, sodium salt
(2-(HO)C6H4CO2Na, 99%) was supplied by BDH (Poole, UK).
All other chemicals were purchased from Sigma-Aldrich
(Gillingham, UK). All the reagents were used without further
purification. All solutions were prepared using deionised water
of resistivity not less than 18.2 MO cmꢀ1 at 298 ꢁ 2 K
(Millipore UHQ, Vivendi, UK).
2.2 Titration of hydrogen peroxide by KMnO4
The simplest and most commonly used method for the determina-
tion of hydrogen peroxide is by titration with potassium perman-
ganate (KMnO4).35–37 Potassium permanganate is a strong
oxidising agent that can quantitatively react with hydrogen
peroxide acidified with diluted sulphuric acid as shown below
in eqn (6) (in net ionic form).35–37
Sonication was carried out using a D-78224 Singen/Htw
sonicator (50/60 Hz, 80 W, UK). pH measurement was made
using a pH213 pH meter (Hanna instrument, UK).
2MnO4ꢀ + 5H2O2 + 6H+ - 2Mn2+ + 5O2 + 8H2O (6)
Proton nuclear magnetic resonance (1H NMR) spectra were
recorded on a Bruker Avance AV400 spectrometer (400 MHz)
at the Chemical Research Laboratory, Department of Chemistry,
Potassium permanganate is intensely purple, even at very
low concentration, while manganous ion (Mn2+) is colourless,
therefore, there is no need to use an indicator to indicate the
Fig. 1 Overlaid cyclic voltammograms of 5 mM Fe(III), recorded at a glassy carbon (GC) electrode, in the absence (a) and presence of 5 mM
(b) and 10 mM (c) H2O2, in 0.1 M Na2SO4 pH 3.0. All scans at 50 mV sꢀ1. Inset: Plot of cathodic peak current (Ipc) vs. concentration of H2O2
(0, 1, 2, 3, 4, 5, 8, 15, 20, 30, 40 and 50 mM) added.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1265–1272 1267

View Online
(EC0) reaction.38,39 In the catalytic reaction scheme, hydrogen
end point of the reaction as it is self-indicating.36 Further
details of the titration of hydrogen peroxide by potassium
permanganate can be found in the Supporting Informationw
(Section I).
peroxide reacts in the following chemical reaction to regenerate
Fe(III), as shown earlier in eqn (7) and (1).
3. Results and discussion
3.1 Voltammetric behaviour of Fe(III) in the presence of H2O2
The behaviour of Fe(III) in the presence of hydrogen peroxide
was first investigated. Fig. 1 displays the overlaid cyclic
voltammograms of 5 mM Fe(III) in the absence (a) and
presence of 5 mM (b) and 10 mM (c) H2O2 in 0.1 M Na2SO4
pH 3.0, measured at 50 mV sꢀ1. As illustrated in Fig. 1a, the
redox couple of Fe(III)/Fe(II) was observed with the cathodic
peak potential of ca. +0.22 V vs. SCE. The ratio of anodic
peak current (Ipa) to cathodic peak current (Ipc) was found to
be close to one, indicating a quasi-reversible one-electron
transfer process, as shown in eqn (7).38,39
Previously, several studies have investigated the reaction between
iron catalysts and hydrogen peroxide (‘‘Fenton’s reagent’’) and its
mechanism is also well-studied in the literature.1–3,7,12,13 In brief,
Fe(II) commonly used in the Fenton process reacts chemically
with H2O2, yielding ꢂOH radical and Fe(III) as shown above in
eqn (1).
According to the Fe(II)/H2O2 Fenton process, the voltam-
metric response of Fe(III) combined with H2O2 obtained in this
study confirms that after the electrochemical reduction of
Fe(III), hydrogen peroxide was catalytically reduced by electro-
generated Fe(II). With the EC0 mechanism, Fe(III) was
continuously regenerated by the coupled chemical reaction,
resulting in the increase in cathodic peak current (Ipc), as
shown in Fig. 1. In Fe(III) solutions containing a large excess of
hydrogen peroxide, the cathodic peak current (Ipc) grew with
the concentration of hydrogen peroxide (plotted as an inset in
Fig. 1) in a manner expected by EC0 process.38,39
Fe3+ (aq) + eꢀ " Fe2+ (aq)
(7)
A scan rate study between 10 to 200 mV sꢀ1 (shown in
Fig. 2) demonstrated that the cathodic peak current (Ipc) was
linear with respect to the square-root of scan rate (K, plotted
as an inset in Fig. 2). This result showed that the electrode
process is diffusion controlled.38,39 Similar electrochemical
behaviour of Fe(III)/Fe(II) redox couple has been previously
reported in the literature.40
Next, when hydrogen peroxide (in 5 mM Fe(III)) was added
to a 5 mM Fe(III) solution (Fig. 1b and c), a response
consistent with an electrochemical catalytic cycle could be
observed. Upon the addition of H2O2, the cathodic peak
current (Ipc) increased in height while the anodic peak for
the oxidation of Fe(II) became smaller, due to the chemical
consumption of Fe(II) by H2O2 corresponding to the catalytic
A scan rate study of 5 mM Fe(III) in the presence of 10 mM
H2O2 was examined and the resulting voltammograms are
shown in Fig. 3. It can be seen that the anodic peak current
(Ipa) corresponding to the oxidation of Fe(II) increased with
increasing scan rates as there was less effect of chemical
reaction involved at high scan rates.38,39 A plot of cathodic
peak current (Ipc) vs. square-root of scan rate (’, plotted as an
inset in Fig. 2) shows a linear response; however, it does not go
through the origin, which is consistent with the EC0 process
suggested.
Fig. 2 Overlaid cyclic voltammograms of scan rate study
(10–200 mV sꢀ1) for 5 mM Fe(III) in 0.1 M Na2SO4 pH 3.0, recorded
at a glassy carbon (GC) electrode. Inset: Plot of cathodic peak current
(Ipc) vs. square-root of scan rate (n1/2) at a GC electrode for 5 mM
Fe(III) itself (K, R2 = 0.998) and 5 mM Fe(III) in the presence of
10 mM H2O2 (’, R2 = 0.999).
Fig. 3 Overlaid cyclic voltammograms of scan rate study
(10–200 mV sꢀ1) for 5 mM Fe(III) in the presence of 10 mM H2O2, in
0.1 M Na2SO4 pH 3.0, recorded at a glassy carbon (GC) electrode.
c
1268 New J. Chem., 2012, 36, 1265–1272
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Online
Fe3+ + 2(C7H6O3) - [Fe(C7H5O3)2]+ + 2H+
(9)
3.2 Voltammetric behaviour of Fe(III) in the presence of
salicylic acid
The cyclic voltammetric behaviour of Fe(III)-salicylic acid
complex was also investigated. As seen in Fig. 4c, one cathodic
peak corresponding to the reduction of Fe(III) could be
observed at ca. 0.21 V (vs. SCE) more negative than that for
Fe(III) alone, owing at least in part to the fact that a stable
Fe(III)-salicylic acid complex is thermodynamically more difficult
to reduce than free Fe(III). In addition, the anodic counterpart of
that cathodic peak shifted towards more positive values and the
peak became very small and ill-defined. Another anodic peak was
also observed at ca. 1.11 V vs. SCE, with a slight shift in potential
towards the positive direction when compared to salicylic acid
itself. It should be noted that the redox couple of Fe(III)/Fe(II)
disappeared after the first cathodic scan, while the second
irreversible oxidation peak got smaller with the cycle numbers,
suggesting the passivation of the electrode surface as described
earlier.
Fig. 4 displays typical cyclic voltammograms of (a) 5 mM
Fe(III), (b) 5 mM salicylic acid and (c) 5 mM Fe(III) in the
presence of 5 mM salicylic acid, in 0.1 M Na2SO4 pH 3.0.
Fig. 4a shows the reduction of Fe(III), whereas Fig. 4b
represents the direct oxidation of salicylic acid. Note that
for the latter process, an irreversible anodic peak (Ia) was
observed only in the first anodic scan, with a peak potential
at ca. 1.06 V vs. SCE. As previously reported in the
literature,30,41–43 this anodic peak can be used for direct
electroanalytical determination of salicylic acid. On the reverse
scan, no reverse reduction peak for salicylic acid was observed,
suggesting an oxidation process followed by a fast irreversible
chemical reaction, which rapidly removed the generated
product.30,38,44 Moreover, it was found that the oxidation
peak current for salicylic acid noticeably decreased in the
second scan as the electrode was passivated. The literature
contains various reports of the electrochemical oxidation of
phenols and substituted phenols in aqueous solution.29,30,44
When the concentration of phenolic compounds was higher
than 0.5 mM, electrode passivation is observed after a single
voltammetric scan, caused by the formation of polymer films
on the electrode surface.29,30,44
A scan rate study of 5 mM Fe(III) in the presence of 10 mM
salicylic aicid was further investigated and is shown in Fig. 5.
The anodic peak current (Ipa) corresponding to the oxidation
of salicylic acid increased linearly with scan rate (plotted as an
inset in Fig. 5), and not with respect to the square root of scan
rate, demonstrating that the process of electrode reaction is
largely controlled by adsorption of the Fe(III)-salicylic acid
complex(s) at the carbon surface.
When an equimolar concentration of salicylic acid (in 5 mM
Fe(III)) was added to a 5 mM Fe(III) solution, the colour of the
solution changed from pale yellow to violet as a result of the
formation of intensely coloured complexes between salicylic
acid and Fe(III).20,45 The reaction between Fe(III) and salicylic
acid under acidic conditions, leading to the formation of a 1 : 1
and/or 1 : 2 stable complex, has been previously reported in the
literature and shown in eqn (8) and (9).20,45–47 In contrast, no
absorption spectrum in the range of 400–600 nm could be
observed for the Fe(II)-salicylic acid complex when 5 mM of
Fe(II) and salicylic acid were mixed, suggesting little, if any,
complexation occurred, at least at the concentration studied.46
3.3 In situ electrochemical generation of H2O2
The reduction of oxygen was initially investigated in a small
ratio of electrode surface area (A) to the volume of the
solution (V) condition using a bare glassy carbon (GC)
electrode in order to confirm its performance characteristics
(e.g. reduction potential of oxygen) before bulk electrolysis
was performed. Fig. 6 displays the typical cyclic voltammograms of
an oxygen-saturated (solid line) and an oxygen-free (dashed line)
0.1 M Na2SO4 pH 3.0 solution. As can be seen from Fig. 6, for
Fe3+ + C7H6O3 - [FeC7H4O3]+ + 2H+
(8)
Fig. 5 Overlaid cyclic voltammograms of scan rate study (5–200 mV sꢀ1
)
Fig. 4 Overlaid cyclic voltammograms of (a) 5 mM Fe(III), (b) 5 mM
salicylic acid and (c) 5 mM Fe(III) in the presence of 5 mM salicylic
acid, in 0.1 M Na2SO4 pH 3.0, recorded at a glassy carbon (GC)
for 5 mM Fe(III) in the presence of 10 mM salicylic acid, in 0.1 M Na2SO4
pH 3.0, recorded at a glassy carbon (GC) electrode. Inset: Plot of anodic
peak current (Ipa) corresponding to the oxidation of salicylic acid vs. scan
rate (n), with correlation coefficient (R2) = 0.994.
electrode, at 50 mV sꢀ1
.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1265–1272 1269

View Online
hydrogen peroxide, each drop was decolourised until all
hydrogen peroxide was used up (reaction shown in eqn (6)).
The next drop of KMnO4 added coloured the solution,
indicating the end-point of the titration. The actual content
of hydrogen peroxide formed by the electrolysis process can
then be determined.
Fig. 7 displays the plot of concentrations of hydrogen
peroxide experimentally determined by the titration (solid
line) against those theoretically calculated based on Faraday’s
law (dashed line). The experiments (either bulk electrolysis or
titration) were performed in triplicate and the results were the
average of three independent experiments. As shown in Fig. 7,
the actual concentrations of hydrogen peroxide electrochemically
generated in situ with four different amounts of charge consumed
(2.41 C, 4.82 C, 7.24 C and 9.65 C) were found to be 1.21 mM,
2.30 mM, 3.60 mM and 4.23 mM, respectively. These experi-
mental values agree with the theoretical data calculated earlier
with the electrolysis current efficiency for in situ hydrogen
peroxide formation of ca. 85–97%.
Fig. 6 Overlaid cyclic voltammograms for a bare glassy carbon (GC)
electrode, recorded in an O2-saturated (solid line) and a N2-saturated
(dashed line) Na2SO4 0.1 M pH 3.0 solution, at 50 mV sꢀ1
.
ꢂ
an oxygen-saturated solution, a large irreversible reduction peak
was observed at ca. ꢀ0.75 V vs. SCE, while no obvious voltam-
metric peak could be seen in the absence of oxygen. These
voltammetric results correspond to the two-proton-two-electron
reduction of oxygen, leading to the formation of hydrogen
peroxide,23 as shown in eqn (5).
3.4 Bulk reaction of the electrogenerated OH radical with
salicylic acid
A preparative-scale hydroxylation of salicylic acid was next
performed at ꢀ1.0 V vs. SCE in a divided electrolysis cell with
a reticulated vitreous carbon (RVC) cathode in 10.00 mL of an
oxygen-saturated 0.1 M Na2SO4 pH 3.0 solution containing
5 mM salicylic acid and 5 mM Fe(II). The electrolysis solution
was thoroughly stirred during the reaction by continuous
bubbling of oxygen gas. At a controlled potential of ꢀ1.0 V
vs. SCE, hydrogen peroxide was first produced in situ via a
two-electron O2 reduction process, followed by a reaction with
Fe(II), generating hydroxyl radical, hydroxide and Fe(III), as
described earlier. Next, the in situ electrogenerated hydroxyl
radical was expected to react with salicylic acid in the
solution.32,33 The total amount of charge (C) passed during
each electrolysis was monitored and set to the exact values,
shown in Table 1. It is important to note that apart from the
two-electron reduction of O2 to H2O2 taking place during the
controlled-potential electrolysis, Fe(III), simultaneously generated
Potentiostatic electrolysis was subsequently conducted in order
to determine whether oxygen could be successfully electrolysed to
produce hydrogen peroxide. In summary, controlled-potential
coulometry was performed in a 10.00 mL of oxygen-saturated
0.1 M Na2SO4 pH 3.0 solution at ꢀ1.0 V vs. SCE in a divided
electrolysis cell using a reticulated vitreous carbon (RVC) electrode
as a working electrode. The electrolysis solution was thoroughly
stirred during the reaction by continuous bubbling of oxygen gas.
The total number of coulombs (Q) consumed in each electrolysis
was monitored and deliberately set to be 2.41 C, 4.82 C, 7.24 C
and 9.65 C. According to Faraday’s law, the amount of oxygen
electrolysed in each electrolysis (which is theoretically in 1 : 1
stoichiometric ratio of hydrogen peroxide formed) can be
calculated by using eqn (10):38
Q = nFN
(10)
where Q represents the total amount of charge passed in the
electrolysis (C); n is the number of electrons involved in
the reaction; F is the Faraday constant (C molꢀ1) and N is the
mole number of substrates that have been electrolysed (mol).
Based on Faraday’s law, the concentration of hydrogen
peroxide generated in a 10 mL of electrolysis solution with
four different amounts of charge passed through (2.41 C,
4.82 C, 7.24 C and 9.65 C) could then be calculated and
inferred to be 1.25 mM, 2.50 mM, 3.75 mM and 5.0 mM,
respectively.
The hydrogen peroxide formed by the electrolysis process
was next quantified by titration with potassium permanganate
(KMnO4),36,37 detailed in Experimental section. Simply,
5.00 mL of each resulting hydrogen peroxide solution
obtained from the electrolysis described above was first
acidified with H2SO4 and subsequently titrated with KMnO4.
When KMnO4 was added dropwise to an acidified solution of
Fig. 7 Plot of H2O2 concentrations electrochemically generated
in situ determined experimentally by titration (solid line) versus
theoretically calculated from Faraday’s law (dashed line).
c
1270 New J. Chem., 2012, 36, 1265–1272
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Online
Table 1 The total amount of charge (C) consumed in each bulk
electrolysis of an O2-saturated 0.1 M Na2SO4 pH 3.0 solution containing
5 mM salicylic acid and 5 mM Fe(II) with expected concentration of
hydrogen peroxide (mM) electrogenerated in situ at ꢀ1.0 V vs. SCE,
calculated from Q = nFN, based on a two-electron O2 reduction process.
Also shown is the estimated percentage of charge passed corresponding
to a 1 : 1 stoichiometric reaction between 5 mM Fe(II) and H2O2 formed
in the system
of repeating the experiments in triplicate. As seen in Fig. 8, the
highest product yield of ca. 21.6 ꢁ 2.5% was obtained when
33.3% of charge (4.82 C) was consumed. After that the reaction
yield was moderately decreased with further increasing electric
charge passed as 2,3- and 2,5-dihydroxybenzoic acids were
oxidised further to other reaction intermediates by more free
ꢂOH radical electrogenerated in situ as discussed earlier.3,15,20,21
Typically, the number of electric charge passed during the
electrolysis corresponded to the amount of H2O2 formed which
then reacted with Fe(II), producing ꢂOH radical. Thus, this
simply implies that the greꢂater the amount of charge passed,
the greater the number of OH radical electrogenerated. The
complete (100%) conversion of salicylic acid to its mono-
hydroxylated products (before further oxidation to other
intermidiates) was not achieved as these primary mono-
hydroxylated products which are more electron-rich aromatic
Amount of charge
passed during each
electrolysis (C)
Expected concentration of Percentage of
H2O2 formed in situ during charge passed
electrolysis (mM)
(%)
0
0
0
6.7
0.97
2.41
4.82
7.24
9.65
14.48
0.34
0.83
1.67
2.50
3.33
5.00
16.7
33.3
50.0
66.7
100.0
ꢂ
compounds could be hydroxylated further by more free OH
in the system by the reaction of H2O2 and Fe(II), could also
accept one electron, being reduced to Fe(II). Considering the
electric charge consumption in the two-electron O2 reduction
reaction, the concentration of hydrogen peroxide formed in the
system could then be theoretically estimated based on Faraday’s
law (shown in Table 1). In addition, the percentage of charge
passed in each electrolysis could be calculated by taking a 1 : 1
stoichiometric reaction between 5 mM Fe(II) and hydrogen
peroxide formed in situ into account (shown in Table 1).
radical electrogenerated in situ, resulting in the formation of
polyhydroxylated products.
Note that the work reported here is a proof-of-concept
study for the aromatic hydroxylation by hydroxyl radical
electrogenerated by electro-Fenton reaction. Future research
is planned to investigate the hydroxylation of aromatic drugs
since the metabolites of aromatic drugs are often hydroxylated
compounds.
The resulting mixture at the end of each electrolysis was
then acidified to pH 1 by the addition of aqueous 1.0 M HCl.
The aqueous layer was next extracted with EtOAc (5 ꢄ 10 mL).
The combined organics were dried over Na2SO4, filtered and
the solvent was removed in vacuo to give the product. The
reaction products obtained from the electrolysed mixture before
and after controlled potential electrolysis reaction were then
analysed by proton nuclear magnetic resonance (1H NMR)
spectroscopy. As demonstrated in the Supporting Informationw
(Section II), four signals of protons could be observed in both
cases at d 6.95 ppm (dd, H-3 of salicylic acid), 7.51 ppm
(td, H-4 of salicylic acid), 6.91 ppm (td, H-5 of salicylic acid)
and 7.80 ppm (dd, H-6 of salicylic acid), corresponding to
1H NMR spectrum of salicylic acid.48–50 After the bulk
reaction, additional proton signals for hydroxylated products
namely 2,3- and 2,5-dihydroxybenzoic acids were observed,
4. Conclusions
The electrochemical behaviour of Fe(III) in the presence of
either hydrogen peroxide (Fenton’s reagent) or salicylic acid
in 0.1 M Na2SO4 pH 3.0 has been investigated by cyclic
voltammetry. The controlled-potential electrolysis was also
conducted for the electrochemical generation of hydrogen
peroxide via a two-proton-two-electron reduction of oxygen.
The major advantage of hydrogen peroxide formed in situ
within the electrochemical cell is that there is no need to store
or transport the commercial one which is difficult to handle
and requires chemical stabilisation. The actual concentration
of hydrogen peroxide obtained from the electrolysis was then
determined by the titration with potassium permanganate and
ꢂ
confirming the hydroxylation of salicylic acid by OH radical
electrogenerated in situ via electro-Fenton process. The signals
at d 7.24 ppm (dd, H-6 of 2,3-dihydroxybenzoic acid) and
7.15 ppm (d, H-6 of 2,5-dihydroxybenzoic acid) were used for
analysis due to their position in an otherwise clear region of
the spectrum.
In order to determine the total yield of these two mono-
hydroxylated products, cyclohexene (–CHQCH–, d 5.60 ppm,
1H NMR not shown) was used as an internal standard. The
conversion percentage of salicylic acid to its hydroxylated
derivatives was obtained by comparing the integration of
corresponding peaks in the product and starting material.
Both signals were normalised with respect to the cyclohexene
signal. Fig. 8 displays the percentage conversion of salicylic
acid to 2,3- and 2,5-dihydroxybenzoic acids (with 2,3-dihydroxy-
benzoic acid being prominent) versus the estimated percentage of
charge passed during the electrolysis. Error bars are the result
Fig. 8 Plot of % conversion of salicylic acid to its mono-hydroxy-
lated derivatives vs. % charge passed during the electrolysis.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1265–1272 1271

View Online
the electrolysis current efficiency for hydrogen peroxide electro-
generated was found to be ca. 85–97%. Furthermore, the
preparative-scale hydroxylation of salicylic acid by hydroxyl
radical electrogenerated from electro-Fenton process has
been studied. An attack of hydroxyl radical on salicylic acid
produced 2,3- and 2,5-dihydroxybenzoic acids with 2,3-dihydroxy-
benzoic acid as a major product. The highest yield of these two
monohydroxylated products was found to be ca. 21.6 ꢁ 2.5%.
Despite the low product yield, this electrochemically assisted
Fenton reaction appears to be a simple, fast and cheap
method for preparing such hydroxylated derivatives in small
amounts.
21 X. M. Chen, D. R. da Silva and C. A. Martı
Chin. Chem. Lett., 2010, 21, 101–104.
´
nez-Huitle,
22 M. A. Oturan, N. Oturan, C. Lahitte and S. Trevin, J. Electroanal.
Chem., 2001, 507, 96–102.
23 M. A. Oturan and E. Brillas, Port. Electrochim. Acta, 2007, 25, 1–18.
24 M. A. Oturan, J. Pinson, N. Oturan and D. Deprez, New J. Chem.,
1999, 23, 793–794.
25 N. Oturan, S. Trajkovska, M. A. Oturan, M. Couderchet and
J.-J. Aaron, Chemosphere, 2008, 73, 1550–1556.
26 M. A. Oturan and J. Pinson, J. Phys. Chem., 1995, 99, 13948–13954.
27 ‘‘Oxygen’’, to be found under http://www.chemicool.com/ele
ments/oxygen.html, 13 Jun. 2011. Web. 1/26/2012.
28 ‘‘Hydrogen peroxide solution’’, to be found under http://www.
sigmaaldrich.com/, Web. 1/26/2012.
29 I. Gualandi, E. Scavetta, S. Zappoli and D. Tonelli, Biosens.
Bioelectron., 2011, 26, 3200–3206.
30 W.-D. Zhang, B. Xu, Y.-X. Hong, Y.-X. Yu, J.-S. Ye and
J.-Q. Zhang, J. Solid State Electrochem., 2010, 14, 1713–1718.
31 M. Tomita, T. Okuyama, S. Watanabe and H. Watanabe,
Arch. Toxicol., 1994, 68, 428–433.
32 M. Grootveld and B. Halliwell, Biochem. J., 1986, 237, 499–504.
33 B. Halliwell and M. Grootveld, FEBS Lett., 1987, 213, 9–14.
34 B. Halliwell and H. Kaur, Free Radical Res., 1997, 27, 239–244.
35 G. H. Jeffery, J. Bassett, J. Mendham and R. C. Denney, Vogel’s
textbook of quantitative chemical analysis, Longman Scientific &
Technical, Harlow, 1989.
Acknowledgements
Syngenta and GlaxoSmithKline are acknowledged for partial
funding of this work.
References
1 M. Pera-Titus, V. Garcı
S. Esplugas, Appl. Catal., B, 2004, 47, 219–256.
2 A. S. Stasinakis, Global NEST Journal, 2008, 10, 376–385.
a-Molina, M. A. Banos, J. Gimenez and
´ ´
36 J. S. Reichert, S. A. McNeight and H. W. Rudel, Ind. Eng. Chem.,
Anal. Ed., 1939, 11, 194–197.
3 E. Guinea, C. Arias, P. L. Cabot, J. A. Garrido, R. M. Rodrı
´
F. Centellas and E. Brillas, Water Res., 2008, 42, 499–511.
guez,
37 C. E. Huckaba and F. G. Keyes, J. Am. Chem. Soc., 1948, 70,
1640–1644.
4 P. R. Gogate and A. B. Pandit, Adv. Environ. Res., 2004, 8, 501–551.
5 J. Wang and L. Xu, Crit. Rev. Env. Sci. Tec., 2012, 42, 251–325.
6 M. A. Oturan, N. Oturan, M. C. Edelahi, F. I. Podvorica and
K. E. Kacemi, Chem. Eng. J., 2011, 171, 127–135.
38 A. J. Bard and L. R. Faulkner, Electrochemical methods: fundamentals
and applications, Wiley, New York, 2001.
39 R. G. Compton and C. E. Banks, Understanding voltammetry,
Imperial College Press, London, 2011.
7 E. Brillas, I. Sire
´
6570–6631.
8 S. H. Lin and C. F. Peng, Environ. Technol., 1995, 16, 693–699.
9 M. Barbeni, C. Minero, E. Pelizzetti, E. Borgarello and
N. Serpone, Chemosphere, 1987, 16, 2225–2237.
10 H. J. H. Fenton, J. Chem. Soc. Trans., 1894, 65, 899–910.
11 C. Walling and S. Kato, J. Am. Chem. Soc., 1971, 93, 4275–4281.
12 S. Wang, Dyes Pigm., 2008, 76, 714–720.
13 M. A. Tony, Y. Q. Zhao and M. F. El-Sherbiny, Chem. Eng.
Commun., 2010, 198, 442–452.
s and M. A. Oturan, Chem. Rev., 2009, 109,
40 M. H. Bridge, E. Williams, M. E. G. Lyons, K. F. Tipton and
W. Linert, Biochim. Biophys. Acta, Mol. Basis Dis., 2004, 1690,
77–84.
41 M. Polakovic, T. Gorner, F. Villieras, P. de Donato and
´
J. L. Bersillon, Langmuir, 2005, 21, 2988–2996.
42 F. Salinas, A. M. de la Pena, I. Duran-Meras and M. S. Duran,
Analyst, 1990, 115, 1007–1011.
43 Y.-S. Fung and S.-F. Luk, Analyst, 1989, 114, 943–945.
44 A. A. J. Torriero, J. M. Luco, L. Sereno and J. Raba, Talanta,
2004, 62, 247–254.
14 P. Kumar, H. Nikakhtari, M. Nemati and G. A. Hill, Environ.
Technol., 2010, 31, 47–52.
45 D. Waqar and W. Hussein, Pak. J. Pharm. Sci., 2011, 24, 31–35.
46 B. Oktavia, L. W. Lim and T. Takeuchi, Anal. Sci., 2008, 24,
1487–1492.
47 M. Rahni and B. Legube, Water Res., 1996, 30, 1149–1160.
48 N. E. Jacobsen, NMR spectroscopy explained: simplified theory,
applications and examples for organic chemistry and structural
biology, Wiley, Hoboken, N.J., 2007.
´
15 C. L. P. S. Zanta and C. A. Martınez-Huitle, Braz. J. Chem. Eng.,
2009, 26, 503–513.
16 W. G. Kuo, Water Res., 1992, 26, 881–886.
17 C. Walling, Acc. Chem. Res., 1975, 8, 125–131.
18 J. M. Peralta-Herna
T. W. Chapman, M. I. Maldonado and L. A. Godı
2006, 40, 1754–1762.
´
ndez, Y. Meas-Vong, F. J. Rodrı
´
guez,
nez, Water Res.,
´
49 A. M. Castillo, L. Patiny and J. Wist, J. Magn. Reson., 2011, 209,
123–130.
19 J. J. Pignatello, Environ. Sci. Technol., 1992, 26, 944–951.
20 A. Goi, Y. Veressinina and M. Trapido, Chem. Eng. J., 2008, 143, 1–9.
50 T. M. Jadrijevic-Mladar and D. V. Topic, Acta Pharmaceut., 2004,
54, 177–191.
c
1272 New J. Chem., 2012, 36, 1265–1272
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012