New evidence against hydroxyl radicals as reactive intermediates in the thermal and photochemically enhanced fenton reactions

Article abstract of DOI:10.1021/jp980129j

During the oxidative degradation of 2,4-dimethylaniline (2,4-xylidine) by means of the H₂O₂/UV method, a series of hydroxylated aromatic amines are formed, this result confirming the role of the hydroxyl radical as an initiator of the oxidative chain reaction. Thermal or photochemically enhanced Fenton reactions in the presence of 2,4-dimethylaniline (2,4-xylidine) yield primarily 2,4-dimethylphenol as an intermediate product, the genesis of which may only be explained by an electron transfer mechanism. Experimental evidence for such a mechanism is presented, and values for the quantum yields of the photochemically enhanced reduction of iron(III) to iron(II) in aqueous solutions of 2,4-xylidine are given.

Full text of DOI:10.1021/jp980129j

5542
J. Phys. Chem. A 1998, 102, 5542-5550
New Evidence against Hydroxyl Radicals as Reactive Intermediates in the Thermal and
Photochemically Enhanced Fenton Reactions
Stefan H. Bossmann,* Esther Oliveros, Sabine Go1b, Silvia Siegwart, Elizabeth P. Dahlen,
Leon Payawan, Jr., Matthias Straub, Michael Wo1rner, and Andre´ M. Braun
Lehrstuhl fu¨r Umweltmesstechnik, Engler-Bunte-Institut, UniVersita¨t Karlsruhe, D-76128 Karlsruhe, Germany
ReceiVed: NoVember 7, 1997; In Final Form: February 10, 1998
During the oxidative degradation of 2,4-dimethylaniline (2,4-xylidine) by means of the H2O2/UV method, a
series of hydroxylated aromatic amines are formed, this result confirming the role of the hydroxyl radical as
an initiator of the oxidative chain reaction. Thermal or photochemically enhanced Fenton reactions in the
presence of 2,4-dimethylaniline (2,4-xylidine) yield primarily 2,4-dimethylphenol as an intermediate product,
the genesis of which may only be explained by an electron transfer mechanism. Experimental evidence for
such a mechanism is presented, and values for the quantum yields of the photochemically enhanced reduction
of iron(III) to iron(II) in aqueous solutions of 2,4-xylidine are given.
Introduction
the iron(II) cation. For reasons of simplicity, we utilize for
Fe2+aq in reaction 1.2 the monomeric complex [Fe(OH)(H2O)5]+
Photochemical degradation processes (also referred to as
advanced oxidation processes, AOP) have been proposed in
recent years for the treatment of ground, surface, and waste-
waters containing biocidal or nonbiodegradable organic com-
pounds.1 AOP are mainly based on oxidative degradation
reactions, most often initiated by hydroxyl radicals which may
be generated by various methods (e.g. UV photolysis of
hydrogen peroxide, TiO2 photocatalysis, vacuum ultraviolet
(VUV) photolysis of water).1,2 Among AOP, the Fenton
reaction3 and especially the photochemically enhanced Fenton
reaction are considered most promising for the remediation of
highly contaminated waters. A wide range of applications have
been reported,4,5 including the successful treatment of industrial
wastewaters on a large pilot scale (500 L) by the photochemi-
cally enhanced Fenton reaction.6
Although the Fenton reagent (a mixture of hydrogen peroxide
and iron(II) salt) has been known for more than a century3a,b
and proven long since a powerful oxidant, the mechanism of
the Fenton reaction is still under intense and controversial
discussion. According to the classic interpretation of Haber and
Weiss,7 the reaction of iron(II) with hydrogen peroxide (H2O2)
in aqueous solution leads to the formation of a hydroxyl radical
(HO•) (reaction 1.1). In 1951, Barb et al.8 investigated further
k ≈ 2 × 106
s
[Fe(OH)(H2O)5]+ + H2O
-18
2 r
[Fe(OH)(H2O2)(H2O)4]+ + H2O (1.2)
instead of [(H2O)4Fe(OH)2Fe(H2O)4]2+, which may prevail
depending upon the reaction conditions. By this exchange
mechanism, a steady-state concentration of iron(II) bound to
H2O2 is reached.
Note the remarkable difference between the monomolecular
rate constant of ligand exchange in high-spin iron(II) complexes
(2 × 106 s-1 10
) and the bimolecular rate constant of the thermal
Fenton reaction (60-80 M-1 s-1).4d From that comparison, it
is clear that the formation of the key intermediate of the thermal
Fenton reaction is not diffusion controlled as would be expected
for an outer-sphere electron-transfer reaction.9 In fact, an inner-
sphere two-electron-transfer reaction slowly takes place within
[Fe(OH)(H2O2)(H2O)4]+ and the intermediate iron(IV) complex,
Fe4+ ([Fe(OH)3(H2O)4]+), is formed (reaction 1.3).
aq
[Fe(OH)(H2O2)(H2O)4]+ f [Fe(OH)3(H2O)4]+ (1.3)
Indeed, the existence of unusually charged metal complexes
Fe4+aq, as well as Fe5+aq and Fe6+aq, has been proven by stopped-
flow experiments in combination with UV/vis absorption
Fe2+aq + H2O2 f Fe3+aq + HO• + HO-
(1.1)
spectroscopy,11 by chemical analysis of products formed during
the mechanism of the Fenton reaction and proposed a second-
order kinetic model, whereas, more than two decades later,
Walling3c presented further evidence of the involvement of
hydroxyl radicals in the oxidation of various organic compounds
by the Fenton reagent.
However, recent thermodynamic calculations have demon-
strated that an outer-sphere electron-transfer reaction between
Fe2+aq and H2O2, as it is rationalized by the classic mechanism
proposed by Haber and Weiss (reaction 1.1), cannot take place,
because the formation of the intermediate H2O2- is not favored.9
In contrast, the formation of a hydrated iron(II)-H2O2 complex
is thermodynamically favored. This reaction consists of a ligand
exchange reaction (H2O2 vs H2O) in the first ligand sphere of
13
the Fenton oxidation,12 and by pulse radiolysis of Fe3+
in
aq
the presence of inorganic ligands (HO- and P2O74-). Further-
more, in the absence of H2O2 as reducing agent for highly
charged iron cations, Fe4+aq and Fe6+aq possess surprisingly long
lifetimes. The decays observed for both species in aqueous
solution at pH 3-7 follow first-order kinetics with rate constants
of approximately 2 s-1 14
.
The intermediate iron(IV) complex (Fe4+aq) may react further
leading to the formation of a free hydroxyl radical and Fe3+
aq
([Fe(OH)(H2O)5]2+), as shown in reaction 1.4.
[Fe(OH)3(H2O)4]+ + H2O f
[Fe(OH)(H2O)5]2+ + HO• + HO- (1.4)
* To whom correspondence should be addressed.
S1089-5639(98)00129-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/17/1998

Evidence against HO• Radicals in Fenton Reactions
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5543
SCHEME 1: Mechanistic Presentation of Possible
Reactions Involved in the Thermal Fenton Reaction with
Simplified Notations Used for the Various Iron
Complexes)
SCHEME 2: Possible Reaction Pathways Involving
Hydroxyl Radicals, Metal Cations and Organic
Substrates in Aqueous Solution
TABLE 1: Reduction Potentials of Fe2+aq, Fe3+aq, H2O2,
and the Reactive Intermediates HO• and Fe4+
aq
redox couple
HO•aq/H2Oaq
HO•aq/HO-
E0 (V vs NHE)
2.59 (pH ) 0)a
1.64 (pH ) 14)a
≈0.9 (pH ) 0) b
≈1.3 (pH ) 7) b
≈1.8 (pH ) 0) b
≈1.4 (pH ) 7) b
0.771 (pH ) 0-3) b
1.776 (pH ) 0)a
0.878 (pH ) 14)a
0.682 (pH ) 0)a
-0.076 (pH ) 14)a
aq
Fe3+aq/FedO2+ (porphyrine chelate)
Fe3+aq/FedO2+ (porphyrine chelate)
Fe3+aq/Fe4+
aq
Fe3+aq/Fe4+
aq
Fe2+aq/Fe3+
aq
H2O2/H2O
H2O2/H2O
H2O2/O2
H2O2/O2
a Reference 19. b References 16-18.
(albeit with lower efficiency), or by electron transfer (Scheme
2a).19 Therefore, aminophenols should be generated during the
reaction of 2,4-xylidine with hydroxyl radicals (Scheme 2b).
The concentration of these products may strongly depend on
the reaction conditions and on the irradiation times, but their
presence would support a hydroxyl radical mechanism.
In contrast, the reaction of a metal cation, as for instance
Fe4+aq, with an aliphatic or an aromatic hydrocarbon proceeds
exclusively by an electron-transfer mechanism21 (Scheme 2c),
because neither addition nor hydrogen abstraction is possible.
Under these conditions, the oxidation of 2,4-xylidine should not
lead to the formation of aminophenols.
Reaction pathways which may be of importance for the
understanding of the mechanism of the Fenton reaction are
shown in Scheme 1, which has been adapted from ref 15.
Although until recently it was a widely accepted paradigm
in the field of AOP research that oxidations using the Fenton
or the photochemically enhanced Fenton reactions are initiated
by free hydroxyl radicals,4-6 the question actually arises whether
HO• production (reaction 1.4) is not too slow to compete with
direct electron transfer between the substrate and an hydrated
higher-valent iron species (most likely Fe4+aq). The reduction
potentials of the reactive intermediates (HO•, Fe4+aq), as well
as Fe2+aq, Fe3+aq, and H2O2, are listed in Table 1.16-19
Depending on the substrate, reactive intermediates other than
the hydroxyl radical have been proposed for the Fenton reaction.
In the case of compounds which may form highly stabilized
iron(II) complexes25 (e.g. EDTA), Sawyer et al.20 suggested that
the base-induced nucleophilic addition of H2O2 to the electro-
philic iron center of these complexes yields the reactive
intermediate of Fenton reagents.
One of the aims of this work was then to provide new
experimental evidence for the involvement of hydrated higher-
valent iron species as key intermediates of both thermal and
photochemically enhanced Fenton reactions. 2,4-Xylidine (2,4-
dimethylaniline) was chosen as the organic substrate, based on
the fact that different reaction products should be formed
depending on the reactive intermediates involved. Indeed,
hydroxyl radicals react with organic compounds by addition to
double bonds possessing a sufficient electron density, by
hydrogen abstraction from alkyl groups or hydroxyl groups
A second topic of this investigation concerns the mechanism
of the photochemically enhanced recycling of Fe2+aq. Indeed,
it has been shown recently that UV/visible irradiation accelerates
both Fenton (H2O2/iron(II)) and Fenton-like (H2O2/iron(III))
reactions, improving the degradation rates of various organic
contaminants.5
The reduction of Fe3+aq to Fe2+aq by H2O2 during the thermal
Fenton reaction proceeds in three consecutive steps.4 The first
step consists of the formation of an hydrated iron(III)-H2O2
complex ([Fe(OH)(HO2)(H2O)4]+ at pH 3, reaction 2.1). An
[Fe(OH)(H2O)5]+ + H2O2 h [Fe(OH)(HO2)(H2O)4]+ +
H3O+ (2.1)
inner-sphere electron-transfer reaction then occurs in the
iron(III)-H2O2 complex, and iron(III) is reduced to iron(II)
(reaction 2.2). A diffusion-controlled outer-sphere electron-
transfer reaction then occurs between a second Fe3+aq complex

5544 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Bossmann et al.
[Fe(OH)(HO2)(H2O)4]+ + H2O h [Fe(OH)(H2O)5]+ +
•
HO2 (2.2)
•
([Fe(OH)(H2O)5]2+) and the hydroperoxyl radical (HO2 ) formed
in reaction 2.2, thus regenerating Fe2+ (reaction 2.3). The
aq
[Fe(OH)(H2O)5]2+ + H2O + HO2 f [Fe(OH)(H2O)5]+ +
•
O2 + H3O+ (2.3)
reaction rate constants for the equilibrium reactions 2.1 and 2.2
are k(2.1) ) 0.020 M-1 s-1 and k-(2.1) ) 1.2 × 106 M-1 s-1 22
.
Reaction 2.3 is irreversible (k(2.3) ) 2.88 × 104 M-1 s-1), and
therefore the reaction sequence is shifted toward the thermal
reduction of Fe3+ to Fe2+ and the corresponding oxidation
Figure 1. Relative spectral distribution of the medium-pressure
mercury lamp (Heraeus, TQ 718) used for polychromatic irradiation
experiments.
aq
aq
of H2O2 to O2. The enhancement of the reaction rate of iron(III)
reduction under irradiation might be due to a photoinduced
oxidation (by an inner-sphere electron-transfer reaction) of the
ligands of various iron(III) complexes (with H2O,5f,g,23 with the
organic substrate, or with its intermediates of degradation5h,24).
from databases (MS, Wiley, Reference Mass Spectra, 1990;
FTIR, EPA, Reference FTIR Spectra, 1992). For the structural
determination of the reaction products, samples of 2,4-xylidine,
2,4-dimethylphenol, 3-hydroxy-2,4-dimethylaniline, and oxalic
acid were co-injected (dissolved in CHCl3), and in all cases, a
superimposition of the corresponding peaks was observed.
Furthermore, calibration curves for the compounds listed above
were recorded. The results for all calibration samples could be
fit to linear calibration curves.
The efficiency and the rate constants of the reduction of Fe3+
aq
to Fe2+ are of particular importance for the optimization of
aq
the photochemically enhanced Fenton reaction. In this publica-
tion, we report values of the quantum yields for the photore-
duction of iron(III) to iron(II) in aqueous solution in the absence
and in the presence of oxalic acid or 2,4-xylidine. It is of special
interest whether the quantum yields reported for the ferrioxalate
actinometry,26 where oxalate serves as ligand for iron(III) (Φ
) 1.24 at λ ) 254 nm) are unique or whether aromatic amines
(and other classes of organic compounds) react with similar
quantum yields.
Ferrioxalate Actinometry. The radiant power [RP (W)]
emitted by the medium-pressure mercury lamp (Heraeus, TQ
718, quartz filter, electrical power consumption 500 W)
employed in all photochemical experiments described here was
determined by using ferrioxalate actinometry26 (Fe3+aq, 2.40 ×
10-2 M; oxalic acid, 2.40 × 10-2 M). The method is based on
the photochemical reduction of iron(III) to iron(II) during
photooxidation of oxalic acid to CO2 (see Results and Discus-
Experimental Section
sion). The photochemically generated Fe2+ was measured
aq
Chemicals. FeSO4‚7H2OFe2(SO4)2‚11H2O, H2O2, H2SO4,
Na3PO4, KMnO4, Na2SO3, KI, oxalic acid, 1,10-phenanthroline,
and acetic acid were purchased from Merck. 2,4-Xylidine, 2,4-
dimethylphenol, and 3-hydroxy-2,4-dimethylaniline were bought
from Aldrich. Water was of bidistilled quality (UHQ II). All
chemicals were ACS grade, except KNO3 (Sigma), which was
ultrapure.
UV/vis spectra were recorded using an Hewlett-Packard
5800(II) diode-array spectrophotometer. A Metrohm pH ana-
lyzer was employed (E 512). H2O2 was analyzed by classic
KMnO4 titration.25
DPP Spectra. All electrochemical experiments were per-
formed with a computer-interfaced potentiostat/galvanostat
(Princeton Applied Research, model 263A and software 270)
using a conventional one-compartment cell and three-electrode
configuration. The redox potentials were determined from
differential pulse voltammetric (DPV) data. The setting of the
pulse amplitude was 50 mV, the scan rate was 5 mV s-1, and
a glassy carbon electrode (3 mm diameter) was used as the
working electrode. Unless otherwise indicated, the electrode
potentials were referenced to the SHE. Before each experiment,
the electrolyte solutions were purged with argon.
GC-MS/FTIR Analysis. For the qualitative and quantitative
analysis of the reaction products generated during the H2O2
photolysis and the thermal and the photochemically enhanced
Fenton reactions, the samples (2 mL) were injected into a GC
(HP 5971A MSD, mass selective detector), coupled with a HP
5965B ID (infrared detector). A HP-INNOWAX capillary
column (cross-linked poly(ethylene glycol)) was employed. All
reaction products were identified by a combination of MS and
FTIR spectroscopy in comparison with analytical data available
quantitatively by using the UV/vis absorption of tris(1,10-
phenanthroline)iron(II) ([Fe(phen)3]2+: ꢀ(510 nm) ) 11 100 M-1
cm-1), which is formed from Fe2+ and 1,10-phenanthroline
aq
in 0.50 M acetic acid.26 Note that the solutions were diluted
by a factor of 8 before taking UV/vis measurements in 1 cm
quartz cells in order to ensure complete solubility of [Fe(phen)3]2+.
Under the experimental standard conditions of the ferrioxalate
actinometry, 9.37 × 10-5 M s-1 of Fe2+ was formed. The
aq
incident radiations between 220 and 600 nm were absorbed by
the chemical actinometer. From this quantitative data and using
the relative emission spectrum of the lamp TQ 718, which is
shown in Figure 1, the value for the radiant power emitted by
the lamp (RP ) 155 ( 4.0 W) was calculated according to eq
3.1, where n(Fe2+) is the number of Fe2+aq ions formed during
n(Fe2+)
RP )
(3.1)
Se,λ
t
(1 - 10-A )Φλ
λ
∑
[
]
Eph,λ
the irradiation time (t in s), Se,λ is the relative spectral distribution
of the radiant power emitted by the lamp (Figure 1), Eph,λ is
the energy of a photon of wavelength λ (in J), Aλ is the average
absorbance of the actinometric solution at wavelength λ during
irradiation, and Φλ is the quantum yield of the chemical
actinometer at wavelength λ.26
As it becomes clear from eq 3.1, the measurement has been
performed under polychromatic integration. The wavelength
dependence of the quantum yield of the photoreduction of
ferrioxalate and the relative photonic emission of the mercury
medium-pressure lamp have been integrated. For the photolysis

Evidence against HO• Radicals in Fenton Reactions
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5545
was pumped (10 ( 1 L min-1) by means of a MPN 80 pump
(Schmitt-Kreiselpumpen) under continuous jet-injection of
compressed air. The photolysis experiments were performed
during 120 min. The starting temperature was 25 ( 2 °C.
During the first 20 min, a linear increase of the reaction
temperature was observed. After this initial time period, the
temperature remained constant at 60 ( 3 °C.
The total volume of the photolysis solution was 2.50 L (2,4-
xylidine, 500 mg of C L-1 (500 ppm C), 5.41 × 10-3 M; H2O2,
5.41 × 10-2 M). The pH was adjusted to the initial value of
3.0 using H2SO4. The analysis was performed immediately after
taking the samples, which were filtered using Nylon Luer-Lock
membrane filters (Roth, 0.22 × 10-6 m). The filtered solution
was extracted with CHCl3 (10 mL of photolysis solution + 1
mL of CHCl3). Since numerous reaction products appeared in
the GC-MS/FTIR traces, only reaction products formed at
concentrations higher than 10-6 M were identified.
TABLE 2: Rates of Photons Absorbed (Pa) by the Iron(III)
Complexes in Solution (pH 3.0; Radiant Power of the TQ
718 Lamp, RP ) 155 ( 4.0 W)
Pa
a
photolyzed system (pH ) 3)
(einstein L-1 s-1
)
Fe3+aq, 2.40 × 10-2 M; oxalic acid, 2.40 × 10-2
M
M
M
8.23 × 10-5
3.32 × 10-5
3.30 × 10-5
3.39 × 10-5
3.26 × 10-5
Fe3+aq, 6.50 × 10-4 M; oxalic acid, 2.08 × 10-2
Fe3+aq, 6.50 × 10-4 M; 2,4-xylidine, 5.20 × 10-3
Fe3+aq, 6.50 × 10-4 M; H2O2, 5.41 × 10-2
Fe3+aq, 6.50 × 10-4 M; H2O, 55.51 M
M
a The standard deviation (from 3 experiments) did not exceed (5
relative %.
Thermal and Photochemically Enhanced Fenton Reac-
tions and Control Experiment. The Fenton oxidation experi-
ments, as well as the irradiation of 2,4-xylidine at pH 3.0 in
the absence of additional chemicals, were carried out in the pilot
reactor already described (Figure 2). The concentration of 2,4-
xylidine was 500 mg of C L-1, 5.41 × 10-3 M. During the
first 120 min of irradiation time, H2O2 was continuously added
(2.25 × 10-3 mol min-1, 0.2705 mol of H2O2 in 0.10 L in total).
In both Fenton reactions, FeSO4‚7H2O (6.50 × 10-4 M) was
dissolved in the presence of 2,4-xylidine. Again, the pH was
adjusted to the initial value of 3.0 by adding H2SO4. The
irradiation was continued for an additional time period of 60
min in order to complete the degradation reactions or to reach
a plateau region, respectively. A control experiment was carried
out by irradiating 2,4-xylidine at pH 3.0 in the absence of
additives.
Figure 2. Photochemical pilot reactor employed in all irradiation
experiments (reservoir, V ) 2.0 L; photochemical reactor, V ) 1.25
L; optical path length ) 2 cm; diameter ) 11 cm; cutoff of the quartz
sleeve, 220 nm).
Treatment of the Samples Prior to CHCl3 Extraction and
HPLC Injection. A 2.0 × 10-3 L volume of “reduction and
precipitation agent” (composed of 0.10 M Na3PO4, 0.10 M KI,
systems investigated here, the radiant power which is absorbed
by the photolyzed solution is distinctly lower, because a lower
concentration of Fe3+aq (6.50 × 10-4 M) was chosen, according
to the conditions employed in pilot scale reactors for photo-
chemically enhanced Fenton oxidation.5,6 Therefore, the rate
of photons absorbed (Pa) by the iron(III) complexes in various
solutions has been included in Table 2. The values Pa (in
Einstein L-1 s-1) have been calculated according to eq 3.2,
and 0.10 M Na2SO3) was added to a volume of 5.0 × 10-3
L
taken from the pilot reactor at different times. This procedure
led to a complete reduction of the residual H2O2 as well as to
the removal of most of the iron(II/III). The precipitate was
removed by filtration using the Nylon filters (Roth) already
described. After this procedure, the solutions were extracted
with CHCl3 and injected into the GC-MS/FTIR spectrometer.
Se,λ
RP
(1 - 10-A
)
(3.2)
λ
Results and Discussion
P )
P
)
a,λ
∑
∑
a
[
]
VNA
Eph,λ
λ
λ
Electrochemical Results. The electrochemical properties of
2,4-xylidine have been investigated by employing differential
pulse voltammography (DPP). 2,4-Xylidine, dissolved in water,
possesses two distinct oxidation waves between pH 6 and pH
3 (Figure 3). At the latter pH, which was chosen for all AOP
experiments described here, two distinct oxidation waves at
0.670 and 0.760 V (vs SHE) are discernible.
where Pa,λ is the rate of photons absorbed at wavelength λ
(wavelength range: 220-600 nm), NA the Avogadro number,
and V the total volume of the solution in L.
Since the emission spectrum of the mercury medium-pressure
lamp is well resolved, and the dependence of the quantum yield
of the ferrioxalate actinometer on the absorbed wavelength is
well-known,26 Pa can be calculated easily in this particular case.
DOC Analyzer. The analyses of the DOC (dissolved organic
carbon) were carried out using a Dohrmann DC-190 TOC (total
organic carbon) analyzer (T ) 680 °C) from Rosemount
Analytical. Calibrations were performed using 2,4-xylidine,
oxalic acid, and potassium hydrophthalate (KHP). The results
for all calibration samples could be fitted with linear calibration
curves.
This result can be rationalized by the equilibrium of the free
and the protonated aromatic amine.27 At pH < 3 only one
oxidation wave was obtained due to complete protonation of
2,4-xylidine. Note that both the hydroxyl radical (E ) 2.38 V
at pH 3.0) and the ferryl ion (Fe4+aq: E ≈ 1.65 V at pH 3.0)
possess a sufficiently high oxidation potential for the oxidation
of 2,4-xylidine (Table 1). Whereas Fe3+ (E ) 0.66 V at pH
aq
3.0) does not oxidize 2,4-xylidine in the dark, the electronically
excited cation Fe3+* (E ≈ 2.43 V at pH 3.0)28 is a strong
aq
H2O2 Photolysis. All photolysis experiments were carried
out in a pilot reactor shown in Figure 2. It consists of a reservoir
(V ) 2.0 L) and a flow-through annular photoreactor, equipped
with a TQ 718 medium-pressure mercury lamp. The solution
oxidant, and therefore, it is capable of oxidizing 2,4-xylidine
when coordinated in iron(III) complexes.
On the basis of the electrochemical literature,27 two reaction
pathways for oxidation/reduction sequences of aromatic amines

5546 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Bossmann et al.
Figure 4. Fe2+aq formation monitored by the increase of the absorbance
of the [Fe(phen)3]2+ complex [A(510 nm)] as a function of irradiation time
(lamp: TQ 718; RP ) 155 ( 4 W). The solutions have been purged
with N2 for 15 min prior to photolysis. Key: (1) [Fe3+aq] ) 6.50 ×
10-4 M, [oxalic acid] ) 2.08 × 10-2 M, pH 3.0, plateau value
corresponds to 100% reduction of iron(III) to iron(II); (2) [Fe3+aq] )
6.50 × 10-4 M, [2,4-xylidine] ) 5.41 × 10-3 M, pH 3.0, plateau value
corresponds to 63% reduction of iron(III) to iron(II); (3) [Fe3+aq] )
6.50 × 10-4 M, [H2O] ) 55.5 M, pH 3.0, plateau value corresponds to
32% reduction of iron(III) to iron(II).
Figure 3. Electrochemical characterization of the redox properties of
2,4-xylidine by differential pulse voltammography: 125 mg of C L-1
of 2,4-xylidine in 0.10 M KNO3 in aqueous solution, helium atmo-
sphere.
according to electron-transfer mechanisms can be proposed.
Whereas the oxidation of aromatic amines at the anode usually
leads to the formation of nitro compounds, the combination of
oxidation and reduction conditions can open pathways for the
formation of phenols. In Scheme 3, two reaction pathways for
the formation of phenols are proposed. In both cases, the initial
step consists of an electron-transfer reaction from 2,4-xylidine
to Fe4+aq and leads to the formation of Fe3+aq and an aromatic
cation radical [2,4-xylidine]+•. The exchange of the amine
group of this reactive intermediate versus water is possible. At
pH 3, the ammonium cation can leave the molecule and the
aromatic phenol cation radical [2,4-dimethylphenol]+• can be
formed. Finally, in a third step, 2,4-dimethylphenol can be
an electron-transfer oxidation of [2,4-dimethylxylidine]+•. In
this case, the aromatic amine group of 2,4-xylidine leaves as
hydroxylamine and again 2,4-dimethylphenol can be formed.
This reaction pathway can be labeled as an oxidation/aromatic
substitution/oxidation (Ox-SEar-Ox) mechanism.
Determination of the Quantum Yields of Several Model
Reactions for the Photochemical Reduction of Iron(III) to
Iron(II). As already discussed in the Introduction, the design
of efficient AOP reactors which rely on the photochemically
enhanced Fenton reaction depends strongly on the knowledge
of fundamental data, such as the quantum yields of the Fe3+
aq
created via an electron-transfer reaction. In principle, Fe2+
,
aq
reduction and the chemical nature of the key intermediate.
Therefore, we have compared (in the pilot reactor already
described (Figure 2)) the ferrioxalate photoreaction with the
reduction of iron(III) in the presence and in the absence of 2,4-
xylidine under polychromatic irradiation (Experimental Section
and Figure 4). In all cases, the pH of the photolyzed solutions
H2O2 (see Table 1), and various organic radicals may serve as
electron donors under the experimental conditions of the Fenton
reactions. This reaction branch can be labeled an oxidation/
aromatic substitution/reduction (Ox-SEar-Red)-mechanism.
The second step of the alternative reaction pathway consists of
SCHEME 3: Proposed Reaction Pathways for the Oxidation/Reduction of 2,4-Xylidine via Electron Transfer Reactionsa
a Two different reaction pathways are possible: (1) oxidation/aromatic substitution/reduction (Ox-SEAr-Red) mechanism; (2) oxidation/aromatic
substitution/oxidation (Ox-SEAr-Ox) mechanism.

Evidence against HO• Radicals in Fenton Reactions
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5547
TABLE 3: Initial Rate Constants (k) and Quantum Yields
of Iron(II) Formation under Polychromatic Irradiation
( Φ(Fe(II)) ) in Model Systems of the Photochemically
Enhanced Fenton Reaction (Corresponding Values for
Hydrogen Peroxide Photolysis Also Indicated) (Standard
Deviation: (3%)
SCHEME 4: Electron-Transfer Reaction between
Electronically Excited Iron(III) and 2,4-Xylidine and
Following Thermal Reaction Stepsa
photolyzed
system
c(Fe(III)) c(substrate) 105k(Fe(II))
Φ(Fe(II))
(lamp TQ 718)
(M)
(M)
(M s-1
)
Fe3+/H2C2O4
Fe3+/H2C2O4
Fe3+/H2O
2.40 × 10-2 2.40 × 10-2
6.50 × 10-4 2.08 × 10-2
6.50 × 10-4 55.5
9.38
6.48
1.08
4.92
1.14a
1.21
0.21
0.92
Fe3+/2,4-xylidine 6.50 × 10-4 5.20 × 10-3
photolyzed
system
c(Fe(III))
c(substrate) -105k(H2O2)
(M)
(M)
(M s-1) × 105
Φ(Fe(II))
≈0.33 (1.32b)
c
Fe3+/H2O2 6.50 × 10-4 2.40 × 10-2
7.07
photolyzed
system
c(Fe(III))
(M)
c(substrate)
-105k(H2O2)
(M)
(M s-1
0.06
)
Φ(HO•)
≈0.02
H2O2
6.50 × 10-3
a Calculated using eqs 3.2 and 3.3, taking into account the relative
emission spectrum of the medium-pressure mercury lamp (TQ 718;
see Figure 1). This value has been used as a reference polychromatic
quantum yield for the determination of the other quantum yields
presented in Table 3. b Quantum yield of H2O2 consumption. c The
second-order rate constant of the thermal H2O2 decomposition at 25
°C in the presence of Fe3+ was k ) 4.48 ((0.1) × 10-3 M-1 s-1
.
a Note that reaction 4.3 most likely proceeds within a hydrated
Iron(III)-2,4-xylidine complex.
aq
has been adjusted exactly to 3.0 by means of H2SO4. The
amount of iron(II) has been measured after formation of the
[Fe(phen)3]2+ complex as described in the Experimental Section
(Ferrioxalate Actinometry).
The “quantum yields” of iron(II) production under polychro-
matic irradiation <Φ(Fe2+)> by means of a medium-pressure
Hg lamp (TQ 718) have been calculated according to the
standard eq 3.3 used under monochromatic irradiation.26
(2) In contrast to case (1), the oxidation of 2,4-xylidine by
Fe3+* led to a plateau region (≈ 4.1 × 10-4 M Fe2+
)
aq
aq
corresponding to a steady-state concentration of approximately
63% (Figure 4). Therefore, it can be concluded that at least
one reaction pathway for the depletion of Fe2+ exists. The
aq
photoinduced redox reactions shown in Scheme 4 explain the
observed reaction behavior.
The formation of 2,4-dimethylphenol was observed by GC-
MS/FTIR. After 12 min of irradiation time, 6.0 ( 0.5 × 10-5
M of 2,4-dimethylphenol was detected. This experimental
finding confirms that an oxidation/reduction pathway from 2,4-
xylidine to 2,4-dimethylphenol exists.
dn
dt Pa
Φ(Fe2+) )
(3.3)
Pa ) rate of photons absorbed (einstein L-1 s-1)
dn/dt ) rate of Fe2+ formation (M s-1)
(3) During the oxidation of water by Fe3+*aq, a plateau region
(≈2.0 × 10-4 M Fe2+aq, corresponding to a steady-state
concentration of approximately 32%, Figure 4) was observed.
It is noteworthy that reaction 4.6 generates hydroxyl radicals.5c,23
As indicated in Table 3, the quantum yield determined for
the reduction of Fe3+aq by 2,4-xylidine is almost unity and hence
surprisingly high. From this result, we may deduce that an
organic compound, which is able to serve as a ligand in iron(III)
complexes and possesses a suitable oxidation potential, can
undergo an efficient inner-sphere electron transfer to electroni-
cally excited iron(III) leading to iron(II) formation. Further-
more, the quantum yield measured in this work for the
photochemical oxidation of water by Fe3+aq* is at the upper
limit of the data reported in the literature, which ranges from
0.065 5c to 0.24.23c
Fe3+aq + H2O
9
hν8 Fe2+aq + HO• + H+
(4.6)
However, its quantum yield ( Φ(Fe(II)) ) 0.21) is significantly
lower than that of the oxidation of oxalic acid ( Φ(Fe(II)) )
1.14) or 2,4-xylidine ( Φ(Fe(II)) ) 0.92) (see Table 3), and
therefore, under AOP conditions, the latter reactions will be
dominant as long as organic compounds are present in solution
and reaction 4.6 will be of minor importance.
Several mechanisms for the depletion of Fe2+ exist:3c (a)
The back-reaction of (4.7) proceeds as diffusion controlled. (b)
aq
In the three cases shown in Figure 4 (oxalic acid, 2,4-xylidine,
and H2O as ligands for iron(III) photooxidation), plateau regions
were reached at different iron(II) concentrations.
Fe2+aq + OH• + H+ f Fe3+aq + H2O
(4.7)
(1) The photolysis of [Fe(C2O4)3]3- led to a complete
The thermal Fenton reaction (reactions 1.2 and 1.3) also takes
place, leading to a depletion of both Fe2+aq and H2O2. (c) The
ferryl ion (Fe4+aq) generated in the thermal Fenton reaction 1.3
reduction of iron(III) to iron(II) (100%, 6.50 × 10-4 M Fe2+
)
aq
according to the established ferrioxalate reaction mechanism:26
is able to react with Fe2+ (reaction 4.8). A reaction rate
aq
hν
[Fe(C2O4)3]3- 98 [Fe(C2O4)2]2- + C2O4
(4.1)
•-
Fe4+aq + Fe2+aq f 2Fe3+
(4.8)
aq
hν
[Fe(C2O4)3]3- + C2O4•- 98 [Fe(C2O4)2]2- + C2O4
+
2-
constant for this process could not be found in the literature.
However, Walling already published in 1975 that the back-
2CO2 (4.2)

5548 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Bossmann et al.
Figure 5. H2O2 concentration as a function of irradiation time (lamp:
TQ 718) in the following systems: (1) [H2O2] ) 5.41 × 10-2 M, TQ
718 irradiation (0) (2) [Fe3+] ) 6.50 × 10-4 M, [H2O2] ) 5.41 ×
10-2 M, dark reaction (O) (3) [Fe3+] ) 6.50 × 10-4 M, [H2O2] ) 5.41
× 10-2 M, TQ 718 irradiation (2).
reaction of the thermal Fenton reaction proceeds diffusion
controlled.3c In the classic mechanistic interpretation, the
formation of the hydroxyl radical as intermediate was postulated.
However, if the ferryl ion and not the hydroxyl radical is the
key intermediate of the thermal Fenton reaction, the measured
rate constant remains unchanged!
(4) Since H2O2/HO2- is competing with water and coordinat-
ing organic molecules as a ligand in hydrated iron(III) com-
plexes, the quantum yield of the photooxidation of H2O2 in
aqueous solution in the presence of iron(III) is of importance
for the mechanistic understanding of the photochemically
enhanced Fenton reaction. The analysis of Fe2+aq formed during
irradiation in analogy to the classic ferrioxalate reaction (Vide
Figure 6. Comparison of the GC/MS traces of 2,4-xylidine and the
reaction intermediates formed after 10 min of H2O2 photolysis and 10
min of the thermal Fenton reaction. (The peak of oxalic acid is not
shown due to its interference with the CHCl3 peak (solvent for
extraction).)
supra) could not be performed, because Fe2+ quickly reacts
aq
with H2O2 (thermal Fenton reaction). We have therefore
followed the consumption of H2O2 during irradiation. As it can
be seen in Figure 5 (case (3)), a linear dependence of the H2O2
concentration has been obtained (photochemical reaction of
pseudo-zero-order). This is typically observed if a photoinitiated
inner-sphere electron-transfer reaction takes place. Thus, the
photoenhancement of the reduction of iron(III) by H2O2 may
be the result of the absorption of a photon by the hydrated
iron(III)-H2O2 complex [Fe(OH)(HO2)(H2O)4]+, followed by
an inner-sphere electron transfer leading to the formation of
the reduction of Fe3+(HO2
-
)
* by H2O2 of Φ(Fe(II)) ) 0.33.
aq
It also can be seen from Figure 5 that the photolysis of H2O2
(without addition of Fe3+aq) can be neglected and that the
thermal reduction of Fe3+aq to Fe2+aq by H2O2 (k ) (4.48 ( 1)
× 10-3 M-1 s-1) (Figure 5) is of minor importance.
2,4-Xylidine Oxidation Mechanisms Employing Hydrogen
Peroxide Photolysis: Thermal and the Photochemically
Enhanced Fenton Reactions. As already discussed in the
Introduction, our experimental strategy for investigating the
existence of the hydroxyl radical as key intermediate of both
Fenton reactions consisted in the comparison of the reaction
products of 2,4-xylidine generated by H2O2 photolysis with the
intermediates created in both Fenton reactions.
[Fe(OH)(H2O)5]+ (Fe2+aq) and the hydroperoxyl radical (HO2 )
•
(reaction 5.1). This process is faster than the reduction of
iron(III) to iron(II) in the ground-state iron(III)-H2O2 complex
(reaction 2.2).
[Fe(OH)(HO2)(H2O)4]+* + H2O f
It is well established that the UV photolysis of H2O2 leads to
[Fe(OH)(H2O)5]+ + HO2 (5.1)
•
the formation of free hydroxyl radicals.1,29
Furthermore, the quantum yield measured for the H2O2
disappearance is equal to 1.32. Therefore, the photoinduced
H2O2 decomposition (reaction 5.1) cannot be neglected in a
kinetic model of the photochemically enhanced Fenton reac-
tion: this photochemical process is always competing with the
H2O2 hν (UV)8 2HO•
(6.1)
Figure 6 and Table 4 show that entirely different intermediates
are formed during H2O2 photolysis on one hand and both Fenton
reactions on the other hand.
oxidation of organic compounds by Fe3+
*
(Scheme 4).
aq
Reaction 5.1 is followed by reaction 2.3 as in the thermal
Before the results of the photochemical experiments per-
formed in the pilot reactor are discussed, it should be noted
that the photolysis of 2,4-xylidine at pH 3.0 did not lead to any
detectable degradation products. To achieve a quantitative
understanding of the AOP reactions investigated, the DOC
(dissolved organic carbon) of the irradiated solutions has been
measured using a catalytic DOC analyzer, and the intermediates
have been identified and determined quantitatively by GC/MS.
process, and a second Fe3+aq is thermally reduced to Fe2+aq by
•
HO2 . Fe2+aq produced in reactions 5.1 and 2.3 participates in
the thermal Fenton reaction (reactions 1.2 and 1.3). Finally,
the reactive intermediate formed in the latter reaction (Fe4+
)
aq
oxidizes H2O2.3c As a consequence, a total of 4 molecules of
-
H2O2 are consumed for each photon absorbed by Fe3+(HO2
)
aq
([Fe(OH)(HO2)(H2O)4]+). It follows that a quantum yield for

Evidence against HO• Radicals in Fenton Reactions
J. Phys. Chem. A, Vol. 102, No. 28, 1998 5549
TABLE 4: Reaction Intermediates Identified by MS and
FTIR and Their GC Retention Times
compd
GC retention time (min)
H2O2 Photolysis
3-hydroxy-2,4-dimethylaniline
furfuryl alcohol
8.75
11.82
13.55
14.10
14.47
16.86
17.62
24.11
2(3H)-5-methylfuranone
2,4-xylidine
4-amino-2-hydroxy-3-methylbenzyl alcohol
acroleine
malondialdehyde
furane
Thermal and Photochemically Enhanced Fenton Reaction
oxalic acid
1.27
2.71
4.37
14.10
14.83
3,5-dimethyl-o-benzoquinone
2-hydroxy-5-methylbenzyl alcohol
2,4-xylidine
Figure 8. Thermal Fenton reaction of 2,4-xylidine (initial concentra-
tion: 5.41 × 10-3 M (500 mg of C L-1); initial FeSO4‚7H2O
concentration, 6.50 × 10-4 M; pH 3.0; H2O2 added continuously (0.2705
mol in 0.10 L during 120 min): 2,4-xylidine (2); 2,4-dimethylphenol
(O), oxalic acid (9); calculated DOC (4); measured DOC (b).
2,4-dimethylphenol
Figure 7. Photochemical oxidative degradation of 2,4-xylidine in the
presence of H2O2 (initial 2,4-xylidine concentration, 5.41 × 10-3
M
(500 mg of C L-1); pH 3.0; initial H2O2 concentration, 5.41 × 10-2
M): 2,4-xylidine concentration (2); measured DOC (b); calculated
DOC (9); minor intermediates (other symbols).
Figure 9. Photochemically enhanced Fenton reaction of 2,4-xylidine
(initial concentration, 5.41 × 10-3 M (500 mg of C L-1); initial FeSO4‚
7H2O concentration, 6.50 × 10-4 M; pH 3.0; H2O2 added continuously
(0.2705 mol in 0.10 L during 120 min): 2,4-xylidine (2); 2,4-
dimethylphenol (O); oxalic acid (9); calculated DOC (4); measured
DOC (b).
The concentrations of the most important intermediates and of
2,4-xylidine have then been added (DOCcalc) and compared to
the DOC measured experimentally.
In Figure 7, the results of the H2O2/UV treatment of a 2,4-
xylidine aqueous solution are shown: the concentration of 2,4-
xylidine decreases according to a pseudo-zero-order kinetic (k
) (1.18 ( 0.05) × 10-1 M s-1 ). However, the DOC does not
change during 120 min of irradiation within the limits of
experimental error [(8 relative %]. Some of the intermediates
are listed in Table 4. Figure 7 shows that a quantitative mass
balance of this reaction could not be measured.
The comparison of Figures 8 and 9 shows that the photo-
chemically enhanced Fenton reaction leads within an irradiation
time of 120 min to an almost complete removal of the DOC.
The complete decomposition of 2,4-xylidine was already
achieved after 60 min of irradiation time. Assuming a pseudo-
first-order kinetic, a reaction rate constant k1 ) 6.61 × 10-4
s-1 was calculated. After 35 min of irradiation, the decomposi-
tion rate of 2,4-xylidine increased and k2 ) 2.3 × 10-3 s-1 was
found. We attribute this increase of the 2,4-xylidine decom-
position rate to a decrease of inner filter effects (appearance of
a purple-brown color), which was strongest at the earliest stages
of the photolyses. The chemical nature of the main intermedi-
ates (>3 mg of C L-1) was exactly the same as that of the
thermal Fenton reaction, but their concentrations were lower.
In this experiment also, the measured and the calculated DOC
values were in very good agreement. The faster removal of
DOC, which at later reaction times consisted mainly of oxalic
acid, can be explained by the photodegradation of [Fe(C2O4)3]3-
(ferrioxalate reactions 4.1 and 4.2).
Figure 8 shows the results of 2,4-xylidine degradation by the
thermal Fenton reaction. In agreement with the analytical results
listed in Table 4, the chemistry of the thermal Fenton reaction
is very different from the chemistry of the H2O2 photolysis.
2,4-Dimethylphenol is the main intermediate of the thermal
Fenton degradation of 2,4-xylidine. This finding strongly
supports the mechanistic hypothesis described in Scheme 4.
Assuming a pseudo-first-order kinetic for the decomposition of
2,4-xylidine, a reaction rate constant k ) 6.59 × 10-4 s-1 was
obtained. Oxalic acid was found to be the main reaction
product, and the loss of DOC was only approximately 150 mg
of C L-1. Further oxidation of oxalic acid was not observed.
This finding can be explained by the high kinetic stability of
Conclusions
the iron(III) tris(oxalate) complex, [Fe(C2O4)3]3- 30
, preventing
the formation of hydrated iron(III)-H2O2 complexes and, thus,
recycling of iron(III) to iron(II). An excellent agreement
between calculated and measured DOC values was observed
for the complete irradiation time.
The products of 2,4-xylidine degradation during H2O2 pho-
tolysis and during the thermal and photochemically enhanced
Fenton reactions have been analyzed using GC-MS/FTIR.
Whereas general agreement exists in the literature that H2O2

5550 J. Phys. Chem. A, Vol. 102, No. 28, 1998
Bossmann et al.
S.; Mailhot, G.; Bolte, M. Solar Energy Mater. Solar Cells 1995, 38, 459-
474. (d) Safarzadeh-Amiri, A.; Bolton, J. R.; Cater, S. R. Solar Energy
1996, 56, 439-444. (e) Bandara, J.; Morrison, C.; Kiwi, J.; Pulgarin, C.;
Peringer, P. J. Photochem. Photobiol. A: Chem. 1996, 99, 57-66. (f)
Pignatello, J. J.; Chapa, G. EnViron. Toxicol. Chem. 1994, 13, 423-427.
(g) Pignatello, J. J. EnViron. Sci. Technol. 1992, 26, 944-951. (h) Sun, Y.;
Pignatello, J. J. Agric. Food Chem. 1993, 41, 1139-1142. (i) Safarzadeh-
Amiri, A.; Bolton, J. R.; Cater, S. R. J. AdV. Oxid. Technol. 1996, 1, 18-
26.
(6) (a) Oliveros, E.; Legrini, O.; Braun, A. M.; Hohl, M.; Mu¨ller, T.
Water Sci. Technol. 1997, 35, 223-230. (b) Oliveros, E.; Legrini, O.; Hohl,
M.; Mu¨ller, T.; Braun, A. M. Chem. Eng. Proc. 1997, 36, 397-405.
(7) (a) Haber, F.; Weiss, J. J. Proc. R. Soc. London, Ser. A 1934, 147,
332-345. (b) Merz, J. H.; Waters, W. A. Discuss. Faraday Soc. 1949, 2,
179-182.
TABLE 5: Summary of the Observed Reaction Behavior
reactive
reactive
electronic
intermediate of
intermediate of
reacn system confign of the thermal reacn the photochem reacn
(aq solutn)
iron(II/III)
(T ) 295 K)
(λirr ) 200-550 nm)
Fe2+
+
high-spin
(t2g)4(eg)2
high-spin
(t2g)5(eg)0
Fe4+
aq
aq
H2O2
Fe3+
HO•
aq
photolysis yields free hydroxyl radicals, a controversy about
the reactive intermediates of the thermal and the photochemi-
cally enhanced Fenton reactions still prevails. The comparison
of the reaction products of 2,4-xylidine clearly demonstrates
that H2O2 photolysis on one hand and both Fenton reactions on
the other hand involve different reactive intermediates. Whereas
hydroxylated aromatic amines are formed during H2O2 pho-
tolysis, 2,4-dimethylphenol is the most important intermediate
in both Fenton reactions. 2,4-Dimethylphenol can be formed
by an electron-transfer mechanism. Therefore, we conclude that
during the thermal reaction of Fe2+aq with H2O2 a cationic iron
intermediate possessing an unusual charge, most likely the ferryl
ion (Fe4+aq), is formed. However, the experiments described
(8) (a) Barb, W. G.; Baxendale, J. H.; George, P.; Hargrave, K. R.
Trans Faraday Soc. 1951, 47, 462-500. (b) Barb, W. G.; Baxendale, J.
H.; George, P.; Hargrave, K. R. Trans. Faraday Soc. 1951, 47, 591-616.
(9) (a) Goldstein, S.; Czapski, G.; Meyerstein, D. Free Radicals Biol.
Med. 1993, 15, 435-445. (b) Marsarwa, M.; Cohen, H.; Meyerstein, D.;
Hickman, D. L.; Bakac, A.; Espensen, J. H. J. Am. Chem. Soc. 1988, 110,
4293-4297.
(10) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry; Verlag
Chemie: Weinheim, Germany, 1980; pp 461-488.
(11) Wink, D. A.; Nims, R. W.; Saavedra, J. E.; Utermahlen, W. E.;
Ford, P. C. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 6604-6608.
(12) Rush, J. D.; Bielski, H. J. J. Am. Chem. Soc. 1986, 108, 523-525.
(13) Kremer, M. L.; Stein, G. Trans. Faraday Soc. 1959, 55, 595-560.
(14) Luzzatto, E.; Cohen, H.; Stockheim, C.; Wieghardt, K.; Meyerstein,
D. Free Radical Res. 1995, 23, 453-463.
here cannot distinguish between Fe4+ and a hydroxyl radical
aq
(15) Sychev, A. Y.; Isaak, V. G. Russ. Chem. ReV. 1995, 64, 1105-
1129.
(16) Sun, Y.; Pignatello, J. J. EnViron. Sci. Technol. 1993, 27, 304-
310.
(17) Wiberg, N.; Wiberg, E.; Hollemann, A. F. Lehrbuch der Anorga-
nischen Chemie; Walter de Gruyter: Berlin, New York, 1985; pp 1136-
1139.
(18) Koppenol, W. H.; Liebman, J. F. J. Phys. Chem. 1984, 88, 99-
101.
(19) (a) Sonntag, C. v.; Schuchmann, H.-P. Angew. Chem. 1991, 103,
1255-79. (b) Turro, N. J. Modern Molecular Photochemistry; University
Science Books: Mill Valley, CA, 1991.
(20) (a) Sawyer, D. T.; Kang, C.; Llobet, A.; Redman, C. J. Am. Chem.
Soc. 1993, 115, 5, 5817-5818. (b) Hage, J. P.; Llobet, A.; Sawyer, D. T.
Bioorg. Med. Chem. 1995, 3, 1383-1388.
(21) (a) Balzani, V. C.; Carassiti, V. Photochemistry of Coordination
Compounds; Academic Press: New York, 1970. (b) Hennig, H.; Rehorek,
D. Photochemische und Photokatalytische Reaktionen Von Koordinations-
Verbindungen; Akademie Verlag: Berlin, 1987.
(22) Walling, C.; Goosen, A. J. Am. Chem. Soc. 1973, 95, 2987-2991.
(23) (a) Langford, C. H.; Carey, J. H. Can. J. Chem. 1975, 53, 2430-
2435. (b) Faust, B. C.; Hoigne, J. J. Atm. EnViron. 1990, 24, 79-89. (c)
Zuo, Y.; Hoigne´, J. EnViron. Sci. Technol. 1992, 26, 1014-1022.
(24) Dobereiner, J. Schw. J. 1833, 62, 90-92.
(25) Jander, G.; Blasius, E. Einfu¨hrung in das anorganisch-chemische
Praktikum; Hirzel: Stuttgart, Germany, 1984.
(26) Braun, A. M.; Maurette, M.-T.; Oliveros, E. Photochemical
Technology; Wiley & Sons: New York, Chichester, U.K., 1991; pp 76-
80 and references therein.
complexed by Fe3+aq. The latter species would possess exactly
the same reactivity as Fe4+
.
aq
In contrast to the thermal Fenton reaction, evidence for the
formation of the hydroxyl radical during the photolysis of Fe3+
aq
at pH 3.0 exists. This reaction pathway, which shows a quantum
yield of 0.21, is suppressed in the presence of most organic
compounds dissolved in water, because the quantum yields for
the photooxidation of iron(III)-coordinated organic ligands are
usually significantly higher. In this work, quantum yields for
the oxidation of 2,4-xylidine (Φ ) 0.92) and H2O2 (Φ ) 1.33)
by electronically excited iron(III) have been measured for the
first time. The observed reaction behavior is summarized in
Table 5.
Acknowledgment. Financial support from Hewlett-Packard
is gratefully acknowledged. L.P. Jr. thanks the United Nations
for a grant supporting his Ph.D. thesis. A part of this research
has also been funded by BMBF.
References and Notes
(1) (a) Legrini, O.; Oliveros, E.; Braun, A. M. Chem. ReV. 1993, 93,
671-698 and references therein. (b) Ollis, D. F.; Al-Ekabi, H., Eds.
Photocatalytic Purification and Treatment of Water and Air; Elsevier:
Amsterdam, 1993.
(2) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513-886.
(3) (a) Fenton, H. J. H. J. Chem. Soc. 1894, 65, 899-901. (b) Fenton,
H. J. H. J. Chem. Soc., Trans. 1899, 75, 1-11. (c) Walling, C. Acc. Chem.
Res. 1975, 8, 125-131.
(27) (a) Christen, H. R. Grundlagen der Organischen Chemie; Otto Salle
Verlag: Frankfurt/Main, Germany, 1982. (b) Gontier, S.; Tuel, A. J. Catal.
1995, 157, 124-132. (c) Tonti, S.; Roffia, P.; Cesana, A.; Mantegazza,
M.; Padovan, M. Eur. Pat. 1988, 314, 147-155. (d) Sakaue, S.; Tsubakino,
T.; Nishiyama, Y.; Ishii, Y. J. Org. Chem. 1993, 58, 3633-3642. (e)
Werkert, E.; Angeli, E. Synth. Commun. 1988, 18, 1331-1339. (f) Tollari,
S.; Vergani, D.; Banfi, S.; Porta, F. J. Chem. Soc., Chem. Commun. 1993,
442-451. (g) Donten, M.; Hyk, W.; Ciszkowska, M.; Stojek, Z. Elec-
troanalysis (Weinheim) 1997, 9, 751-754. (h) Hoon, M.; Fawcett, R. W.
J. Phys. Chem. A 1997, 101, 3726-3730. (i) Spitsyn, M. A.; Andreev, V.
N.; Kazarinov, V. E. Russ. J. Electrochem. 1995, 31, 1092-1097. (k) Wang,
L.; Goodloe, G. W.; Stallman, B. J.; Cammarata, V. Chem. Mater. 1996,
8, 1175-1181.
(4) (a) Cahill, A. E.; Taube, H. J. Am. Chem. Soc. 1952, 74, 2312-
2316. (b) Eisenhauer, H. R. J. WPCF 1964, 36, 1116-1128. (c) Bishop,
D. F.; Stern, G.; Fleischman, M.; Marshall, L. S. Ind. Eng. Chem. Proc.
Design DeV. 1968, 7, 110-117. (d) Pasek, E. A.; Straub, D. K. Inorg. Chem.
1972, 11, 259-266. (e) Feuerstein, W.; Gilbert, E.; Eberle, S. H. Vom
Wasser 1981, 56, 35-54. (f) Barbeni, M.; Minero, C.; Pelizzetti, E.;
Bogarello, E.; Serpone, N. Chemosphere 1987, 6, 2225-2237. (g) Sedlak,
D. L.; Andren, A. W. EnViron. Sci. Technol. 1991, 25, 777-782. (h)
Lipczynska-Kochany, E. Chemosphere 1991, 22, 529-536.
(28) The value of E ≈ 2.43 V was estimated using Fe3+aq (E ≈ 0.66 V
at pH 3.0) and E0-0 ) 1.77 V, corresponding to λ0-0 (Fe3+*aq) ) 700 nm
at pH 3.0.
(29) Maillard, C.; Guillard, C.; Pichat, P. Chemosphere 1992, 24, 1085-
1094 and references therein.
(5) (a) Goldstein, S.; Czapski, G.; Meyerstein, D. Free Radicals Biol.
Med. 1993, 15, 435-445. (b) Ruppert, G.; Bauer, R.; Heisler, G. J.
Photochem. Photobiol. A: Chem. 1993, 73, 75-78. (c) Andrianirinaharivelo,
(30) Cotton, F. A. Coord. Chem. ReV. 1972, 8, 185-227.