Geochimica et Cosmochimica Acta, Vol. 69, No. 1, pp. 83-93, 2005
Copyright © 2005 Elsevier Ltd
Printed in the USA. All rights reserved
0016-7037/05 $30.00 ⫹ .00
doi:10.1016/j.gca.2004.05.043
Oxidation of iron (II) nanomolar with H2O2 in seawater
MELCHOR GONZÁLEZ-DAVILA, J. MAGDALENA SANTANA-CASIANO, and FRANK J. MILLERO*
Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA
(Received January 23, 2004; accepted in revised form May 20, 2004)
Abstract—The oxidation of Fe(II) with H2O2 at nanomolar levels in seawater have been studied using an
UV-Vis spectrophotometric system equipped with a long liquid waveguide capillary flow cell. The effect of
pH (6.5 to 8.2), H2O2 (7.2 ⫻ 10⫺8 M to 5.2 ⫻ 10⫺7 M), HCO⫺
3 (2.05 mM to 4.05 mM) and Fe(II) (5 nM
to 500 nM) as a function of temperature (3 to 35 °C) on the oxidation of Fe(II) are presented. The oxidation
rate is linearly related to the pH with a slope of 0.89 ⫾ 0.01 independent of the concentration of HCO⫺
3 . A
kinetic model for the reaction has been developed to consider the interactions of Fe(II) with the major ions
in seawater. The model has been used to examine the effect of pH, concentrations of Fe(II), H2O2 and HCO⫺
3
as a function of temperature. FeOH⫹ is the most important contributing species to the overall rate of oxidation
from pH 6 to pH 8. At a pH higher than 8, the Fe(OH)2 and Fe(CO3)2⫺
species contribute over 20% to the
2
rates. Model results show that when the concentration of O2 is two orders of magnitude higher than the
concentration of H2O2, the oxidation with O2 also needs to be considered. The rate constants for the five most
kinetically active species (Fe2⫹, FeOH⫹, Fe(OH)2, FeCO3, Fe(CO3)2⫺
2 ) in seawater as a function of
temperature have been determined. The kinetic model is also valid in pure water with different concentrations
of HCO⫺
Copyright © 2005 Elsevier Ltd
3 and the conditions found in fresh waters.
Farlow, 2000), these measurements were made in dilute solutions and with different buffers to extract oxidation rate constants for the Fe(II)-hydrolysis and carbonate species. This
model has been applied to describe the behavior of Fe(II) in
natural fresh water and with solutions at a high ionic strength
(King and Farlow, 2000; Pullin and Cabaniss, 2003).
The reaction mechanism for the oxidation of Fe2⫹ with
H2O2 has not been resolved satisfactorily. Most researchers
have accepted the free radical chain mechanism, initially proposed by Haber and Weiss (1934) and modified by Barb et al.
(1951) working under very acidic conditions. The mechanism
of the reaction between Fe(II) and H2O2 has been widely
assumed to be the following
1. INTRODUCTION
The oxidation of Fe(II) with H2O2 has been studied in
seawater by a number of authors. H2O2 is an intermediate in the
reduction of oxygen to water and can act as an oxidant in the
reaction with Fe(II) (Moffett and Zika, 1987). The H2O2 in
surface water is also generated by photochemical processes,
due to the presence of organic compounds (Moffet and Zika,
1983). The H2O2 is present at a concentration of ⬃10⫺7 M in
surface seawater (Zika et al., 1985a, 1985b; Moore et al.,
1993). Thus, hydrogen peroxide in seawater is in excess with
respect to the concentration of Fe(II) and could be an important
oxidation pathway for the Fe(II) oxidation.
Moffett and Zika (1987) and Millero and Sotolongo (1989)
using the bathophenantroline method (Sung and Morgan, 1980)
have studied this reaction at micromolar levels of Fe(II). Millero et al. (1991) studied the effect of ionic interactions on the
rates of oxidation of Fe(II) with hydrogen peroxide at micromolar levels in different media. In recent years, the chemiluminiscence technique using the luminol reagent has been used
to study the oxidation of Fe(II) at nanomolar levels in different
media (King et al., 1995; Millero et al., 1995a; King and
Farlow, 2000). The utilization of a long liquid waveguide
capillary flow cell (LWCFC) permits the spectrophotometric
determination of Fe(II) at nanomolar levels (Zhang et al.,
2001). The advantage of this technique, apart from the low
level of detection that could also be obtained with the chemiluminiscense technique, is that one can to follow the spectrum
of the complex between Fe(II) and ferrozine or bathophenantroline indicators.
Although some studies have been carried out in recent years
on the oxidation of Fe(II) with O2 and H2O2 at nanomolar
levels of Fe(II) to define a kinetic model (King, 1998; King and
Fe2⫹ ⫹ H2O2 → Fe3⫹ ⫹ OH● ⫹ OH⫺
(1)
Fe2⫹ ⫹ OH● → Fe3⫹ ⫹ OH⫺
(2)
OH● ⫹ H2O2 → HO2● ⫹ H2O
(3)
Fe2⫹ ⫹ HO2● → Fe3⫹ ⫹ HO2⫺
(4)
Fe3⫹ ⫹ O2●⫺ → Fe2⫹ ⫹ O2
(5)
In equation (5), the equilibrium between the two forms of
superoxide has been considered and O●⫺
is used instead of
2
HO●⫺
(pK ⫽ 4.8, Barb et al., 1951, Dunford, 2002). This
2
scheme is a chain mechanism in which Fe2⫹ is regenerated.
However, other authors consider a non-free radical mechanism.
Kremer (1999) published a reinvestigation of the reaction of
ferrous iron with H2O2 at low pH considering the formation of
an intermediate oxidant, the ferryl ion, FeO2⫹. This species
was first proposed by Bray and Gorin (1932) as one step in the
ferric ion catalyzed decomposition of H2O2.
* Author to whom correspondence should be addressed (fmillero@
rsmas.miami.edu).
83
Fe2⫹ ⫹ H2O2 → FeO2⫹ ⫹ H2O
(6)
FeO2⫹ ⫹ H2O2 → Fe2⫹ ⫹ H2O ⫹ O2
(7)
84
M. González-Davila, J. M. Santana-Casiano and F. J. Millero
Reaction (7) takes places when H2O2 is in excess. When the
ferrous ion is in excess equation (6) will be followed by the
equation
2H⫹
FeO2⫹ ⫹ Fe2⫹ → 2Fe3⫹ ⫹ H2O
(8)
This mechanism shows both ferrous and ferryl ions attacked by
hydrogen peroxide, but not ferric ions. According to Dunford
(2002) there is no obvious kinetic way to distinguish the two
mechanisms and the conflict cannot be settled at the present
time. In seawater this distinction is probably not important as
both species should react rapidly with Br⫺ and HCO⫺
3 , even●
tually forming Br⫺
2 and HCO3 , which will be the principal
oxidants in subsequent reactions (Moffet and Zika, 1987;
Emmenegger et al., 1998).
The objective of this work is to examine the oxidation of
Fe(II) in seawater with hydrogen peroxide over a range of
concentrations similar to that found in the ocean. The effect of
pH, concentration of HCO⫺
3 and Fe(II) as a function of temperature on the oxidation rate of Fe(II) has been considered. A
kinetic model that describes the behavior of Fe(II) under different experimental conditions has been developed to explain
the results. The contribution of each Fe(II) species to the
overall rates of Fe(II) oxidation has also been determined.
2. MATERIAL AND METHODS
2.1. Chemicals
Fe(II) stock solutions (2 ⫻ 10⫺3 M) were prepared using ferrous
ammonium sulfate hexahydrate (Fisher), acidified with Suprapur HCl.
The initial concentration of Fe(II) was kept at 250 nM in the reaction
vessel in most of the studies. All chemicals used for the Fe(II) determination were trace analysis grade.
2.2. Oxidation Experiments
The reactions were studied in a 250 mL glass thermostated vessel.
The temperature was controlled to ⫾ 0.02°C with a NesLab circulating
bath in the range of 3 to 35 °C. The top of the vessel had four openings,
one for a glass frit to bubble N2 through the solutions, two for both the
glass and the reference electrode, and one to insert a calibrated pipette,
from which the samples were taken. The solutions were stirred with a
teflon-coated magnetic stirrer. The samples were taken from the vessel
with a 10 mL calibrated automatic pipette and added to the 25 mL glass
flasks where the ferrozine reagents had been previously added. Considering the 2.5 dilution factor for the Fe(II) determination, the initial
concentration in the 200 mL of seawater solution was fixed at 250 nM.
After bubbling the solution with N2 for 1 h, the pH was adjusted to the
desired value with additions of small amounts of HCl and the required
amount of H2O2 was added to the seawater. The addition of the Fe(II)
stock solution (25 ␮L of 2 mM Fe(II) in HCl 0.01M) to the seawater
corresponds to the zero time of reaction. The pH for the study was
recorded during the reaction to account for any change after the
addition of the Fe(II). The change in pH was always less than 0.02 U,
with the highest effects occurring at low pH where the buffer capacity
of the carbonate is lowest. In all cases, the gas stream was passed
through a MnO⫺
4 solution to eliminate any H2O2 and through a trap
with MilliQ 18 M⍀ pure water. The N2 was continuously bubbled
through the solutions during the experiments.
Fe(II) oxidation experiments at different bicarbonate concentrations
were carried out under the same conditions after increasing the concentration of NaHCO3 from the initial value (2.05 mM) to the desired
level.
To determine if Fe(III) is reduced by H2O2 under our experimental
conditions, control experiments were carried out with Fe(III), 2.5 ⫻
10⫺7 M, in seawater in the pH range from 7 to 8.2 in the presence of
1 ⫻ 10⫺7 M H2O2.
2.3. pH Measurements
Tris-(hydroxymethyl)aminomethane (Tris)-artificial seawater buffers (Millero, 1986; Millero et al., 1987) were used to calibrate the
electrode system used to determine the pH of the solutions. These
buffers were prepared with a concentration of both TRIS and TRISHCl of 0.005 m. The pH was measured on the free scale with an Orion
pH meter using an Orion glass electrode and an Orion Ag/AgCl
reference electrode. The outer sleeve of the reference electrode was
filled with 0.7 m NaCl. The effect of temperature on the pK* of the
Tris-buffers was considered in each study (Millero, 1986).
2.4. Fe(II) Analysis
The Fe(II) concentration was determined spectrophotometrically using a modified version of the ferrozine method (Gibbs, 1976). In this
method, 10 mL samples react with ferrozine (50 ␮L, 0.01 M) in an
acetate buffer solution (2 mL, pH ⫽ 5.5) to form a pink Fe(II)-ferrozine
complex that absorbs at 562 nm. The Ferrozine solution was prepared
by dissolving 0.51 g of ferrozine (C20H13N4NaO6S2 · 2H2O) in 100 mL
of water. The buffer was made up with a ratio of 1:8 using 6.9 M HCl
and 5 M HAc/ammonia solutions. The 5 M solution was prepared by
mixing 338 mL ammonium hydroxide and 286 mL acetic acid and
diluted to 1 L with MilliQ-18⍀ pure water. Different authors have
reported some interferences in this method due to the presence of
Fe(III). Murray and Gill (1978) found that the ferrozine added to an
Fe(III) solution showed a slow increase of color with the time that may
be due to reduction of Fe(III) to Fe(II) by ferrozine. Hong and Kester
(1986) estimated that the reduction of Fe(III) was of order of 10% after
10 mins. and could be ⬃25% for longer periods of time. Fe(III) can be
masked by the addition of strong ligands such us F⫺, NTA and EDTA
(Viollier et al., 2000). NTA and EDTA have been shown to affect the
kinetic of Fe(II) oxidation (Santana-Casiano et al., 2000). F⫺ forms a
FeF2⫹ ion pair that is nonreactive to reduction (Millero et al., 1995a).
Considering that in our studies the Fe(III) is present due to both, the
Fe(II) oxidation and natural Fe(III), NaF was used at a final concentration 1.25 ␮M to complex any soluble Fe(III) and to avoid any
interference. When the NaF was used with the acetate buffer and
ferrozine solution, a stable absorbance reading was observed for over
30 mins.
A 5 m long waveguide capillary flow cell (LWCFC) from Ocean
Optics was used to carried out measurements at nanomolar levels of
Fe(II) concentration. The LWCFC was connected to two fiber optics
connectors for light path, one coming from a tungsten lamp and the
another to the UV detector S2000 (Ocean Optics) and the spectra
recorded using the OOIBase32 computer program provided by Ocean
Optics. The sample was pumped using a Rabbit peristaltic pump
through the two fluid connectors on the front panel of the LWCFC case.
During the absorbance reading, the circulating pump was turned off to
obtain a stable reading. The utilization of this long flow cell significantly enhances the sensitivity of spectrophotometric analysis of iron
by the ferrozine method (Zhang et al., 2001). The response of the
system was linear over two orders of magnitude Fe(II) ([Fe(II)]
⫽ 0.408 ⫹ 82.92 Abs), with a standard error for the Fe(II) determination of ␴ ⫽ 1.0 nM in [Fe(II)].
2.5. H2O2 Determination
Hydrogen peroxide was determined using an enzyme-mediated fluorescent decay method (Zika and Saltzman, 1982) utilizing horseradish
peroxidase and scopoletin. Batch hydrogen peroxide determinations
were performed with a Turner Designs model 10 fluorometer equipped
with a stirred large volume cuvette adapter, using 25 ⫻ 150 mm
borosilicate culture tubes (Moore et al., 1993).
2.6. Oxygen Determination
Dissolved oxygen concentration in seawater was determined using a
modified Winkler method described by Hansen (1999). This was done
to check the initial dissolved oxygen concentration after bubbling the
seawater solution with nitrogen.
Oxidation of Fe(II) by H2O2
Fig. 1. Values of the log k1 vs. log [H2O2] for the oxidation of Fe(II)
with H2O2 at 25°C and pHF ⫽ 8.17 and S ⫽ 36.244. The pseudofirst
order rate constants show a linear relationship both with [H2O2] in
excess (solid line) and in the full range studied (dashed line).
2.7. Numerical Model
A kinetic model has been used to explain the oxidation kinetics of
Fe(II) with H2O2 in seawater. The Gepasi Version 3.21 (Mendes, 1997)
software system was used to simulate the chemical kinetics and to
compute the time-dependent concentrations of all the reactants. The
individual rate constants ki are obtained by adjusting the observed
[Fe(II)] concentration/time pair of data for the different experimental
conditions with the kinetic model output. A response surface methodology (Box and Draper, 1987) together with initial conditions and ki
values were used in the Gepasi program to generate theoretical [Fe(II)]
concentrations as a function of time. The Statistica Program for Windows (1995) was used in the minimization procedure. Initial conditions
and concentration and time data were inserted in the Statistica Program
until the sum of squared residuals from the difference between model
and data were minimized for the entire set of experiments (SantanaCasiano et al., 2004). In the Gepasi software, chemical equilibrium is
treated as a series of forward and backward reactions with bimolecular
rate constants of 1010 M⫺1s⫺1 for the acid bases and complex formation backward reactions following Buerge and Hug (1998).
3. RESULTS
85
The reduction of Fe(III) has also been proposed to occur in
the presence of H2O2 (Pignatello, 1992) generating Fe(II).
Control experiments carried out with Fe(III), 2.5 ⫻ 10⫺7 M, in
seawater in the pH range from 7 to 8.2 in the presence of
1 ⫻ 10⫺7 M H2O2 did not show any formation of Fe(II),
possibly due to the formation of kinetically inert ferric hydroxy
colloids (Moffet and Zika, 1987). Consequently, a Fe(II) regeneration process was not considered in our studies and in the
kinetic model.
To examine the order of the reaction with respect to H2O2,
we have made a number of measurements with different initial
concentrations of H2O2 at pH 8.17 (S ⫽ 36.244). Figure 1
shows the resulting pseudofirst-order rate constant for the oxidation of Fe(II) with different concentrations of H2O2. When
the concentration of H2O2 was over the stoichiometric ratio 2:1
([H2O2]o ⬎ 1.25 ⫻ 10⫺7 M), the resulting slope is 1.00 ⫾ 0.01.
This result indicates that the reaction is first-order with respect
to [H2O2] and in agreement with the past micromolar studies
(Millero and Sotolongo, 1989). At low concentrations of H2O2
([H2O2]o ⬍ 1.25 ⫻ 10⫺7 M), deviations from a linear dependence were observed and assuming a pseudofirst order rate is
not valid. Changes in the contributions of the major species at
pH over 8 and the presence of trace amounts of oxygen may
control the rates at low [H2O2]. The slope for the linear relationship with [H2O2] for all the data shown in Figure 1, at times
lower than t1/2, yield a slightly higher value of 1.26 ⫾ 0.03.
As found by other authors (Millero and Sotolongo, 1989),
our results show the oxidation of Fe(II) is strongly dependent
on pH. Our results are shown in Figure 2 over the pH range of
6.5 to 8.2, with 275 nM hydrogen peroxide at 25°C (Fe(II):
H2O2 ratio 1:1) assuming pseudofirst order kinetics. At pH
values lower than 7.5, a pseudofirst order rate for the Fe(II)
oxidation was clearly observed at times longer than t1/2. At pH
over 7.5 as the case at low [H2O2], an increased role of the
oxygen and changes in the Fe(II) speciation can account for
a higher order of the reaction. The values of log k (k
⫽ k1[H2O2]⫺1) assuming a pseudofirst order rate at times lower
than t1/2 over the entire pH range, have been fitted to the linear
equation
The oxidation of Fe(II) in seawater in the presence of H2O2
can be described by
Fe(II) ⫹ H2O2 → Products
(9)
where Fe(II) represents all the iron (II) species present in the
solution. At micromolar levels, this reaction has been found to
be first order with respect to total Fe(II) and H2O2 concentration (Millero and Sotolongo, 1989; Millero et al., 1991)
d[Fe(II)]
dt
⫽ ⫺k[Fe(II)][H2O2]
(10)
The brackets denote the total molar concentration. When the
reactions are studied with an excess of [H2O2], the reaction
becomes pseudofirst-order (Moffet and Zika, 1987; Millero and
Sotolongo, 1989)
d[Fe(II)]
dt
where k1 ⫽ k · [H2O2]
⫽ ⫺k1[Fe(II)]
(11)
Fig. 2. Effect of pH on the values of log k (M⫺1 s⫺1) for the
oxidation of Fe(II) with [H2O2] in seawater (S ⫽ 36.244, total dissolved inorganic carbon 2.05 mM) and with NaHCO3 added at 25°C.
86
M. González-Davila, J. M. Santana-Casiano and F. J. Millero
Table 1. Fe(II)-H2O2 oxidation rate constants (M⫺1 s⫺1) in seawater under different
experimental conditions.
25°C
pHF in seawater*
8.17
8.05
7.94
7.75
7.71
7.36
7.06
6.99
6.89
6.70
6.54
7.94
7.94
6.99
6.70
(2.9 mM HCO⫺
3)
(4.05 mM HCO⫺
3)
(4.05 mM HCO⫺
3)
(4.05 mM HCO⫺
3)
10°C
Log k
(M⫺1 s⫺1)
4.77
4.64
4.54
4.37
4.36
4.06
3.82
3.73
3.60
3.45
3.30
4.64
4.73
3.83
3.53
pHF in seawater*
8.17
8.00
7.55
7.29
7.01
6.53
8.17
8.00
7.01
6.53
8.17
8.17
8.17
(4.05 mMHCO⫺
3)
(4.05 mMHCO⫺
3)
(4.05 mMHCO⫺
3)
(4.05 mMHCO⫺
3)
(20°C)
(3°C)
(35°C)
Log k
(M⫺1 s⫺1)
4.17
4.04
3.53
3.35
2.93
2.64
4.40
4.24
3.26
2.78
4.52
3.84
5.00
* [HCO3] ⫽ 2.05 mM
log k ⫽ ⫺2.55(⫾0.09) ⫹ 0.89(⫾0.01)pH
(12)
Over this pH range, the dependence with pH gives a similar
slope to the values determined at micromolar levels by Moffet
and Zika (1987) and Millero and Sotolongo (1989).
Previous studies (Millero et al., 1991; King and Farlow,
2⫺
2000) have shown that the addition of HCO⫺
3 (or CO3 ) causes
the rates of oxidation of Fe(II) with H2O2 to increase. Total
dissolved inorganic carbon concentration in the seawater used
in this study was 2.05 mM. To explain the effect of carbonate
on the oxidation of Fe(II) in seawater, the concentration of
HCO⫺
3 was increased to 2.95 mM and 4.05 mM over the pH
range 6.5 to 8.2. Figure 2 shows the effect of HCO⫺
3 on the
oxidation rates of Fe(II) giving a slope of 0.96 ⫾ 0.01 when the
total carbonate concentration was 4.05 mM. At pH ⫽ 7.94, the
oxidation rate increases from 0.046 min⫺1 in natural seawater
to 0.056 min⫺1 when the concentration of HCO⫺
3 was 4.05
mM. As we will show later, changes in the Fe(II) speciation
and formation of the reactive species FeCO3 and Fe(CO3)2⫺
2
are responsible for the observed behavior.
In our next series of experiments, we determined the effect of
temperature on the rates of oxidation of Fe(II) with H2O2 in
seawater. The experiments were made at the natural pH of
seawater 8.17 and a Fe(II):H2O2 ratio of 1:1 (Table 1). A
decrease in the oxidation rate was observed when the temperature decreases from 35 to 3 °C. The values of k over the entire
temperature range have been fitted to the equation.
ln k ⫽ 35.4(⫾1.3) ⫺ 7279(⫾383) 1 ⁄ T
(13)
with a standard error of 0.05. This gives an energy of activation
of 60 ⫾ 3 kJ mol⫺1 close to the value determined by Millero
and Sotolongo (1989) of 56 ⫾ 2 kJ mol⫺1 at micromolar level
and a pH ⫽ 6.
Similar studies were carried out at 10°C to study the effect of
pH and HCO⫺
3 on the Fe(II) oxidation rate (Table 1). Linear
relationships were obtained for the pH dependence in seawater
with different levels of HCO⫺
3 at 10°C.
log k ⫽ ⫺3.74(⫾0.39) ⫹ 0.97(⫾0.05)pH
< HCO3⫺ = ⫽ 2.05 mM (14)
log k ⫽ ⫺3.64(⫾0.02) ⫹ 0.98(⫾0.01)pH
< HCO3⫺ = ⫽ 4.05 mM (15)
These results show that the Fe(II) oxidation by hydrogen peroxide is first-order with respect to the proton concentration and
a two-fold increases in the total HCO⫺
3 concentration increases
the log k by 0.1 U over the pH range studied.
The effect of ionic strength [I ⫽ 0.0199 · S/(1–10⫺3 · S),
(where S is the salinity)] on the oxidation rates (k, M⫺1s⫺1)
was evaluated by Millero and Sotolongo (1989) at micromolar
levels of Fe(II) and can be represented at 25°C and pH ⫽ 6.0
by
log k ⫽ 3.84 ⫺ 1.70I1⁄2 ⫹ 1.20I
(16)
King and Farlow (2000) studied the Fe(II) oxidation at nanomolar levels with H2O2 for pure water with 2.0 mM NaHCO3
at 25°C and obtained a value of log k (M⫺1s⫺1) of 4.26 at pH
⫽ 6.98. In seawater, with a similar concentration of NaHCO3
and pH, we have obtained a value of log k ⫽ 3.72 (Table 1). If
we assume that Eqn. 16 is valid at nM levels of Fe(II) and at a
fixed pH, we obtain log k ⫽ 4.29 in pure water (2 mM
NaHCO3) at pH 6.98, similar to the experimental value of King
and Farlow (2000). This indicates that the effect of ionic
strength on the Fe(II) oxidation rate by H2O2 at nM is similar
to the values at ␮M. At pH ⫽ 8.17 and taking into account that
in seawater with a salinity of 36.244 (I ⫽ 0.748), log k ⫽ 4.77
(Table 1), Eqn. 16 be expressed as
log k ⫽ 5.34 ⫺ 1.70I1⁄2 ⫹ 1.20I
(17)
According to Eqn. 17, in pure water with 2.0 mM NaHCO3 and
pH ⫽ 8.17, log k ⫽ 5.34.
Oxidation of Fe(II) by H2O2
87
Table 2. Stability constants for the formation of Fe(II) and Fe(III) inorganic complexes
considered for the kinetic model.
N°
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Species
⫹
⫺
H2O N H ⫹ OH
⫹
CO2 ⫹ H2O N HCO⫺
3 ⫹ H
2⫺
HCO⫺
⫹ H⫹
3 N CO3
Na⫹ ⫹ HCO⫺
3 N NaHCO3
Na⫹ ⫹ CO2⫺
N NaCO⫺
3
3
⫹
Ca2⫹ ⫹ HCO⫺
3 N CaHCO3
2⫹
2⫺
Ca ⫹ CO3 N CaCO3
⫹
Mg2⫹ ⫹ HCO⫺
3 N MgHCO3
Mg2⫹ ⫹ CO2⫺
N
MgCO
3
3
2 Mg2⫹ ⫹ CO2⫺
N Mg2(CO3)2⫹
3
2⫹
⫺
⫹
Mg ⫹ OH N MgOH
⫹
Fe2⫹ ⫹ HCO⫺
3 N FeHCO3
Fe2⫹ ⫹ CO2⫺
N
FeCO
3
3
Fe2⫹ ⫹ 2 CO2⫺
N Fe(CO3)2
3
2⫹
2⫺
⫺
Fe ⫹ CO3 ⫹ OH N Fe(CO3)(OH)⫺
Fe2⫹ ⫹ H2O N Fe(OH)⫹ ⫹ H⫹
Fe2⫹ ⫹ 2 H2O N Fe(OH)2 ⫹ 2 H⫹
Fe2⫹ ⫹ Cl⫺ N FeCl⫹
Fe2⫹ ⫹ SO2⫺
N FeSO4
4
H⫹ ⫹ SO2⫺
N HSO⫺
4
4
3⫹
⫺
Fe ⫹ Cl N FeCl2⫹
Fe3⫹ ⫹ 2 Cl⫺ N FeCl⫹
2
Fe3⫹ ⫹ H2O N Fe(OH)2⫹ ⫹ H⫹
3⫹
⫹
Fe ⫹ 2 H2O N Fe(OH)⫹
2 ⫹ 2 H
Fe3⫹ ⫹ 3 H2O N Fe(OH)3 ⫹ 3 H⫹
⫹
Fe3⫹ ⫹ 4 H2O N Fe(OH)⫺
4 ⫹ 4 H
Log K
(0.7 mol L⫺1, 25°C)
Ref
⫺13.69
⫺6.005
⫺9.6
⫺0.53
0.42
0.33
2.1
0.28
1.94
2.59
1.70
0.97
4.33
6.09
8.90
⫺9.66
⫺20.87
⫺0.12
0.96
⫺0.10
0.57
0.13
⫺2.62
⫺6.0
⫺12.5
⫺21.8
1
1
1
2
2
2
2
2
2
2
2
3
4
4
4
5
5
4
4
1
5
5
5
5
5
5
1. Millero, 1995. 2. Millero and Schreiber, 1982. 3. Millero and Hawke, 1992. 4. King,
1998. 5. Millero et al., 1995b.
4. DISCUSSION
The oxidation of Fe(II) in natural waters has been proposed
to occur through the mechanism previously described but considered all the iron species present in the media (Moffet and
Zika, 1987; Millero and Sotolongo, 1989)
Fe(II) ⫹ H2O2 → Fe(III) ⫹ OH● ⫹ OH⫺
(18)
Fe(II) ⫹ OH● → Fe(III) ⫹ OH⫺
(19)
Under the conditions studied in this work, at relatively low
Fe(II) : [H2O2] ratios the oxidation of Fe(II) is second order
overall in the reactants as described by
d[Fe(II)]
dt
⫽ ⫺2k[Fe(II)][H2O2]
(20)
The values of k (Millero and Sotolongo, 1989) for the overall
Fe(II) oxidation rate is a complex function of pH and can be
explained in terms of the weighted sum of the oxidation rate of
individual Fe(II) species presented in the solution. The factor of
2 in Eqn. 20 accounts for the fast reaction with the hydroxyl
radical (Eqn. 19) with Fe(II). The intermediate OH● radical
reacts rapidly and unselectively with many natural species
(Hoigné, 1988). The oxidation rate of Fe(II) with OH● should
be dependent on [Fe(II)], if other substances are competing for
the same oxidant. In seawater solutions (pH ⫽ 7.2) and with
[Fe(II)] from 100 nM to 1000 nM, the pseudofirst-order rate
constant, k1, does not change appreciably, from 0.11 ⫾ 0.01
min⫺1 to 0.13 ⫾ 0.02 min⫺1 when the [H2O2] added was kept
at 4.24 ␮M. This indicates that OH● or any resulting reactive
●
species formed in seawater (Br⫺
2 , HCO3 , organic radicals) must
react predominantly with Fe(II) as was found by King and
Farlow (2000) in pure water. Emmenegger et al. (1998) in lake
waters with high dissolved organic carbon ([DOC] ⫽ 3.2 mg
L⫺1) showed the OH● was scavenged by dissolved organic
matter, HCO⫺
3 and Fe(II) according to Larson and Zepp (1988).
The formation of HCO●3 was particularly favorable under such
conditions and the main reaction pathway for the radical is the
scavenging by DOC. In our studies, for the pH range 6 to 8 and
in seawater with a total alkalinity of 2.4 mM and dissolved
organic carbon of 87 ␮M, the formation of HCO●3 (King 1998)
and Br⫺
2 (Moffet and Zika, 1987) are favorable and the main
reaction pathway is expected to be Fe(II).
The overall Fe(II) oxidation rate is expressed in terms of the
weighted sum of the oxidation rates of the individual Fe(II)
species
k ⫽ kFe2⫹␣Fe2⫹ ⫹ kFeOH⫺␣FeOH⫺ ⫹ kFe(OH)2␣Fe(OH)2
⫹ kFeHCO⫺3 ␣FeHCO⫺3 ⫹ kFe(CO3)␣Fe(CO3) ⫹ kFe(CO3)2⫺
␣Fe(CO3)2⫺
2
2
⫹ kFe(CO3)(OH)⫺␣Fe(CO3)(OH)⫺ ⫹ kFeCl⫹␣FeCl⫹ ⫹ kFeSO4␣FeSO4
(21)
where ␣i is the molar fraction of Fe(II) species in the solution.
To verify this rate law and to determine the individual rate
constants, we have defined a kinetic model (Santana-Casiano et
al., 2004) with all pertinent dissociation, complex formation
and oxidation reaction for the Fe(II) species in seawater (Table
2). The presence of the different major inorganic species in the
88
M. González-Davila, J. M. Santana-Casiano and F. J. Millero
Fig. 3. Fe(II) speciation in seawater media with an ionic strength of
0.7 M following the equilibrium constants presented in Table 2.
Fig. 4. Experimental and predicted Fe(II) concentration as a function
of pH using the individual rate constants presented in Table 3. Initial
[Fe(II)]o ⫽ 250 nM and [H2O2]o ⫽ 275 nM.
seawater solution, Ca2⫹, Mg2⫹, K⫹ and Na⫹ that affect the
carbonate and Fe(II) speciation were also considered in the 0.7
M seawater solution to account for the formation of ion-pairs.
To gain an insight into the role played by the different Fe(II)
species in the oxidation kinetic of Fe(II), Figure 3 shows the
Fe(II) speciation in seawater between pH 6.5 to 8.2. In this pH
range Fe(II) speciation is dominated by the Fe2⫹, FeCl⫹ and
FeSO4 species while only at pH higher than 8.2 does the FeCO3
species become higher than Fe2⫹. The concentrations of
FeOH⫹ and Fe(OH)2 only reach values of 1.1 ⫻ 10⫺9 M and
7.9 ⫻ 10⫺13 M at pH 8 for a total Fe(II) concentration of 1.25
⫻ 10⫺7 M. To get values for the nine individual rate constants
for the oxidation of Fe(II) the kinetic model was applied to the
experimental results at the different pH and HCO⫺
3 concentrations. Using the methods described earlier, the computed rate
constants describing all the experimental conditions presented
in this work with a 95% confidence, are given in Table 3. The
fitting of these rate constants in describing the experimental
data are shown in Figure 4 where the lines represent the output
from the kinetic model. The studies carried out both at different
carbonate concentrations and pH (Fig. 5) allowed us to obtain
reasonable values for the oxidation rates of FeHCO⫹
3 , FeCO3,
Fe(CO3)2⫺
and Fe(CO3)(OH)⫺ species and Fe2⫹, Fe(OH)⫹
2
and Fe(OH)2 species, respectively. Figure 6 shows the contributions of the different Fe(II) species to the total Fe(II) oxidation rate. For the pH range of this study, FeOH⫹ is the most
important contributing species to the overall oxidation rate
while FeCO3 contributes half of the FeOH⫹ value. These two
Fe(II) species are consistent with the first order pH-dependence
on Fe(II) oxidation reported by previous studies (Moffet and
Zika, 1987; Millero and Sotolongo, 1989) and shown in Figure
3. At values of pH higher than 8, the Fe(OH)2 and Fe(CO3)2
species contribute over 20%. The Fe(OH)2 complex is the most
important species at pH higher than 8.1. Thus, in seawater at
low pH the first order pH dependence is due to the FeOH⫹
complex; while at high pH the Fe(OH)2 complex causes the
second order pH dependence. This is similar to the oxidation of
Fe(II) with O2 in seawater at carbonate levels of 2 ␮M (Millero
et al., 1987).
King and Farlow (2000) have computed individual Fe(II)
oxidation rates with H2O2 at nanomolar levels in pure water
with different concentrations of HCO⫺
3 . They also found a
second order pH-dependence for the oxidation of Fe(II) with
H2O2 for pure water with different bicarbonate concentrations
at pH over 8. Their values for the two species, kFeOH ⫽ 3.8 ⫻
105 and kFeCO3 ⫽ 2.2 ⫻ 104 (M⫺1 s⫺1), are 75% lower and
15% higher, respectively, than our values in seawater. Their
studies in NaHCO3 were carried out in the pH range 4.62 to
7.05, very far from the pH of seawater, where the change in the
speciation of both Fe(II) and HCO⫺
3 occurs (Fig. 3). The
presence of major ions Ca2⫹ and Mg2⫹ in seawater interacting
both with carbonate and hydroxide species and differences in
Table 3. Fe(II) oxidation rates with hydrogen peroxide in seawater (log k, M⫺1 s⫺1) at different temperatures valid for the pH range 6 to 8.2
calculated from the kinetic model. Values in pure water according to Eqn. 16 (25°C, 2 mM NaHCO3) are also included.
Species
2⫹
Fe
FeHCO⫹
3
FeCO3
Fe(CO3)2⫺
2
Fe(CO3)OH⫺
⫹
FeOH
Fe(OH)2
FeCl⫹
FeSO4
3°C
10°C
20°C
25°C
35°C
Pure Water
1.2
⬍1.6
3.5
5.5
⬍1.4
5.9
8.5
⬍1.6
⬍1.7
1.6
⬍1.6
3.7
6.0
⬍1.4
6.0
8.8
⬍1.6
⬍1.7
2.1
⬍1.6
4.1
6.7
⬍1.4
6.1
9.0
⬍1.6
⬍1.7
2.4
⬍1.6
4.3
7.2
⬍1.4
6.2
9.3
⬍1.6
⬍1.7
2.8
⬍1.6
4.6
7.9
⬍1.4
6.3
9.6
⬍1.6
⬍1.7
2.9
⬍1.8
4.9
7.8
⬍1.9
6.8
9.9
Oxidation of Fe(II) by H2O2
Fig. 5. Experimental and predicted Fe(II) concentration as a function
of pH in the presence of different NaHCO3 concentrations at 25°C
using the kinetic model and constants presented in Table 3.
the ionic strength between these two studies can account for the
differences in the rate constants. The concentration of free
2⫹
HCO⫺
in seawater is reduced com3 available to complex Fe
pared to the same amount in to pure water due to the presence
of high concentrations of Ca2⫹ and Mg2⫹ (see below).
Our model has also been applied to the Fe(II) oxidation with
different concentrations of H2O2. The total Fe(II) concentration
determined from the model and the experimental rate data are
in very good agreement. When, however, the concentration of
H2O2 was reduced and kept below the stoichiometry ratio, the
model output gave slower rates than those determined experimentally. At a very low concentrations of hydrogen peroxide,
the presence of trace amounts of oxygen in the solution even
after of bubbling with N2 can also affect the concentration of
Fe(II) in the solutions. After one hour bubbling with N2, the O2
concentration in the solution was still 17 ␮M (210 ␮M before
bubbling with N2). This amount of O2 has been neglected in all
the studies using high amounts of H2O2. To account for this
effect, oxidation of the different Fe(II) species with molecular
oxygen were also included in our kinetic model following
Fig. 6. Contribution of specific Fe(II) species in total Fe(II) oxidation
rate by hydrogen peroxide in seawater (S ⫽ 36.244) and 25°C.
89
Fig. 7. Experimental and predicted Fe(II) concentration as a function
of [H2O2] concentration in seawater (S ⫽ 36.244, pHF ⫽ 8.17, T ⫽
25°C). Lines represent model output considering both the absence of
dissolved O2 (solid lines) and the presence of O2 at the concentration
of 17 ␮M (dashed lines).
Santana-Casiano et al. (2004). By considering the amount of O2
in the solutions, faster oxidation rates were obtained and give a
good fit of the experimental results (the dashed lines in Fig. 7).
The model shows that when the dissolved O2:H2O2 ratio is
higher than 100, the oxidation with oxygen becomes important
and needs to be considered. A 20% decrease in the concentration of the most active Fe(II) species is observed after reacting
for two minutes when the initial H2O2 concentration was 7.22
⫻ 10⫺8 M and O2 equal 17 ␮M. In most of our studies, the
O2:H2O2 ratio was 60 and the values predicted by the model
with and without considering the O2 contribution are within the
estimated error.
A similar effect was also observed when the model was
applied to describe the effect of Fe(II) concentrations on the
oxidation rate with H2O2. At low Fe(II) concentrations,
below the stoichiometric ratio, the kinetic constants presented in Table 3, describe the experimental oxidation rates.
However, when the concentration of [Fe(II)]o ⫽ 5 ⫻ 10⫺7 M
and [H2O2]o ⫽ 2.75 ⫻ 10⫺7 M, the kinetic model predicts a
lower rate for the oxidation of Fe(II). The presence of
dissolved O2 at 17 ␮M decreases the concentration of Fe(II)
in the solution by 15% after 3 mins. when oxygen is not
considered. When the O2 is considered the kinetic model
agrees with the experimental data.
The effect of temperature on the rates of oxidation of
Fe(II) with H2O2 was also examined using the kinetic model.
The hydrolysis constants, carbonate stability constants with
Fe(II) and with the major ions at 10°C, are given in Table 4.
The experimental data at 10°C as a function of pH and
concentrations of total bicarbonate and the data at temperature of 3°, 20° and 35°C (Table 1) were used in the fitting
procedure to compute the nine individual oxidation rate
constants for each temperature. The computed values for
these rate constants are given in Table 3. Figure 8 shows the
model output for the oxidation of Fe(II) with H2O2 with
different concentrations of HCO⫺
3 and pH at 10°C. The
contributions of the five kinetically active species at 10° and
25°C (Fig. 6) are shown in Figure 9. Our model predicts at
90
M. González-Davila, J. M. Santana-Casiano and F. J. Millero
Table 4. Stability constants for the formation of Fe(II) and Fe(III)
inorganic complexes considered for the kinetic model at 10°C.
Species
Log K
(0.7 mol
L⫺1, 10°C)
Ref
H2O N H⫹ ⫹ OH⫺
⫹
CO2 ⫹ H2O N HCO⫺
3 ⫹ H
2⫺
HCO⫺
⫹ H⫹
3 N CO3
Na⫹ ⫹ HCO⫺
3 N NaHCO3
Na⫹ ⫹ CO2⫺
N NaCO⫺
3
3
⫹
Ca2⫹ ⫹ HCO⫺
3 N CaHCO3
2⫹
2⫺
Ca ⫹ CO3 N CaCO3
⫹
Mg2⫹ ⫹ HCO⫺
3 N MgHCO3
Mg2⫹ ⫹ CO2⫺
N
MgCO
3
3
2 Mg2⫹ ⫹ CO2⫺
N Mg2(CO3)2⫹
3
2⫹
⫺
⫹
Mg ⫹ OH N MgOH
⫹
Fe2⫹ ⫹ HCO⫺
3 N FeHCO3
Fe2⫹ ⫹ CO2⫺
N
FeCO
3
3
Fe2⫹ ⫹ 2 CO2⫺
N Fe(CO3)2
3
2⫹
2⫺
⫺
Fe ⫹ CO3 ⫹ OH N Fe(CO3)(OH)⫺
Fe2⫹ ⫹ H2O N Fe(OH)⫹ ⫹ H⫹
Fe2⫹ ⫹ 2 H2O N Fe(OH)2 ⫹ 2 H⫹
Fe2⫹ ⫹ Cl⫺ N FeCl⫹
Fe2⫹ ⫹ SO2⫺
N FeSO4
4
H⫹ ⫹ SO2⫺
N HSO⫺
4
4
⫺14.27
⫺5.91
⫺9.38
⫺0.53
0.42
0.33
1.94
0.28
1.82
2.59
1.90
0.97
4.14
6.19
9.0
⫺10.17
⫺21.98
⫺0.12
0.77
0.17
1
1
1
3
3
3
2
3
2
3
1
1
1
1
4
1
1
1
1
1
1. Millero and Pierrot, 2002. 2. Smith and Martell, 1976. 3. No
significant effect was observed for these constants. 4 The same temperature coefficient as for Fe(CO3)2
⫹
10°C that the FeOH species is the most important one,
which is in agreement with the first-order pH-dependence of
the Fe(II) oxidation at this temperature (Eqn. 14 and 15).
The FeCO3 species is the second most active species until
pH 8 where the Fe(CO3)2⫺
contributes over 30% to the
2
overall rate constant. At this pH, carbonate species have a
greater effect at 10°C than 25°C. This accounts for the larger
effect of the addition of carbonate to the oxidation rates
(Eqn. 15) at 10°C. The effects of temperature on the individual oxidation rate constants are shown in Figure 10. The
values for the individual oxidation rate constants for the five
Fig. 8. Experimental and predicted Fe(II) concentration as a function
of pH in natural seawater ([NaHCO3]T ⫽ 2.05 mM) and with 2.0 mM
NaHCO3 added at 10°C using the kinetic model and constants presented in Table 3 and Table 4.
Fig. 9. Contribution of specific Fe(II) species in total Fe(II) oxidation
rate by hydrogen peroxide in seawater (S ⫽ 36.244) and 10°C.
most kinetically reactive species in seawater can be determined from
ln kFe2⫹ ⫽ 38.0(⫾1.1) ⫺ 9529(⫾320)1 ⁄ T
Ea ⫽ 79.2 ⫾ 2.7 kJ mol⫺1 (22)
ln kFeOH⫹ ⫽ 24.20(⫾0.8) ⫺ 2757(⫾232)1 ⁄ T
Ea ⫽ 22.9 ⫾ 1.9 kJ mol⫺1 (23)
ln kFe(OH)2 ⫽ 44.4(⫾1.5) ⫺ 6658(⫾450)1 ⁄ T
Ea ⫽ 55.4 ⫾ 3.7 kJ mol⫺1 (24)
ln kFeCO3 ⫽ 33.2(⫾0.3) ⫺ 6757(⫾92)1 ⁄ T
Ea ⫽ 56.2 ⫾ 0.8 kJ mol⫺1 (25)
⫽ 65.6(⫾1.2) ⫺ 14460(⫾337)1 ⁄ T
ln kFe(CO3)2⫺
2
Ea ⫽ 120.2 ⫾ 2.8 kJ mol⫺1 (26)
Equations 22–26 give energies of activation from 22.9 ⫾ 1.9 kJ
mol⫺1 to 120.2 ⫾ 2.8 kJ mol⫺1, respectively, for FeOH⫹ and
Fig. 10. The effect of temperature for the most contributing Fe(II)
species on the oxidation rate constants by H2O2, ki, in seawater at pHF
⫽ 8.17.
Oxidation of Fe(II) by H2O2
Table 5. Comparison of experimental Fe(II)-H2O2 oxidation rate
constants at nanomolar level in pure water with added NaHCO3 (King
and Farlow, 2000) with values determined by two different kinetic
models.
pH
mM
[NaHCO3]
(M⫺1 s⫺1)
Exp.
log k
Model log k
King and Farlow
(2000)
Model log k
This work
5.20
6.00
6.98
6.97
2.0
3.0
2.0
7.0
3.06
3.61
4.26
4.48
3.00
3.51
4.15
4.49
3.07
3.57
4.25
4.49
Fe(CO3)2⫺
2 . The values of log kFeCl, log kFeSO4, log kFeHCO3
and log kFe(CO3)OH, from our model are lower than 1.5 (ki in
M⫺1 min⫺1) and do not contribute to the overall oxidation rate
under the experimental conditions of this study. The values for
the individual kinetic oxidation rates determined using Equations 22–26 and estimates of the fraction of each Fe(II) species
following Eqn. 21 agree with the experimental values on the
average to 3% as can be observed in Figure 8.
Individual oxidation rates are expected to vary as a function
of the ionic strength (King, 1998). To apply our model to
natural waters at different ionic strengths, Eqn. 16 was assumed
to be valid to describe the ionic strength dependence for all of
the individual oxidation rates in Table 3. The resulting oxidation rates together with equilibrium constants valid at I ⫽ 0
(King, 1998) were used to compute the Fe(II) oxidation rate in
the presence of different NaHCO3 concentrations in the pH
range 5 to 7. The experimental and model values determined by
King and Farlow (2000) were compared with our model output
using the individual oxidation rate constants shown in Table 3
(Table 5). The excellent agreement between experimental values and our model output makes our oxidation rate constants
for the individual Fe(II) species applicable to both seawater and
pure water with different concentrations of bicarbonate and
using the ionic strength dependence shown in Eqn. 16.
Results from this study have been compared with the
calculated oxidation rate determined by Millero and Sotolongo (1989) in seawater at 25°C and micromolar levels of
Fe(II) in Figure 11. Our rates are slightly lower than the
values determined by these authors. However, if we consider
the results from Millero and Sotolongo (1989) at micromolar
levels with a 2:1 Fe(II):H2O2 stoichiometry ratio (Eqn. 18
and 19), the overall rate constants presented in Table 2 from
their work, kMS, should be equal to kMS ⫽ 2 · k following
Eqn. 20 and 21. When this adjustment is made both results
are in excellent agreement. Figure 11 shows the predicted
overall rate constant from our model in the range of pH 6 to
8.2 determined according to Eqn. 21. The predicted rates fit
the overall rate constants computed from experimental
pseudofirst order oxidation rate at pH lower than 7.5. At pH
higher than 7.5, as was observed in the pseudofirst order data
treatment, deviations occur between measured and predicted
rate constants. When the experimental data were treated as a
pseudosecond order rate constant for a pH higher than 7.5
(Fig. 11), the results are in good agreement. This finding is
in accordance with the contribution computed from our
model for the different species to the overall rate constant
(Fig. 6 at 25°C and Fig. 9 at 10°C) and indicates that the
91
Fe(OH)2 and Fe(CO3)2⫺
species are important at the pH
2
range above 7.5 and trace amounts of oxygen also contributes to the process. Figure 11 also shows the predicted rates
using the pure water model of King and Farlow (2000). The
effect of ionic strength on the Fe(II)-H2O2 (Millero et al.,
1991, and in this work) and Fe(II)-O2 (King, 1998) oxidation
rates have been previously described. If we assume the same
ionic strength dependence for the Fe(II)-H2O2 oxidation
rates (Eqn. 16) to be applicable to the individual oxidation
rates presented by King and Farlow (2000) in pure water
solutions and taking into account also the corrections for the
media composition on the speciation of Fe(II), their model
predictions are not consistent with the experimental rate
measurements. Although King and Farlow (2000) used their
set of constants to describe the oxidation of Fe(II) with H2O2
in pure water at nanomolar Fe(II) concentrations and high
concentrations of NaCl and NaClO4, they did not make
corrections for the changes in the ionic strength. Our set of
constants determined in seawater considers the effects of
changes in the media and the interactions affecting the
speciation of Fe(II) with carbonate and hydroxide ions. The
effects due to changes in the ionic strength are also considered to compare experimental data with model output for
pure water with different carbonate concentrations (Table 5).
The discrepancies between kinetic data from both models
can be ascribed to the role of the carbonate complex FeCO3
over the hydroxide complex FeOH⫹ in the King and Farlow
model. The parallel behavior presented by both species in
the speciation diagram (also applicable to Fe(OH)2 and
Fe(CO3)2⫺
species) as a function of pH (Fig. 3) together
2
with the interactions between both ligands with major ions in
solutions makes the set of constants presented in this work
applicable over a larger range of experimental conditions.
5. CONCLUSIONS
The results obtained in this study have important implications for the redox chemistry of iron in the marine envi-
Fig. 11. Comparison of published oxidation rates in seawater by
H2O2 at 25°C by Millero and Sotolongo (1989) and this work considering first order pH dependence and second order pH dependence (pH
⬎7.5). Model predictions for oxidation rate constants of Fe(II) by H2O2
as a function of pH following King and Farlow (2000) corrected for
ionic strength effects (Eqn. 16) and media composition. Our model rate
constants in seawater and 25°C (Table 3) are also included.
92
M. González-Davila, J. M. Santana-Casiano and F. J. Millero
ronment. In seawater, the oxidation rate of nanomolar Fe(II)
with H2O2 is a function of pH, temperature and the concentration of HCO⫺
3 and H 2O 2. The oxidation rate constants
determined for the different Fe(II) species allow us to estimate the relative importance of each species and the competitive effect between the reaction with H2O2 and O2.
FeOH⫹ is the most important iron species controlling the
Fe(II) oxidation with H2O2 in the pH range of natural
seawater systems. At a pH higher than 8, the Fe(OH)2 and
Fe(CO3)2⫺
species contribute over 20% to the rates. Our
2
model predicts that at 10°C the FeOH⫹ species is the more
important than FeCO3 species below pH 8. Above a pH of 8
Fe(CO3)2⫺
species contributes over 30% to the overall rate
2
constant. At 25°C and pH 8 the carbonate species are at a
higher contribution, resulting in the greater effect of added
carbonate to the oxidation rates. The results from this study
indicate that the oxidation rate of nanomolar Fe(II) is predicted with a kinetic model taking into account both the
speciation of Fe(II) as a function of pH, carbonate concentration and the ionic interactions between the major cations
in seawater and the carbonate species. When these interactions are removed and ionic strength effects are considered,
our model is found to be applicable to pure water solutions.
Acknowledgments—We thank C. Moore for her advice on H2O2
analysis and for her helpful suggestions. We also would like to
thank Dr. Enrique González-Dávila who applied the “Statistica
Program” to our data and advice on how to select the best set of
experiments to determine the individual oxidation rates. This study
was supported by the Project BQU2003-04010 of Ministerio de
Ciencia y Tecnología from Spain and the Oceanographic section of
National Science Foundation. The authors wish to thank the reviewers for providing many helpful suggestions to improve the paper.
Associate editor: W. H. Casey
REFERENCES
Barb W. C., Boxendale J. H., George P. and Hargrove K. R. (1951)
Reactions of ferrous and ferric ions with hydrogen peroxide. Part
II. The ferric ion reaction. Trans. Faraday Soc. 47, 591– 616.
Box G. E. P. and Draper. N. R (1987) Empirical Model-Building and
Response Surfaces. Wiley, New York.
Bray W. C. and Gorin M. H. (1932) Ferryl ion, a compound of
tetravalent iron. J. Am. Chem. Soc. 54, 212419.
Buerge I. J. and Hug S. J. (1998) Influence of organic ligands on
chromium (VI) reduction by iron(II). Environ. Sci. Technol. 32,
2092–2099.
Dunford H. B. (2002) Oxidation of iron(II)/(III) by hydrogen peroxide from aquo to enzyme. Coordination Chem. R. 233–234,
311–318.
Emmenegger L., King D. W., Sigg L. and Sulzberger B. (1998)
Oxidation kinetics of Fe(II) in a eutrophic Swiss lake. Environ. Sci.
Technol. 32, 2990 –2996.
Gibbs C. R. (1976) Characterization and application of ferrozine
iron reagent as a ferrous iron indicator. Anal. Chem. 48, 1197–
1201.
Haber F. and Weiss J. (1934) The catalytic decomposition of hydrogen peroxide by iron salts. Proc. Roy. Soc. London, Ser. A,
332–351.
Hansen H. P. (1999) In Methods of Seawater Analysis (eds. K. Grasshoff, K. Kremling and M. Ehrhardt). Wiley-VCH. Germany.
Hong H. and Kester D. H. (1986) Redox state of iron in the offshore
waters of Peru. Limnol. Oceanogr. 31, 512–524.
Hoigné J. (1988) The chemistry of ozone in water. In Process Technologies for Water Treatment. (ed. S. Stucki), pp 121–143. Plenum
Press, New York.
King D. W. (1998) Role of carbonate speciation on the oxidation
rate of Fe(II) in aquatic systems. Environ. Sci. Technol. 32,
2997–3003.
King D. W. and Farlow R. (2000) Role of carbonate speciation on the
oxidation of Fe(II) by H2O2. Mar. Chem. 70, 201–209.
King D. W., Lounsbury H. A. and Millero F. J. (1995) Rates and
mechanism of Fe(II) oxidation at nanomolar total iron concentration. Environ. Sci. Technol. 29, 818 – 824.
Kremer M. L. (1999) Mechanism of the Fenton reaction. Evidence for
a new intermediate. Phys. Chem. Chem. Phys. 1, 3595–3605.
Larson R. A. and Zepp R. G. (1988) Reactivity of the carbonate radical
with aniline derivatives. Environ. Toxicol. Chem. 7, 265–274.
Mendes P. (1997) Biochemistry by numbers: simulation of biochemical
pathways with Gepasi 3. Trends Biochem. Sci. 22, 361–363.
Millero F. J. (1986) The pH of estuarine waters. Limnol. Oceanogr. 31,
839 – 847.
Millero F. J. (1995) Thermodynamics of the carbon dioxide system in
the oceans. Geochim. Cosmochim. Acta 59, 661– 677.
Millero F. J. and Hawke D. J. (1992) Ionic interactions of divalent
metals in natural waters. Mar. Chem. 40, 19 – 48.
Millero F. J. and Pierrot D. (2002) Speciation of metals in natural
waters. In Chemistry of Marine Water and Sediments (eds. A.
Gianguzza, E. Pelizzetti and S. Sammartano), pp. 191–220.
Springer, Berlin.
Millero F. J. and Schreiber D. R. (1982) Use of the ion pairing model
to estimate activity coefficients of the ionic components of natural
waters. Am. J. Sci. 282, 1508 –1540.
Millero F. J., Sotolongo S. and Izaguirre M. (1987) The kinetics of
oxidation of Fe(II) in seawater. Geochim. Cosmochim. Acta 51,
793– 801.
Millero F. J. and Sotolongo S. (1989) The oxidation of Fe(II) with
H2O2 in seawater. Mar. Chem. 53, 1867–1873.
Millero F. J., González-Dávila M. and Santana-Casiano J. M. (1995a)
Reduction of Fe(III) with sulfite in natural waters. J. Geophys. Res.
100, 7235–7244.
Millero F. J., Yao W. and Aicher J. (1995b) The speciation of Fe(II)
and Fe(III) in natural waters. Mar. Chem. 50, 21–39.
Millero F. J., Sotolongo S., Stade D. J., and Vega C. A. (1991) Effect
of ionic interactions on the oxidation of Fe(II) with H2O2 in
aqueous solutions. J. Sol. Chem. 20, 1079 –1092.
Moffet J. W. and Zika R. G. (1983) Reaction kinetics of hydrogen
peroxide with copper in seawater: Implications for its existence in
the marine environment. Mar. Chem. 13, 239 –251.
Moffet J. W. and Zika R. G. (1987) Reaction kinetics of hydrogen
peroxide with copper and iron in seawater. Environ. Sci. Technol.
21, 804 – 810.
Moore C. A., Farmer C. T. and Zika R. G. (1993) Influence of the
Orinoco River on hydrogen peroxide distribution and production in
the eastern Caribbean. J. Geophys. Res. 98, 2289 –2298.
Murray J. W. and Gill G. (1978) The geochemistry of iron in Puget
Sound. Geochim. Cosmochim. Acta 42, 9 –19.
Pignatello J. J. (1992) Dark and photoassisted Fe3⫹ catalyzed degradation of chlorophenoxy herbicides by hydrogen peroxide. Environ.
Sci. Technol. 26, 944 –951.
Pullin M. J. and Cabaniss S. E. (2003) The effects of pH, ionic
strength and iron-fulvic acid interactions on the kinetics of
non-photochemical iron transformations. I. Iron (II) oxidation
and iron (III) colloid formation. Geochim. Cosmochim. Acta 67,
4067– 4077.
Santana-Casiano J. M., González-Dávila M., Rodríguez M. J. and
Millero F. J. (2000) The effect of organic compounds in the
oxidation kinetics of Fe(II). Mar. Chem. 70, 211–222.
Santana-Casiano J. M., González-Dávila M. and Millero F. J. (2004)
The oxidation of Fe(II) in NaCl-HCO⫺
3 and seawater solutions in
the presence of phthalate and salicylate ions: a kinetic model. Mar.
Chem. 85, 27– 40.
Smith R. M. and Martell A. E. (1976) Inorganic complexes. Vol. 4.
Plenum Press. New York.
Oxidation of Fe(II) by H2O2
Sung W. and Morgan J. J. (1980) Kinetics and product of ferrous iron
oxygenation in aqueous solutions. Environ. Sci. Technol. 14, 561–
568.
Viollier E., Inglett P. W., Hunter K., Roychoudhury A. N. and van
Cappellen P. (2000) The ferrozine method revisited: Fe(II)/
Fe(III) determination in natural waters. App. Geochem. 15, 785–
790.
Zhang J. Z., Kelble C. and Millero F. J. (2001) Gas-segmented continuous flow analysis of iron in water with a long liquid waveguide
capillary flow cell. Anal. Chim. Acta 438, 49 –57.
93
Zika R. G. and Saltzmann E. S. (1982) Interaction of ozone and
hydrogen peroxide in water: Implication for analysis of H2O2 in air.
Geophys. Res. Lett. 9, 231–234.
Zika R. G., Saltzman E. S. and Cooper W. J. (1985a) Hydrogen
peroxide concentrations in the Peru upwelling Area. Mar. Chem.
17, 265–275.
Zika R. G., Moffett J. W., Pestane R. G., CooperW. J., and Saltzman
E. S. (1985b) Spatial and temporal variations of hydrogen peroxide in Gulf of Mexico waters. Geochim. Cosmochim. Acta 49,
1173–1184.