Effects of Fe III Concentration Speciati

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Author's personal copy

Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49

Contents lists available at SciVerse ScienceDirect

Journal of Photochemistry and Photobiology A:


Chemistry
journal homepage: www.elsevier.com/locate/jphotochem

Effects of Fe(III)-concentration, speciation, excitation-wavelength and light


intensity on the quantum yield of iron(III)-oxalato complex photolysis
Christian Weller, Sabrina Horn, Hartmut Herrmann ∗
Leibnitz-Institut für Troposphärenforschung, Permoserstraße 15, 04318 Leipzig, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Iron(III)oxalato complexes do frequently occur in the environment, specifically in surface waters,
Received 24 October 2012 in atmospheric waters (clouds, rain, fog) or in waste waters. Due to their high photo-reactivity
Received in revised form 11 January 2013 and their absorption overlap with the actinic spectrum, Fe(III)oxalato complex photochemistry is
Accepted 19 January 2013
widespread and of broad interest. Fe(III)oxalato complex photolysis in deaerated solutions using
Available online xxx
single excimer laser flash photolysis at 308 and 351 nm and continuous Hg(Xe)-lamp irradiation at
313, 366 and 436 nm was quantified via Fe(II) quantum yield measurements with phenanthroline
Keywords:
complexometry and UV–vis detection. Measured Fe(II) quantum yields showed a dependence on initial
Iron
Oxalate
Fe(III)ferrioxalate concentration and irradiation energy at below millimolar concentrations. Individual
Complex molar extinctions (in l mol−1 cm−1 ) and individual quantum yields (˚) were determined for initial Fe(III)
Photochemistry concentrations of 4.85 × 10−4 M for the 1:2 (FeOx2 − ) and 1:3 (FeOx3 3− ) complexes applying a regression
Quantum yield analysis for solutions containing variable ratios of 1:2 and 1:3 complexes: ε1:2, 308 nm = 2300 ± 90,
ε1:3, 308 nm = 2890 ± 40, ˚1:3, 308 nm = 0.93 ± 0.09; ε1:2, 351 nm = 1040 ± 30, ε1:3, 351 nm = 1120 ± 20,
˚1:3, 351 nm =0.88 ± 0.08; ε1:2, 313 nm = 2055 ± 111, ε1:3, 313 nm = 2663 ± 37, ˚1:3, 313 nm = 0.12 ± 0.05;
ε1:2, 366 nm = 753 ± 357, ε1:3, 366 nm = 709 ± 10, ˚1:2, 366 nm = 1.17 ± 1.46, ˚1:3, 366 nm = 0.91 ± 0.09;
ε1:2, 436 nm = 55 ± 9, ε1:3, 436 nm = 22 ± 2, ˚1:2, 436 nm = 1.40 ± 0.40, ˚1:3, 436 nm = 1.00 ± 0.20. Individual
quantum yields for the 1:2 complex could only be determined for the excitation wavelengths 366
and 436 nm due to non-linearity of the data for 308, 351 and 313 nm. The non-linearity is ascribed
to complicated interactions of secondary reactions involving Fe(III)oxalato educt-complexes, carboxyl
radicals and Fe(II)-radical complexes. The 1:2 complex has generally a higher quantum yield compared
to the 1:3 complex at all considered wavelengths.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction investigations of the subject in relation to waste water treat-


ment via so called advanced oxidation processes [2,4]. Oxalate
The photochemistry of the Fe(III)oxalato or ferrioxalate-system forms three different complexes with Fe(III) in acidic to neutral
is one of the most thoroughly investigated Fe(III) polycarboxy- solutions: mono(oxalato)iron(III), [Fe(C2 O4 )]+ ; bis(oxalato)ferrate,
late systems with many different backgrounds and applications. [Fe(C2 O4 )2 ]− and tris(oxalato)ferrate, [Fe(C2 O4 )3 ]3− , which occur
It was proposed as a chemical actinometer to measure light inten- simultaneously in different proportions depending on the pH
sities in a very convenient, inexpensive and reliable way at well and the Fe(III):oxalate ratio. Iron(III)oxalate complexes exhibit
defined conditions [1] and is still frequently used in photochem- strong LMCT absorption bands in the UV range. Fe2+ and reactive
istry. Furthermore, the photochemistry of ferrioxalate complexes oxalate fragments are ultimately produced upon photolysis of an
as a one representative of the class of Fe(III)polycarboxylate Fe(III)oxalato complex. The reactive fragments lead to the produc-
complexes can play a role in environmental systems such as tion of CO2 •− , O2 •− /HO2 • and ultimately H2 O2 and • OH from the
atmospheric waters (cloud droplets, aqueous particles, rain) [2,3] Fenton reaction.
where the direct degradation of oxalate via photolysis and the The mechanism of photoreduction and subsequent reactions in
secondary production of reactive radicals that will react fast the ferrioxalate system were subject of numerous investigations
with any present substance can cause the turnover of organic [5–16]. There are two different basic ideas of the ferrioxalate reac-
compounds. The radical generating potential has also led to tion mechanism, following the excitation of the complex [17]. One
perception is the intramolecular electron transfer from the oxalate
ligand to the center ion Fe3+ and the formation of a long lived radical
complex (R1) or the formation of a C2 O4 •− radical (R2) [11,13]:
∗ Corresponding author. Tel.: +49 341 2717 7024; fax: +49 341 2717 997024.
E-mail addresses: weller@tropos.de (C. Weller), horns@tropos.de (S. Horn),
herrmann@tropos.de (H. Herrmann). [Fe(III) (C2 O4 )3 ]3− + hv → [Fe(II) (C2 O4 • )(C2 O4 )2 ]3− or (R1)

1010-6030/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jphotochem.2013.01.014
Author's personal copy

42 C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49

[Fe(III) (C2 O4 )3 ]3− + hv → [Fe(II) (C2 O4 )2 ]2− + C2 O4 •− (R2) 2.2. Excimer laser flash photolysis

The C2 O4 •− radical will then decarboxylate instantly and form CO2 For the photolysis experiments, a volume of 2.6 ml of
and CO2 •− [18]: iron(III)oxalato-complex solution was transferred to a standard
1 cm optical path length spectroscopy cuvette made of fused sil-
C2 O4 •− → CO2 •− + CO2 k1st = 2 × 106 s−1 (R3)
ica. Deoxygenated samples were transferred under an argon flow
Another possibility is the sequential cleavage of the Fe(III) O bond and the cuvette was capped with a Teflon seal. Then, the cuvette
between iron and one oxalate ligand and its C C bond which would was placed in a cuvette-holder in front of an excimer laser (Com-
yield a biradical complex or two carboxylate radicals [9,14,19]: pex 200 and 205, Lambda Physik). The lasers were operated either
at 308 nm or at 351 nm. The cuvette was exposed to a single laser
[Fe(III) (C2 O4 )3 ]3− + hv → [Fe(III) (CO2 • )2 (C2 O4 )2 ]3− or (R4) flash of about 20 ns duration. The output energy of the laser was
measured with a pyroelectric sensor (Coherent). A volume of 2 ml
[Fe(III) (C2 O4 )3 ]3− + hv → [Fe(III) (C2 O4 )2 ]− + 2CO2 •− (R5) exposed solution was transferred to a prepared flask containing
the necessary agents (see Section 2.4) to detect the photochemi-
Despite a controversial argument on the two pathways [17,19–21],
cally formed Fe2+ . In a standard course of the measurement of one
the discussion on the primary mechanism of ferrioxalate photolysis
quantum yield, the exposure and transfer to the flask have been
still open because neither point of view was sufficiently confirmed
repeated six times to calculate an average quantum yield from the
throughout the discussion. It has to be emphasized that in the end
results, including an error with Student t-factor on a 95% confidence
both mechanisms will lead to the formation of Fe2+ and at least one
interval.
CO2 •− radical, which is important for secondary reduction reac-
tions.
Regardless of the exact photochemical mechanism, there are a
large number of references that report quantum yields of Fe2+ for-
2.3. Hg(Xe) lamp continuous irradiation
mation from ferrioxalate photolysis which were determined for the
UV–vis wavelength range from 205 to 578 nm. The values that were
For the continuous irradiation experiments, the preparation of
obtained at the actinometry conditions specified by Hatchard and
the Fe(III) complex solutions was the same as for the laser flash
Parker [1] do generally agree [22–25] but values obtained at differ-
photolysis experiments. The irradiation cell was a 1 cm fused silica
ent experimental conditions, [26–31] which involve mostly lower
cuvette with a small inset for a magnetic stirrer out of optically non-
initial Fe3+ concentrations, are quite scattered. These observed
transmittive black glass (Hellma). The cell was placed in a holder
inconsistencies in the reported ferrioxalate quantum yields gave
above a magnetic stirrer and in front of the output of a prism-
rise to a reinvestigation of the system quantum yield in the present
monochromator (model M4QII, Carl Zeiss). A 200 W Hg(Xe) lamp
work and the results will be discussed. Furthermore, the influence
(Hamamatsu) was used as irradiation source and the 313, 366 or
of Fe(III) oxalate speciation, irradiation wavelength and light inten-
436 nm lines were selected with the monochromator. Irradiations
sity on the photochemistry and measured Fe(II) quantum yields
have been performed under magnetic stirring typically for 2, 4, 6
will be addressed on the basis of excimer laser flash photolysis and
and 8 min and no longer than 10 min to observe the kinetics of
continuous irradiation experiments.
Fe2+ formation. Similar to the flash photolysis procedure, a volume
of 2 ml exposed solution was transferred to a prepared flask for
2. Experimental photochemically generated Fe2+ analysis.
The measurement of the photon flux entering the photolysis cell
2.1. Preparation of experimental solutions from the Hg(Xe) lamp/monochromator system was carried out with
standard ferrioxalate actinometry. A 6 mM ferrioxalate solution has
The species distribution of the different oxalate and possibly been used for the wavelengths 313 and 366 nm and for 436 nm a
inorganic iron(III) complexes was calculated with the com- 0.15 M ferrioxalate solution has been used with a quantum yield
puter program Visual MINTEQ (Department of Land and Water value of 1.01 [1]. Typical measured photon fluxes reaching the
Resources Engineering, Sweden), an equilibrium speciation model. cell from the lamp-monochromator system were qn,p = 10−10 to
For all experimental solutions Milli-Q water (Millipore, USA) with 10−9 mol s−1 .
R = 18 M was used. Fe(III) stock solutions were prepared from
Fe2 (SO4 )2 (Aldrich, 97%) and FeClO4 (Aldrich, low chloride) salts
and kept at pH ∼ 1 with H2 SO4 (Fluka, 98%) or HClO4 (Baker, 70%)
to prevent precipitation of insoluble Fe(III) species. 2.4. Detection of photochemically produced Fe2+
Solutions for the photochemical experiments were always
freshly prepared. When necessary, the ionic strength was adjusted Photochemically generated Fe2+ was analyzed by VIS-
by addition of NaClO4 . Adjustments of the pH were done with NaOH spectroscopy as 1,10-phenanthroline complex (phen; Fluka,
and HClO4 . Argon bubbling was used for deoxygenation through- puriss, p.a.) [Fe(phen)3 ]2+ using its local maximum extinction at
out all Fe(III) oxalate experiments in the present work. Preparations  = 510 nm with ε510 nm = 11,100 l mol−1 cm−1 [22] from which the
of Fe(III) complex solutions and photolysis experiments have been Fe2+ concentration was calculated. The photolyzed solutions con-
carried out in the dark under illumination by red light, to avoid tained both a major fraction of unreduced Fe(III) besides a smaller
photochemical Fe2+ production prior to the actual photolysis. The amount of Fe2+ which had to be detected. Phenanthroline forms
investigated complexes are not photoreactive in the red region of strong complexes with both Fe3+ and Fe2+ , with the [Fe(phen)3 ]3+
the spectrum. Additionally, the flasks containing Fe(III) complex complex having a much lower extinction coefficient than the
solutions were covered with aluminum foil. All experiments have [Fe(phen)3 ]2+ complex at 510 nm. Nevertheless, it causes a small
been carried out at room temperature 20 ◦ C. An UV–vis spectrum of but noticeable interference with the Fe2+ analysis. Therefore, a
the Fe(III) complex solutions has been recorded for each quantum suppression of this interference was applied, were the Fe3+ is
yield measurement to determine the extinction coefficient at the masked by nitrilo-triacetate (NTA: Fluka, ≥99%) in presence of a
excitation wavelengths (Lambda 900 double beam spectrometer, glycine–HCl buffer adjusted to pH = 2.9 (glycine: Sigma, 99%, HCl:
Perkin Elmer). Fluka, 37%, Riedel de Haen, 1 M) [32].
Author's personal copy

C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49 43

2.5. Calculation of Fe2+ quantum yield

The quantum yields ˚ from the excimer laser flash photolysis


experiments were calculated in the following way. The incident
amount of photons per pulse [Ph]I0 was calculated from the mea-
sured excimer laser energy and the amount of photons absorbed
by the ferrioxalate-complex(es), [Ph]abs , was computed via:

[Ph]abs = [Ph]I0 · (1 − 10−εcd ) (1)

Finally, the quantum yield is calculated by:

[Fe2+ ]
˚= (2)
[Ph]abs

The calculation of quantum yields from continuous Hg(Xe) lamp


irradiation experiments was performed with the absorbed pho-
ton flux Iabs from the incident photon flux, qn,p , the rate of Fe2+
production, d[Fe2+ ]/dt and the reaction volume, Vr :

d[Fe2+ ] Vr
˚= · (3)
dt Iabs

Iabs = qn,p · (1 − 10−εcd ) (4)

2.6. Treatment of speciation effects on quantum yield

To measure individual quantum yields of [Fe(Ox)2 ]− and


[Fe(Ox)3 ]3− complexes, it is necessary to distinguish their con-
tributions to the overall quantum yield. It is impossible to fully
isolate all individual Fe(III) oxalato complexes in an aqueous solu-
tion through variation of pH and Fe(III):oxalate ratio because their
stability constants lie too close together or other complexes such
as [FeIII (OH)n ]3−n will form simultaneously. However, the pH and Fig. 1. Top: concentration dependent effective overall Fe2+ quantum yields from
308 nm excimer laser flash photolysis. Conditions: at pH = 1.2 (actinometry con-
Fe(III):oxalate ratio can be adjusted in such way that the sum of ditions) Fe(III):Ox = 1:3, [H2 SO4 ] = 0.05 M; at pH 3 and 5 Fe(III):Ox = 1:10; bottom:
bis- and tris(oxalato)ferrate complexes in a solution contains 100% concentration dependent fractions of 1:1, 1:2, 1:3 Fe(III) oxalate complexes of the
of the present Fe3+ . total [Fe3+ ] for each quantum yield measurement data point of the top plot (back-
The measured total molar extinction εtot of the Fe(III) complexes ground color codes the pH of the respective ˚ measurement series in the plot at the
top).
can be expressed as:

εtot = εi fi (5) 3. Results and discussion
i
3.1. Dependence of Fe2+ -quantum yield on initial
with εi being the individual molar extinction coefficient of the ith Fe3+ -concentration
Fe(III)-complex representing a fraction fi of the total Fe(III) [33]. In
this work, the complex speciation and thus fi was calculated using 3.1.1. Overview
the speciation program Visual MINTEQ. Consequently, the over- Measurements of Fe2+ quantum yields after excimer laser flash
all extinction coefficient εtot which is the sum of the individual photolysis at 308 nm using initial Fe3+ concentrations between
extinction coefficients of bis- and tris(oxalato)ferrate complexes is 10−5 and 10−4 M gave much smaller values than expected from
proportional to the fraction of one complex species fi . In analogy earlier works [1,23]. Consequently, Fe2+ quantum yield measure-
to the calculation of the individual molar extinction, the individual ments as a function of initial Fe3+ concentration in the range of
quantum yields of different complex species can be determined in 10−6 to 10−3 M were made using excimer laser flash photolysis at
a similar manner: 308 nm (Fig. 1). At actinometry conditions of Hatchard and Parker
 [1] with milli-molar Fe3+ concentration the measured quantum
(˚tot ) · (εtot ) = (˚i · εi · fi ) (6) yield of the present work (˚308 nm = 1.27) agrees well with the val-
i
ues of ˚300 nm = 1.25 and ˚313 nm = 1.24 by Goldstein and Rabani
[23] and Hatchard and Parker [1], respectively. Only one work
˚tot is the measured overall quantum yield of a solution and ˚i reports a similar behavior of quantum yields under decreasing Fe3+
is that of the individual complex species. The repetition of at least concentrations using Hg(Xe) lamp continuous photolysis [30].
a number of i photolysis experiments under experimental condi-
tions where fi is varied, gives a set of linear equations which can 3.1.2. Possible effect of speciation change along with [Fe3+ ]
be solved for ˚i , if εi has been determined before. The measured decrease
quantity ˚tot × εtot can be regarded as a weighted average, where A decrease of the Fe3+ concentrations at actinometry conditions
the weighting is connected to the speciation of the metal [33]. (Fe(III):Ox = 1:3, [H2 SO4 ] = 0.05 M, pH = 1.2) down to 5 × 10−4 M
Author's personal copy

44 C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49

CO2 •− during photolysis. The precondition of having a constant


amount of approximately 70% of the initial [Fe3+ ] unphotolyzed
was kept during the above described experiments. Considering the
formation of CO2 •− and Fe2+ via:

[Fe(III) (C2 O4 )n ]3−2n + hv → Fe2+ + (n − 1)C2 O4 2− + CO2 + CO2 •−

(R6)

and the immediately following secondary reduction of unpho-


tolyzed [Fe(III) Oxn ]m complex:

[Fe(III) (C2 O4 )n ]3−2n + CO2 •− → Fe2+ + nC2 O4 2− + CO2 (R7)

leads to the conclusion that the amount of Fe2+ formed through the
secondary reaction path (R7) is most likely diminished between
initial Fe3+ concentrations from 5 × 10−4 to 1 × 10−5 M and even-
tually shut down at initial Fe3+ concentrations below 5 × 10−6 M.
Fig. 2. Measured effective overall Fe2+ -quantum yield at 308 nm excimer laser flash Reaction path (R7) becomes less favorable when the initial Fe3+ con-
photolysis at pH = 1.2 (actinometry conditions) and at pH = 3 compared to Gepasi centration is lowered because competing reactions, which divert
modeled overall Fe2+ -quantum yields with different sets of kinetic constants and or “scavenge” CO2 •− , gain more importance. Such competing reac-
initial conditions (see text and Table 1).
tions are the recombination of 2 CO2 •− (R8) and the reaction with
residual dissolved O2 (R9), which can be still around 5 × 10−7 M
did not change the quantum yield (red data points in Fig. 1). A [34], despite thorough Argon-purging [35].
further decrease of the Fe3+ concentrations at actinometry con-
ditions (pH = 1.2) would have systematically lowered the part of 2CO2 •− → C2 O4 2− kR8 = 1.2 × 109 M−1 s−1 (R8)
Fe3+ bound by oxalate in favor of Fe(III)hexaaqua, Fe(III)sulfato CO2 •− + O2 → CO2 + O2 •− kR9 = 6.5 × 109 M−1 s−1 (R9)
and Fe(III)hydroxo complexes. These complexes would have con-
tributed to the photochemistry which should be exclusively that (R9) is only competing when the formed O2 reacts with other
•−

of Fe(III)oxalato complexes. Therefore, a tenfold (molar) excess of components in the solution than [Fe(III) (C2 O4 )n ]3−2n because O2 •−
oxalate over Fe(III) was used for the measurement series at pH and Fe(III) produce also Fe2+ .
3 and 5, to ensure complete complexation by oxalate. The sum
of Fe(III)(H2 O)6 , Fe(III)sulfato and Fe(III)hydroxo complexes never 3.1.4. Simulation of kinetic effect through [Fe3+ ] decrease
exceeded one percent of the total [Fe3+ ] in all experiments. It has been attempted to quantify the experimentally observed
Solutions at pH = 3 are dominated by the bis(oxalato) complex quantum yield dependence on [Fe3+ ] with available rate constants
and solutions at pH = 5 dominated by the tris(oxalato) complex for (R8) and (R9) and following HOx /HOx + iron reactions
(Fig. 1). Solutions at both pH = 3 and 5 show a decrease of the quan- [36–46] using Gepasi, a (bio)chemical kinetic simulation software
tum yield when the initial Fe3+ concentration is lowered, with the (http://www.gepasi.org, [47]). The concentration dependence of
solution at pH = 5 having always the lower value. This points toward the quantum yields has been modeled with three different sets of
a lower individual quantum yield of the tris(oxalato) complex com- initial conditions and rate constants (Table 1). From Fig. 1 can be
pared to the bis(oxalato) complex which is confirmed by this work seen that ˚ = 1.26 is a limiting value for the quantum yield at high
and Faust and Zepp [28]. When the concentration of Fe3+ is low- concentrations. The quantum yield of 1.26 cannot be further max-
ered, even at a constant pH (3 or 5) and a constant Fe(III):oxalate imized by an increase of Fe(III) concentration. Secondary reactions
ratio, the speciation gradually changes with an increasing amount must clearly contribute to this value, because it exceeds unity. But
of bis(oxalato) complex accompanied by a decreasing amount of under the applied experimental conditions the effect of the sec-
tris(oxalato) complex. The bis(oxalato) complex has in fact a higher ondary reactions reached its upper limit, which results into the
quantum yield than the tris(oxalato) complex, thus the gradu- value of 1.26. Therefore, the theoretical value of 1.26 has been used
ally changing speciation toward lower Fe3+ concentrations cannot together with the measured overall extinction at 308 nm to calcu-
explain the decrease of the quantum yield. Because if the changing late the formed Fe2+ and CO2 •− as input for initial concentrations
speciation would be the major cause for a quantum yield change, in the model. Two different scenarios have been used for the ini-
then the quantum yield would be expected to increase, which is not tial photolysis mechanism that affects the definition of the initial
the case. Additionally, the four measurements at actinometry con- model concentrations of Fe2+ and CO2 •− . Mechanism (R5), which
ditions (mM [Fe3+ ], red data points in Fig. 1) show a considerable involves formation of 2 CO2 •− , leads to an initial CO2 •− concentra-
change in the speciation, while the quantum yield is stable around tion obtained with ˚ = 1.26 while the initial [Fe2+ ] = 0. The other
1.25. This leads to the conclusion that the speciation change has no possibility, mechanism (R2), is represented by an initial concentra-
effect and the observed quantum yield decrease with decreasing tion of CO2 •− which is equal to the initial [Fe2+ ] and both are half
[Fe3+ ] is mainly caused by the concentration decrease. the value of the final [Fe2+ ]. This should account for the initial for-
mation of Fe2+ through inner sphere electron transfer. Model A is
3.1.3. Kinetic effect on secondary Fe(III) to Fe(II) reduction very simplistic containing only two reactions, the secondary reduc-
The concentration dependence of the quantum yield can be tion of Fe(III) by CO2 •− and the recombination of CO2 •− as well as
explained by a change of the reaction regime for the secondary mechanism (R5).
reactions of the carboxyl radical, CO2 •− . The formation of CO2 •− Model A provides the best fit of the experimental data. The
from the photolysis of [Fe(III) (C2 O4 )n ]3−2n complexes is described observed trend of the concentration dependence is at least qualita-
by (R1)–(R5). Theoretically, there is the possibility of the formation tively reproduced. Through model B, only the invariable quantum
of two Fe2+ ions per one absorbed photon by any [Fe(III) (C2 O4 )n ]3−2n yields at high [Fe3+ ] are reproduced, but not the decrease with
complex, if there are enough [Fe(III) (C2 O4 )n ]3−2n complexes which concentration decrease. Model B uses the same mechanism (R5)
do not receive light, meaning they remain unphotolyzed and as A, but a published rate constant for (R7) which is four times
can thus serve as reactants for secondary reductions by formed higher than in model A, the rate constant for (R8) is similar to
Author's personal copy

C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49 45

Table 1
Parameters used in the Gepasi model sets to simulate measured Fe2+ quantum yields with secondary radical reactions displayed in Fig. 2.

Model kR7 [M−1 s−1 ] Ref. kR8 [M−1 s−1 ] Ref. HOx /HOx + Fe(II)/(III) reactions Photolysis mechanism
9 9
A 1 × 10 Estimated to lie on 1 × 10 Estimated Excluded (R5)
the lower end of
the range
109 –1010 M−1 s−1
[50] [51]
B 4 × 109 [52] 1.2 × 109 [35] Included (R5)
C 4 × 109 [52] 1.2 × 109 [35] Included (R2)

that used in A. The main factor for the disagreement is the higher reaction product Fe2+ depends also on secondary reactions of the
kR7 , which reveals the high sensitivity of the final [Fe2+ ] toward second reaction product, the carboxyl radical (CO2 •− ). In this case,
changes in (R7). Furthermore, model B includes HOx –iron reac- the decrease of the effective overall Fe2+ quantum yield with an
tions, but their contribution to total Fe2+ formation is below 1% increasing amount of absorbed photons can be attributed to a
and therefore negligible. Model C uses the same rate constants as decrease of the amount of non-photoreduced [Fe(III) (oxalate)n ]3−2n
B, but simulates mechanism (R2) which leads to the poorest agree- complexes available for secondary reactions with CO2 •− . At con-
ment with the observed data among all three models. Concluding ditions with high ferrioxalate concentrations as used in chemical
for the concentration dependence it can be stated that the reac- actinometry ([[Fe(III) (oxalate)n ]3−2n ] ≥ 6 × 10−3 M), the ferrioxalate
tion system modeling strongly highlights the great influence of the system was reported to have Fe2+ quantum yields independent
secondary reactions of the radical species CO2 •− as the main cause of the excitation energy [1], which is of course a prerequisite of a
for the observed concentration dependence. This is a major finding chemical actinometer. According to the author’s knowledge, this
of the present paper and it has to be emphasized that the investi- is the first report of an energy or photon flux dependence of Fe2+
gated concentration dependence has to be taken into account when quantum yields at lower (≤6 × 10−3 M) ferrioxalate concentrations.
ferrioxalate photochemistry at below millimolar initial concentra- Hatchard and Parker [1] stated, that under their specified actinom-
tions is dealt with. etry conditions with [ferrioxalate] ≥ 6 × 10−3 M no quantum yield
dependence on incident photon flux was observed, which would
3.2. Dependence of Fe2+ -quantum yield on amount of initially be a basic requirement for an actinometer. It is important that the
absorbed photons identified energy dependence of ferrioxalate Fe2+ quantum yields
is considered for the interpretation of photolysis experiments with
The amount of absorbed exciting radiation at 308 nm was dilute (<millimolar) initial Fe(III) concentrations.
changed via increase or decrease of the excimer laser energy
output while keeping all other experimental parameters con- 3.3. Speciation dependence of Fe2+ -quantum yield
stant. The ratio [Phabs ]/[Fe3+ ]initial was calculated with the
measured laser energy and the overall extinction coefficient 3.3.1. Laser flash photolysis of bis- and trisoxalatoferrate at 308
of [Fe(III) (oxalate)n ]3−2n complex mixtures in the experimental and 351 nm
solution. From Fig. 3 can be seen that the effective overall Fe2+ Individual extinction coefficients of the bis- and
quantum yields decrease with an increase of the number of tris(oxalato)ferrate complexes have been determined for the
absorbed photons. In a photochemical reaction with only one wavelengths of interest. Fig. 4 shows a set of measured overall
primary step, the quantum yield should not change upon variation molar extinction coefficients at 308 and 351 nm as a function of the
of the amount of excitation radiation over a certain range, because fraction of [Fe(Ox)2 ]− in the solution including a linear regression.
the more molecules are excited, a proportional larger amount Note that the higher degree of scatter at 351 nm is due to a
of reaction product should be formed. As mentioned before, in much smaller scale of the y-axis and because the isosbestic point
the Fe(III)oxalato complex photolysis, the amount of one formed at 362 nm (where [Fe(Ox)2 ]− and [Fe(Ox)2 ]3− have the same molar
extinction) is closer to 351 nm than to 308 nm. Therefore the change
of εi at 351 nm with f[Fe(Ox)2 ]− is minor, because [Fe(Ox)2 ]−
and [Fe(Ox)3 ]3− have very similar values at close distance to the

Fig. 3. Dependence of effective overall Fe2+ quantum yield on the amount of


absorbed photons per initial Fe3+ concentration, pH = 3, [Fe3+ ] = 4.85 × 10−4 M, Fig. 4. Overall extinction coefficients (εtot ) of [Fe(Ox)2 ]− and [Fe(Ox)3 ]3− mixtures
Fe(III):oxalate = 1:10, excimer laser flash, photolysis = 308 nm. at 308 (top) and 351 nm (bottom) as a function of calculated [Fe(Ox)2 ]− fraction.
Author's personal copy

46 C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49

Table 2
Individual extinction coefficients of bis- and tris(oxalato)ferrate complexes at 308
and 351 nm.

Species ε308 nm [l mol−1 cm−1 ] ε351 nm [l mol−1 cm−1 ]

[Fe(Ox)2 ]−
2300 ± 90 1040 ± 30
[Fe(Ox)3 ]3− 2890 ± 40 1120 ± 20

isosbestic point, so that computational and experimental errors


produce a more pronounced scatter a the given scale of the y-axis.
The regression parameters are as follows:

 = 308 nm; εtot = (−595 ± 78) · f[Fe(OX) − + (2891 ± 39);


2]

R = 0.95, n = 27 (7)

Fig. 5. Product of overall quantum yields (˚) and overall extinction coefficients
 = 351 nm; εtot = (−87 ± 29) · f[Fe(OX) − + (1122 ± 15); (εtot ) of [Fe(Ox)2 ]− and [Fe(Ox)3 ]3− mixtures as a function of calculated [Fe(Ox)2 ]−
2]
fraction, laser flash photolysis at 308 and 351 nm, Ar purged (O2 depleted), I = 0.01 M
R = 0.75, n = 30 (8) (NaClO4 ), [Fe3+ ]initial = 4.85 × 10−4 M.

At the intercept of the regression line, where f[Fe(Ox)2 = 0, it fol- ]−


lows that f[Fe(Ox)3 ]3− = 1 and therefore the individual extinction because the detection of photoproduced Fe2+ was in this case per-
coefficient of the tris(oxalato)ferrate complex equals the value of formed well above the quantification limit. The product of overall
the intercept. The individual extinction coefficients calculated with quantum yield and (εtot ) does not seem to be much affected over
the help of Eqs. (7) and (8) are displayed in Table 2, no literature a wide range of changing amounts of bis- and tris(oxalato)ferrate
values are available for these wavelengths. complexes. Only at the left side of the plots in Fig. 5, (εtot × ˚tot )
A similar regression analysis has been done for additional wave- shows a marked decrease, when the content of f[Fe(Ox)2 ]− is
lengths between 290 and 436 nm to obtain the individual spectra smaller than 10%. Thus it can be concluded, that solutions clearly
of the bis- and tris(oxalato)ferrate complexes, shown in Fig. 7. dominated by [Fe(Ox)3 ]3− have a lower quantum yield than
After the individual extinction coefficients of the ferrioxalate com- solutions dominated by [Fe(Ox)2 ]− , which is in agreement with
plexes were determined, the individual quantum yields of the the photochemical behavior of bis- and tris(oxalato)ferrate com-
bis- and tris(oxalato)ferrate complexes were measured in two plexes at 436 nm in continuous lamp irradiation studies ([28]
series at both a higher and a lower initial concentration of Fe3+ . and next section). Because the solutions on the leftmost side of
Because of the observed concentration dependence reported ear- the plot are clearly dominated by [Fe(Ox)3 ]3− with 99% bound
lier, the concentration range of changing quantum yields between iron, these individual quantum yields are assigned to [Fe(Ox)3 ]3− :
5 × 10−6 and 4 × 10−4 M [Fe3+ ] was avoided. One measurement ˚308 nm = 0.93 ± 0.09 and ˚351 nm = 0.88 ± 0.08. The values have
series was performed at constant low initial iron concentration been determined for the first time for these wavelengths.
[Fe3+ ] = 4 × 10−6 M, but levels of photochemically produced Fe2+
were too close to the quantification limit of the phenanthroline 3.3.2. Continuous photolysis of bis- and trisoxalatoferrate with
spectroscopic detection method. Therefore, the effects on the quan- Hg(Xe) lamp at 313, 366 and 436 nm
tum yields through a speciation change were most likely within The speciation dependent photochemistry of bis- and
the magnitude of error introduced by the measurement procedure tris(oxalato)ferrate complexes at 313, 366 and 436 nm was
(data not shown). measured at constant high [Fe3+ ]initial = 4.85 × 10−4 M. Quantum
The other series was performed at higher constant initial iron yields were obtained from the slope of the regression line of a
concentration of [Fe3+ ] = 4.85 × 10−4 M. The fractions of bis- and plot of Fe2+ production vs. time as described in the experimental
tris(oxalato)ferrate complexes in each series were only varied by section. The product (εtot × ˚tot ) was plotted against f[Fe(Ox)2 ]−
changing the pH and the amount of excess oxalate. Taking into together with the measured values of εtot (Fig. 6).
account the observed quantum yield dependence on the amount From the regression lines in Fig. 6, the individual molar extinc-
of absorbed photons, the amount of photons absorbed by the mix- tions were determined for bis- and tris(oxalato)ferrate complexes
ture of ferrioxalate complexes was kept always between 3 and 5% at 313, 366 and 436 nm, which are compared with literature val-
of the initial [Fe3+ ] through adjustment of the excimer laser output ues in the case of 436 nm (Table 3). The individual quantum yields
energy. could not be retrieved by regression analysis for 313 nm, which
The performance of εtot and ˚tot measurements under will be further discussed. Unexpectedly, the individual quantum
variable [Fe(Ox)2 ]− /[Fe(Ox)3 ]3− conditions at a constant yields determination at three different wavelengths led to three
[Fe3+ ]initial = 4.85 × 10−4 M led to the results displayed in Fig. 5. distinct results. The plots for 313, 366 and 436 nm have in common
At both wavelengths the measured εtot correlate well with that the measured total extinction coefficient correlates well with
f[Fe(Ox)2 ]− , which is an indication for a correct speciation calcu- f[Fe(Ox)2 ]− . A straight correlation of (εtot × ˚tot ) with f[Fe(Ox)2 ]−
lation. However, the product εtot × ˚tot displays a curved behavior on the basis of Eq. 6 did however only appear at 436 nm. At 366 nm,
instead of a linear dependence on f[Fe(Ox)2 ]− at both 308 and (εtot × ˚tot ) shows considerable scatter compared to 436 nm, but
351 nm. Speciation dependent measurements of (εtot × ˚tot ) in fer- an increasing trend toward higher f[Fe(Ox)2 ]− is clearly visible.
rioxalate solutions at 436 nm using continuous lamp photolysis It has to be noted that for both 366 and 436 nm (εtot × ˚tot ) and
instead of laser flash photolysis showed a clearly increasing trend εtot are increasing with increasing f[Fe(Ox)2 ]− . The higher scatter
of (εtot × ˚tot ) vs. f[Fe(Ox)2 ]− in a study by Faust and Zepp [28] and at 366 nm is most likely introduced by the close proximity of the
a good linear correlation in this work at 436 nm (see next section). isosbestic point. Thus, the individual quantum yield and extinction
On the contrary, results from continuous lamp photolysis at 313 nm coefficient for [Fe(Ox)2 ]− at 366 nm contains a large error (Table 3).
(next section) show also such unexpected non-linear behavior. For 313 nm, a change of f[Fe(Ox)2 ]− between 0.09 and 0.75 shows
This unexpected result is not caused by experimental errors, only slight variations in (εtot × ˚tot ) but at f[Fe(Ox)2 ]− smaller than
Author's personal copy

C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49 47

Table 3
Individual extinction coefficients and quantum yields from Hg(Xe)-lamp photolysis of bis- and tris(oxalato)ferrate complexes at 313, 366 and 436 nm, this work and literature.

photolysis 313 nm 366 nm 436 nm

[Fe(Ox)2 ] −
a
εi [l mol−1 cm−1 ] 2055 ± 111 753 ± 357 55 ± 9 62 ± 6
b
˚i (≥1.23) 1.17 ± 1.46 1.4 ± 0.40 a 1.0 ± 0.25

[Fe(Ox)3 ]3−
a
εi [l mol−1 cm−1 ] 2663 ± 37 709 ± 10 22 ± 2 24 ± 4
c
˚i 0.12 ± 0.05 0.91 ± 0.09 1.0 ± 0.20a 0.60 ± 0.46
Regression, ε m = −609, b = 2663, R = 0.95 m = 45, b = 709, R = 0.86 m = 33, b = 21, R = 0.99
Regression, ˚ – m = 161, b = 648, R = 0.68 m = 56, b = 21, R = 0.98
a
Ref. [28].
b
Highest ˚ measured in FeOx2 /FeOx3 mixture, therefore lower limit for FeOx2 .
c
Measured at 99% FeOx3 .
m = slope, b = intercept, R = correlation coefficient.

0.09 an abrupt decrease is seen, similar to the observations made


using laser flash photolysis at 308 and 351 nm (previous section).
Because of the unusual low quantum yields at low f[Fe(Ox)2 ]− , sev-
eral measurements from the 313 nm plot have been repeated to
exclude random experimental errors. The repeated experiments
involved complete renewal of all stock solutions and use of another
lamp-monochromator photolysis set-up, identical in construction
to the previous one. The quantum yield of ˚i = 0.12 that was mea-
sured in solutions dominated to 99% by [Fe(Ox)3 ]3− was assigned
to the 1:3 complex. Due to the non-linearity of the data, the indi-
vidual quantum yield could not be determined for the 1:2 complex
by linear regression. Since the highest measured overall quantum
yield in the [Fe(Ox)2 ]− /[Fe(Ox)3 ]3− mixtures was ˚ = 1.23, this val-
ues can be regarded as an lower limit for the [Fe(Ox)2 ]− complex
at 313 nm.
The individual extinction coefficients obtained at 436 nm in this
work agree well with the values published [28]. However, higher
values of individual quantum yields have been obtained in this
work, compared with Faust and Zepp [28]. This difference in indi-
vidual quantum yields is attributed to a difference in the [Fe3+ ]initial
in both studies, because it has been shown in the present work that
the quantum yield can depend on initial [Fe3+ ] (see Fig. 1). Taking
into account that Faust and Zepp’s [28] measurements were made
using lower [Fe3+ ]initial of 1 × 10−4 and 5 × 10−5 M which lie in the
region of decreasing and low overall quantum yields shown in Fig. 1,
the lower individual quantum yields could be readily explained by
the observed concentration dependence in the present work.

3.3.3. Concluding remarks to individual quantum yield analysis


Wavelength dependent individual molar extinctions and quan-
tum yields are summarized in Fig. 7. A major outcome of this work is
to provide new data not only for one wavelength but also to present
and compare measurements from several wavelengths both using
continuous and laser flash irradiation techniques. The individual
quantum yields of 1:2 oxalato complexes are higher than those of
the 1:3 complexes which can be clearly seen from the experimental
data. Furthermore, the individual quantum yields of the 1:3 com-
plexes are almost constant in the considered range between 308
and 436 nm, except for the value at 313 nm.
Because it has been already shown, that the observed overall
quantum yield of ferrioxalate photolysis depends on reactions of
the carboxyl radical and thus strongly on changes in experimental
parameters, tentative interpretations of the unexpected nonlinear
behavior of (εtot × ˚tot ) vs. f[Fe(Ox)2 ]− for 308 nm (laser), 313 nm
Fig. 6. Product of overall quantum yields ˚ and overall extinction coefficients εtot (lamp) and 351 nm (laser) photolysis and the very low quantum
(regression: dashed line) of [Fe(Ox)2 ]− and [Fe(Ox)3 ]3− mixtures and εtot (regres-
yield of [Fe(Ox)3 ]3− at 313 nm in the present study can be made
sion: solid line) as a function of calculated [Fe(Ox)2 ]− fraction, continuous photolysis
with Hg(Xe) lamp at 313 (top), 366 (middle) and 436 nm (bottom), Ar purged (O2
accordingly.
depleted), I = 0.01 M (NaClO4 ), [Fe3+ ]initial = 4.85 × 10−4 M. The conditions for the secondary reaction of the carboxyl radi-
cal are different comparing flash photolysis with continuous lamp
Author's personal copy

48 C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49

ferrioxalate photolysis and added CuSO4 , which remarkably


increased the quantum yield from 1.21 to 1.8 at 365 nm [16]. It was
postulated that Cu2+ ions enhanced the decomposition of the Fe(II)
radical complex and decreased the back electron transfer while
forming Cu+ , which could act as a further reductant of Fe(III) species.
The relevance for the present work is to show that radical complex
reactions can have a large impact, but they are not characterized
well enough to draw conclusions about their influence in solutions
with a changing speciation.

4. Summary and conclusions

The present contribution highlights that the chemistry of the


ferrioxalate system is by far not well understood although numer-
ously investigated by several researchers. It was shown that the
measured overall effective Fe2+ quantum yield of ferrioxalate pho-
tolysis can be a function of initial ferrioxalate concentration in the
range 10−4 M to 10−5 M. The concentration dependence is caused
Fig. 7. Individual quantum yields ˚i and individual extinction coefficients εi for the
[Fe(Ox)3 ]3− and [Fe(Ox)2 ]− complexes as a function of wavelength. by a decrease of the efficiency of the secondary Fe(III) reduction
via the CO2 •− radical and the simultaneous increase of CO2 •− rad-
ical recombination to reform oxalate. It has to be emphasized that
photolysis. In flash photolysis the generation of carboxyl radicals this concentration dependence should be taken into account when
happens in 20 ns which then react instantly with non-photolyzed published quantum yields for ferrioxalate photolysis are used for
Fe(III) producing approximately micro-molar concentrations of the interpretation of systems with dilute concentrations, such as
CO2 •− , which is in the moles per second range. On the contrary, environmental compartments like cloud or surface waters. For the
the production of CO2 •− in continuous lamp photolysis is in the first time it has been shown that in dilute ferrioxalate solutions,
nano-mol per second range, so the ratio of the reactants CO2 •− the measured quantum yields after flash photolysis decreased with
and non-photolyzed Fe(III) differs by many orders of magnitude increasing photolysis light intensity suggesting a changing kinetic
between the experiments, which could possibly cause a differ- regime for secondary reactions of the photofragments. Moreover,
ent overall production of Fe2+ relative to the absorbed photons the Fe2+ quantum yield depends on the contributions from the indi-
given the sensitivity of ferrioxalate photolysis to the experimen- vidual coordination compounds [Fe(Ox)3 ]3− and [Fe(Ox)2 ]− , where
tal conditions. However, because similar curved behavior is seen 1:2 complexes have a higher quantum yield compared to the 1:3
also at 313 nm lamp photolysis, the considerations made above complexes. Ultimately, it can be concluded that a differentiation
might account for differences regarding flash versus continuous between the two different proposed mechanisms (R2) (concerted
photolysis but cannot explain the non-linearity in general. electron transfer) and (R5) (oxalate C-C bond cleavage) is not pos-
It is likely that both [Fe(Ox)2 ]− and [Fe(Ox)3 ]3− have a different sible on the basis of the Fe2+ quantum yield measurements in the
initial photoreactivity. [Fe(Ox)2 ]− and [Fe(Ox)3 ]3− have different present work. The reason can be found in the design of the exper-
redox potentials which would affect the reaction rate of the reduc- iments that measures only the net formed Fe2+ produced by a
ing radical CO2 •− with them and contribute differently to the number of processes, some of which remain poorly characterized.
overall quantum yield. The standard reduction potential E◦ of CO2 •− The benefit of the present work is to highlight that the depend-
equals −1.85 V [48]. The E◦ of solutions containing a lower excess ence of the Fe2+ quantum yield on secondary reactions is very
of oxalate is higher (0.035 V), than in solutions dominated by pronounced, which are sensitive to initial concentration and irra-
[Fe(Ox)3 ]3− where E◦ = 0.005 V [49]. Hence, it follows that the reac- diation intensity. The ferrioxalate photolysis experiments appear
tion of CO2 •− with [Fe(Ox)2 ]− should be favored and proceed faster. similar to radical scavenging experiments, where the obtained yield
This could be one reason for the higher quantum yield of the 1:2 can depend on reactive species, mainly CO2 •− but possibly also
complex. Increased steric hindrance of the carboxyl radical attack Fe(II) radical complexes or excited states of ferrioxalate complexes,
on the metal center caused by one more oxalate ligand in case of and the concentration of the scavenger, which would be ground
the 1:3 complex could be another reason for a lower quantum yield state ferrioxalate complexes. Hopefully, the present paper could
compared to the 1:2 complex. The above mentioned effects of dif- raise some awareness about the complicated interactions of the
ferent initial photoreactivity and secondary reaction of the carboxyl ferrioxalate system and maybe trigger future investigations which
radical could possibly lead to the curved dependence of (εtot × ˚tot ) further aim at the investigation of secondary processes, possibly
vs. f[Fe(Ox)2 ]− observed in flash photolysis at 308 and 351 nm in under consideration of speciation effects.
the present work (Fig. 5) and at 313 nm (Fig. 6).
Possible effects of free excess oxalate on secondary reactions References
were considered for solutions with a high fraction of 1:3 com-
plex. Theoretically, it is improbable that formed carboxyl radicals [1] C.G. Hatchard, C.A. Parker, A new sensitive chemical actinometer. 2. Potas-
could be scavenged by excess oxalate. CO2 •− is a reducing rad- sium ferrioxalate as a standard chemical actinometer, Proceedings of the Royal
Society of London Series A 235 (1956) 518–536.
ical which should be unreactive toward oxalate. A key to fully
[2] P. Ciesla, P. Kocot, P. Mytych, Z. Stasicka, Homogeneous photocatalysis by tran-
understand the ferrioxalate system might lie in the reactions of sition metal complexes in the environment, Journal of Molecular Catalysis A:
the excited state and the formed radical complexes, about which Chemical 224 (2004) 17–33.
[3] L. Deguillaume, M. Leriche, K. Desboeufs, G. Mailhot, C. George, N. Chaumer-
not much information is published. Excited states of [Fe(Ox)3 ]3−*
liac, Transition metals in atmospheric liquid phases: sources, reactivity, and
were reported to undergo quenching by reaction with ground sensitive parameters, Chemical Reviews 105 (2005) 3388–3431.
state [Fe(Ox)3 ]3− (k = (4.5 ± 1.2) × 107 M−1 s−1 ) [11] which could [4] W. Feng, N.S. Deng, Photochemistry of hydrolytic iron (III) species and pho-
also affect the quantum yield of ferrioxalate photolysis. Nothing toinduced degradation of organic compounds. A minireview, Chemosphere 41
(2000) 1137–1147.
was reported about quenching with excess oxalate. The importance [5] G.D. Cooper, B.A. Degraff, Photochemistry of ferrioxalate system, Journal of
of radical complex reactions was shown by an experiment with Physical Chemistry 75 (1971) 2897–2902.
Author's personal copy

C. Weller et al. / Journal of Photochemistry and Photobiology A: Chemistry 255 (2013) 41–49 49

[6] G.D. Cooper, B.A. Degraff, Photochemistry of monoxalatoiron(III) ion, Journal of [28] B.C. Faust, R.G. Zepp, Photochemistry of aqueous iron(III) polycarboxylate
Physical Chemistry 76 (1972) 2618–2625. complexes—roles in the chemistry of atmospheric and surface waters, Envi-
[7] D.C. Doetschman, D.W. Dwyer, K.L. Trojan, A study of ferric trisoxalate ronmental Science and Technology 27 (1993) 2517–2522.
photochemistry in single-crystals by electron-paramagnetic resonance—an [29] E. Fernandez, J.M. Figuera, A. Tobar, Use of the potassium ferrioxalate actinome-
intermediate exchange-coupled ferrous bisoxalate-radical species, Chemical ter below 254 nm, Journal of Photochemistry 11 (1979) 69–71.
Physics 129 (1989) 285–294. [30] H. Funayama, K. Ogiwara, T. Sugawara, H. Ohashi, Dependence of quantum
[8] D.J.E. Ingram, W.G. Hodgson, C.A. Parker, W.T. Rees, Detection of labile yield on concentration in potassium trisoxalatoferrate chemical actinometer,
photochemical free radicals by paramagnetic resonance, Nature 176 (1955) Kagaku Kogaku Ronbunshu 10 (1984) 534–536.
1227–1228. [31] L. Vincze, S. Papp, Individual quantum yields of Fe3+ Oxn 2− Hm + complexes in
[9] R.A. Jamieson, S.P. Perone, Electroanalytical measurements of flash-photolyzed aqueous acidic solutions (Ox2− = (C2 O4 )2− , N = 1–3, M = 0,1), Journal of Photo-
ferrioxalate, Journal of Physical Chemistry 76 (1972) 830–839. chemistry 36 (1987) 289–296.
[10] A.V. Loginov, S.B. Katenin, I.V. Voyakin, G.A. Shagisultanova, Mechanism [32] H. Fadrus, J. Maly, Suppression of iron(III) Interference in determination
of the photolysis of the ferrioxalate ion and the sensitized reduction of iron(II) in water by 1,10-phenanthroline method, Analyst 100 (1975)
of copper, Soviet Journal of Coordination Chemistry 12 (1986) 1621– 549–554.
1626. [33] B.C. Faust, Experimental determination of molar absorptivities and quantum
[11] V. Nadtochenko, J. Kiwi, Dynamics of light-induced excited state quenching yields for individual complexes of a labile metal in dilute solution, Environ-
of ferrioxalate complexes by peroxides. Fast kinetic events and interaction mental Science and Technology 30 (1996) 1919–1922.
with toxic pollutants, Journal of Photochemistry and Photobiology A 99 (1996) [34] F.R. Smith, P. Delahay, Coulostatic study of residual oxygen in electro-
145–153. chemical investigations, Journal of Electroanalytical Chemistry 10 (1965)
[12] C.A. Parker, Induced autoxidation of oxalate in relation to the photolysis 435–450.
of potassium ferrioxalate, Transactions of the Faraday Society 50 (1954) [35] N. Getoff, Peroxyl radicals in the treatment of waste solutions, in: Z.B. Alfassi
1213–1221. (Ed.), Peroxyl Radicals, Wiley, Chichester, 1997.
[13] C.A. Parker, C.G. Hatchard, Photodecomposition of complex oxalates - some [36] W.G. Barb, J.H. Baxendale, P. George, K.R. Hargrave, Reactions of ferrous and
preliminary experiments by flash photolysis, Journal of Physical Chemistry 63 ferric ions with hydrogen peroxide, Nature 163 (1949) 692–694.
(1959) 22–26. [37] B.H.J. Bielski, D.E. Cabelli, R.L. Arudi, A.B. Ross, Reactivity of HO2 /O2 − radicals in
[14] J.I.H. Patterson, S.P. Perone, Spectrophotometric and electrochemical studies of aqueous-solution, Journal of Physical and Chemical Reference Data 14 (1985)
flash-photolyzed trioxalatoferrate(III), Journal of Physical Chemistry 77 (1973) 1041–1100.
2437–2440. [38] H. Christensen, K. Sehested, Pulse-radiolysis at high-temperatures and high-
[15] D. Rehorek, M. Benedix, P. Thomas, Photocatalytic systems. 13. Detection of pressures, Radiation Physics and Chemistry 18 (1981) 723–731.
oxalate free-radicals by EPR spin trapping in photolysis of tris(oxalato)iron(III), [39] H. Christensen, K. Sehested, E. Bjergbakke, Radiolysis of reactor water: reaction
Inorganica Chimica Acta 25 (1977) L100. of hydroxyl radicals with superoxide (O2 − ), Water Chemistry of Nuclear Reactor
[16] D. Rehorek, H. Grikos, R. Billing, Über den Einfluss von Kupfer(II)-Ionen auf die Systems 5 (1989) 141–144.
Photoreduktion von Tris(oxalato)ferrat(III), Zeitschrift fur Chemie 30 (1990) [40] H. Christensen, K. Sehested, H. Corfitzen, Reactions of hydroxyl radicals with
378–379. hydrogen-peroxide at ambient and elevated-temperatures, Journal of Physical
[17] I.P. Pozdnyakov, O.V. Kel, V.F. Plyusnin, V.P. Grivin, N.M. Bazhin, New insight Chemistry 86 (1982) 1588–1590.
into photochemistry of ferrioxalate, Journal of Physical Chemistry A 112 (2008) [41] A.J. Elliot, G.V. Buxton, Temperature-dependence of the reactions OH + O2 − and
8316–8322. OH + HO2 in water up to 200 ◦ C, Journal of the Chemical Society, Faraday Trans-
[18] Q.G. Mulazzani, M. Dangelantonio, M. Venturi, M.Z. Hoffman, M.A.J. Rodgers, actions 88 (1992) 2465–2470.
Interaction of formate and oxalate ions with radiation-generated radicals in [42] G.G. Jayson, A.J. Swallow, B.J. Parsons, Oxidation of ferrous ions by hydroxyl
aqueous-solution—methylviologen as a mechanistic probe, Journal of Physical radicals, Journal of Chemical Society, Perkin Transactions 1 (68) (1972)
Chemistry 90 (1986) 5347–5352. 2053–2058.
[19] J. Chen, H. Zhang, I.V. Tomov, M. Wolfsherg, X.L. Ding, P.M. Rentzepis, Transient [43] R. Losno, Trace metals acting as catalysts in a marine cloud: a box model study,
structures and kinetics of the ferrioxalate redox reaction studied by time- Physics and Chemistry of the Earth Part B-Hydrology Oceans and Atmosphere
resolved EXAFS, optical spectroscopy, and DFT, Journal of Physical Chemistry 24 (1999) 281–286.
A 111 (2007) 9326–9335. [44] J.D. Rush, B.H.J. Bielski, Pulse radiolytic studies of the reactions of HO2 /O2 − with
[20] J. Chen, A.S. Dvornikov, P.M. Rentzepis, Comment on “new insight into pho- Fe(II)/Fe(III) ions—the reactivity of HO2 /O2 − with ferric ions and its implication
tochemistry of ferrioxalate”, Journal of Physical Chemistry A 113 (2009) on the occurrence of the Haber–Weiss Reaction, Journal of Physical Chemistry
8818–8819. 89 (1985) 5062–5066.
[21] I.P. Pozdnyakov, O.V. Kel, V.F. Plyusnin, V.P. Grivin, N.M. Bazhin, Reply to “Com- [45] N.F. Tyupalo, Y.A. Dneprovskii, Studying the reactions of ozone with iron(II) ions
ment on ‘New insight into photochemistry of ferrioxalate”’, Journal of Physical in aqueous-solutions, Zhurnal Neorganicheskoi Khimii 26 (1981) 664–667.
Chemistry A 113 (2009) 8820–8822. [46] J. Ziajka, F. Beer, P. Warneck, Iron-catalyzed oxidation of bisulfite aqueous-
[22] J.N. Demas, W.D. Bowman, E.F. Zalewski, R.A. Velapoldi, Determination solution—evidence for a free-radical chain mechanism, Atmospheric Environ-
of the quantum yield of the ferrioxalate actinometer with electrically ment 28 (1994) 2549–2552.
calibrated radiometers, Journal of Physical Chemistry 85 (1981) 2766– [47] P. Mendes, Biochemistry by numbers: simulation of biochemical pathways with
2771. Gepasi 3, Trends in Biochemical Sciences 22 (1997) 361–363.
[23] S. Goldstein, J. Rabani, The ferrioxalate and iodide–iodate actinometers in [48] P.S. Surdhar, S.P. Mezyk, D.A. Armstrong, Reduction potential of the CO2 −
the UV region, Journal of Photochemistry and Photobiology A 193 (2008) radical-anion in aqueous-solutions, Journal of Physical Chemistry 93 (1989)
50–55. 3360–3363.
[24] J. Lee, H.H. Seliger, Quantum yield of ferrioxalate actinometer, Journal of Chem- [49] J.A. Bard, R. Parsons, J. Jordan, Standard Potentials in Aqueous Solution, IUPAC,
ical Physics 40 (1964), 519-&. New York, 1985.
[25] D.E. Nicodem, O.M.V. Aquilera, Standardization of the potassium ferrioxalate [50] P. Neta, R.E. Huie, A.B. Ross, Rate constants for reactions of inorganic radicals in
actinometer over the temperature-range 5–80 ◦ C, Journal of Photochemistry aqueous-solution, Journal of Physical and Chemical Reference Data 17 (1988)
21 (1983) 189–193. 1027–1284.
[26] H.B. Abrahamson, A.B. Rezvani, J.G. Brushmiller, Photochemical spectroscopic [51] K.A. Hislop, J.R. Bolton, The photochemical generation of hydroxyl radicals in
studies of complexes of iron(III) with citric-acid and other carboxylic-acids, the UV–vis/ferrioxalate/H2 O2 system, Environmental Science and Technology
Inorganica Chimica Acta 226 (1994) 117–127. 33 (1999) 3119–3126.
[27] G.G. Duka, D.G. Batyr, L.S. Romanchuk, A.Y. Sychev, Photochemical transfor- [52] J. Chen, H. Zhang, I.V. Tomov, P.M. Rentzepis, Electron transfer mechanism and
mation of hydroxy acids in the presence of iron(III) ions, Soviet Journal of photochemistry of ferrioxalate induced by excitation in the charge transfer
Coordination Chemistry 16 (1990) 93–106. band, Inorganic Chemistry 47 (2008) 2024–2032.

You might also like