Photoinduced reduction of chromium(VI)

PAPER

www.rsc.org/pps | Photochemical & Photobiological Sciences

Photoinduced reduction of chromium(VI) by iron aminopolycarboxylate complex (FeNTA) Otman Abida,*†a Gilles Mailhot,a,b Hana Mestankova,a Marta Litterc and Michèle Boltea,b Received 22nd October 2009, Accepted 6th May 2010 First published as an Advance Article on the web 13th May 2010 DOI: 10.1039/b9pp00138g Chromium(VI) reduction photoinduced by iron(III) nitrilotriacetate (FeNTA) was investigated under monochromatic excitation. 313 nm was used as the irradiation wavelength in order to minimize the absorption of the light by chromium(VI): 91% of the photons were absorbed by FeNTA at pH = 7.0. Quantum yields of FeNTA and chromium(VI) disappearance and iron(II) formation were measured at pH 2.0, 4.0 and 7.0. In all the cases, chromium(VI) reduction follows first order kinetics with maximum efficiency at pH 2.0. The observed rate constant is proportional to FeNTA concentrations up to a maximum of 6.0 ¥ 10-4 mol L-1 . The effect of oxygen was also investigated. If there is no large pH effect and no effect of oxygen on the quantum yields of chromium(VI) and FeNTA disappearance, the pH strongly influences the nature of the reduced chromium species. At pH 2.0, only chromium(III) was detected, whereas at pH 4.0 and 7.0 no chromium(III) resulting from chromium(VI) reduction was observed. Chromium(V) is supposed to be formed and stabilized by the chelating groups of NTA or NTA photoproducts.

Introduction The presence of heavy metals such as chromium(VI) in aquatic bodies is a problem because it represents a threat to aquatic and human life.1 Chromium(VI) is a highly oxidizing, soluble, carcinogenic, mutagenic and toxic form of the metal that is produced as an effluent from metal plating, tanning, paper making and the other industries.2,3 The worldwide annual mining of chromite (FeCr2 O4 ) has exceeded a level of 10 million tons3 and, as a result of the extensive use, waste disposal and natural water contamination has become an important environmental problems.1,4 However, chromium(VI) itself is unreactive toward DNA and it has been shown that chromium(V) generated by reducing agents is the actual DNA damaging agent. Next to chromium(VI), chromium(III) is the other chromium oxidation state found in the environment. Chromium(III) is considered non-toxic in most forms and is even essential to human life. Chromium(III) tends to associate with solid phases and is quite immobile in the environment. Accordingly, processes leading to the reduction of chromium(VI) to chromium(III) are of great interest from an environmental point of view. The photochemical reduction of chromium(VI) might be useful in remediation procedures and is likely to be important in a Clermont Université, Université Blaise Pascal, Laboratoire de Photochimie Moléculaire et Macromoléculaire, BP 10448, F-63000, Clermont-Ferrand, France b CNRS, UMR 6505, Laboratoire de Photochimie Moléculaire et Macromoléculaire, F-63177, Aubière, France c Unidad de Actividad Qu´ımica, Centro Atómico Constituyentes, Comisión Nacional de Energ´ıa Atómica, Av. Gral. Paz 1499, 1650, Buenos Aires, Argentina † Present address: Department of Chemistry, University of Calgary, 2500 University Drive NW, Calgary, AB, T2N 1N4, Canada. E-mail: [email protected]. Tel: +1-403-220-8815.

natural aquatic systems where Fe(II,III) and organic compounds are present, particularly for slightly acidic and neutral surface waters (e.g., in contaminated wetlands) and in atmospheric cloud and fog droplets. Therefore, the interaction of chromium(VI) with iron(II), largely present in the surface waters, under visible light or solar irradiation has been described in the literature.5 In addition, iron(III) carboxylate complexes such as oxalate and citrate5 or natural organic matter (NOM), were described as efficient photoinducers of chromium(VI) reduction.6 However, photochemical production of Fe(II) from Fe(III) organic complexes is accompanied by the production of HO2 ∑ /O2 ∑ - , H2 O2 , and the other reactive oxygen species that together reoxidize Fe(II).7 The net results are photochemical Fe(II,III) cycles in which steady state concentrations of Fe(II), Fe(III), HO2 ∑ /O2 ∑ - , and H2 O2 are established, and the complexing organic ligands are ultimately oxidized.7,8 In addition, strong Fe(III)-complexing ligands shift the steady state equilibrium to Fe(III), thereby lowering the E H for Fe(II) oxidation at a given pH (Fe(II) becomes a stronger reductant). Fe(II,III) cycling in the presence of oxalate was described by Sedlak and Hoigne.9 Reduction of chromium(VI) by Fe(II) is thus in competition with oxidation of Fe(II) by O2 and other oxygen species. While rate laws for reduction of chromium(VI) by Fe(II) between pH 1 and 2 have been determined.10 The photodegradation of FeNTA, the potential photoinducer of chromium(VI) reduction, was previously investigated in our laboratory.11 It was shown to undergo a very fast photoredox process, the efficiency of which depends on FeNTA speciation and on the irradiation wavelength.11 The major photoproducts were Fe(II), CO2 , HCHO and iminodiacetic acid. Both components, FeNTA and chromium(VI) undergo protolytic equilibrium according to and

This journal is © The Royal Society of Chemistry and Owner Societies 2010

Photochem. Photobiol. Sci., 2010, 9, 823–829 | 823

off and dried overnight at 60 ◦ C. The elemental microanalysis results were as follows: experimental C 26.46, H 3.48, N 5.12, Fe 20.55%, calculated for FeC6 H10 O8 N, C 25.7, H 3.6, N 5.0, Fe 20.0%. The elementary analysis confirmed the formula of the ferric salt Fe[N(CH2 CO2 )3 ]·2H2 O; the analysis gave evidence for a 1 : 1 stoichiometry. Irradiation setup The dimer Cr2 O7 2- formed from HCrO4 2HCrO4 - → Cr2 O7 2- + H2 O, K d = 48 mol L-1 was negligible under our experimental conditions: [FeNTA] was in the range 1.0 ¥ 10-4 – 9.0 ¥ 10-4 M and [chromium(VI)] was kept constant and equal to 2.0 ¥ 10-4 M. At pH 2.0 and 4.0 FeNTA and HCrO4 - are present, whereas at pH 7.0 the monoanionic form of FeNTA together with 70% of CrO4 2- and 30% of HCrO4 - are present in aqueous solution. In this paper, the results concerning chromium(VI) reduction photoinduced by Fe(III)-nitrilotriacetic acid complex (FeNTA), an iron salt considered as a suitable model of iron complexes with natural organic matter, are reported. Special emphasis was given to the mechanistic aspects including speciation, monitoring of the intermediates and the final oxidation state of chromium.

In order to measure the quantum yield, monochromatic irradiations were carried out with a high-pressure polychromatic mercury-vapor lamp (Osram HBO 200 W) equipped with a grating monochromator (Bausch and Lomb) providing a collinear and homogenous beam. The photoreactor was a cylindrical quartz cell with an internal diameter of 1.85 cm and an internal length of 2.0 cm, the light beam passing along the axis. The light intensity (I 0 ) was measured by ferrioxalate actinometry:12 I 0(365 nm) = 1.30 ¥ 1015 photons s-1 cm-2 and I 0(313 nm) = 3.9 ¥ 1014 photons s-1 cm-2 . The calculation of the initial quantum yields was performed at conversion percentage lower than 10%. The spectra of the different species involved in the system (chromium(VI) and FeNTA) are presented in Fig. 1. The fraction of photons absorbed (Fa ) by each partner was calculated when considering I a /I 0 of the mixture.

Experimental Materials All reagents were of the purest commercially available grade and used without further purification. Potassium dichromate >99% and nitrilotriacetic acid (NTA) 99% were Merck and Aldrich products respectively. Ferric perchlorate nonahydrate (Fe(ClO4 )3 ·9H2 O, > 97%) was a Fluka product kept in a desiccator. The pH of the solutions was measured with a pH-meter (JENWAY 3310) and a micro-electrode (Fisherbrand W 84908, 8 ¥ 115 mm) filled with liquid electrolyte of KCl 4 M saturated with AgCl. The calibration is performed each day with pH buffer at 4.01, 7.01 and 10.01. The precision of the measurements was 0.1 units. For the initial pH adjustments, diluted solution of HCl acid 0.1 M and NaOH 0.1 M were used. The solutions were made up with deionised water (Milli-Q) in equilibrium with air, saturated with oxygen or deaerated by bubbling with argon or nitrogen for 30 min at 22 ◦ C. For prolonged irradiations, the bubbling with nitrogen or oxygen was maintained all along the experiments. Some experiments were carried out in the dark. We checked the concentrations of both, Fe(III)-NTA and chromium(VI) by HPLC and UV-visible spectrophotometry, which were stable with time (7 days) in aqueous solution and were not affected by the bubbling of nitrogen or oxygen. The ionic strength of the solution was not controlled. Fe(III)-NTA was prepared as solid salt by the following method. A weighed sample of NTA, corresponding to a 0.2 M solution was dispersed in H2 O. The dissolution of NTA was then obtained by addition of NaOH and was complete at pH 4.0. The concentration of the aqueous solution Fe(ClO4 )3 was 0.67 M. Then 100 cm3 of NTA solution, 30 cm3 of Fe(III) solution and 70 cm3 of buffer (0.05 M potassium acetate solution, adjusted to pH 4.0 with acetic acid) were then mixed. The so obtained precipitate was filtered 824 | Photochem. Photobiol. Sci., 2010, 9, 823–829

313 nm appears to be the most convenient irradiation wavelength, present in the mercury lamp, to optimise the absorption of the photons by FeNTA. In addition, the concentration of chromium(VI) in the environment is far lower than the one used in this work and 313 nm is present in the solar emission and energetic enough to induce the photoredox process. The percentage of photons absorbed, at 313 nm, by chromium(VI) and FeNTA are summarized in Table 1.

Fig. 1 UV-visible absorption spectrum of FeNTA (0.3 mM) and chromium(VI) (0.2 mM) (1 cm optical path length) at different initial pH values.

For the kinetic studies concerning chromium(VI) and iron complexes, an experimental setup with larger volume (100 mL) and more powerful source was used. The batch mode photoreactor was a cylindrical stainless steel container in which six fluorescent


tubes (DUKE Sunlamp FL 20 W) are located around a Pyrex reactor (2.8 cm in diameter) containing the sample. These lamps emitted radiation as a broad band (l max = 310 nm) extending from 280 to 380 nm; wavelengths shorter than 300 nm were cut off by the wall of Pyrex tube. The photonic flux was determined by the ferrioxalate method.12 The unit delivered an intensity I a ª 4.8 ¥ 1015 photons s-1 .cm-3 over a large volume (100 mL). The reaction medium was continuously stirred. Experiments in field conditions were performed in a Pyrex cylindrical reactor (100 mL) on a sunny day in July in ClermontFerrand (latitude 46◦ N, 400 m above sea level). UV-vis spectra were recorded on a CARY 3 and CARY 300 SCAN double beam spectrophotometer (Varian), absorption precision ±0.002.

Chemical analysis Fe(II) concentration was determined by complexometry with ortho-phenanthroline, using e510 nm = 1.118 ¥ 104 M-1 cm-1 for the Fe(II)-phenanthroline complex.12 Chromium(VI) concentration was determined by UV-visible spectrophotometry titration with the diphenylcarbazide (DPC).13 Chromium(III) concentration was determined by spectrofluorimetry titration with o-vanillin-8-aminoquinoline (OVAQ).14

Transformation of chromium(VI) photo-induced by FeNTA complex The UV-visible spectra of the mixture FeNTA-chromium(VI) (3 ¥ 10-4 M – 2 ¥ 10-4 M) at neutral pH (7.0) and at pH 2.0 as a function of the irradiation time at 313 nm (monochromatic irradiation, see Experimental) are presented in Fig. 2a and 2b respectively. At pH 7.0, a drastic spectral evolution is observed with a complete disappearance of the two maxima at 262 and 370 nm and the presence of three well-defined isosbestic points at 304, 345 and 406 nm. These spectral modifications correspond to the disappearance of FeNTA and the transformation of chromium(VI) with the irradiation time. Considering that FeNTA has no absorption at wavelengths higher than 380 nm, the isosbestic point at 406 nm can be attributed to the lone transformation of chromium(VI) into another chromium species. Moreover, in the differences of the spectra (T irr - T 0 ), an absorption band with a maximum at 460 nm clearly appeared (insert of Fig. 2a). For the same mixture at pH 2.0, a fast disappearance of the band at

EPR measurements The EPR spectra of the different chromium species formed during the photochemical process were recorded at room temperature with a Bruker ESP 300 spectrometer, with a microwave frequency of 9.64 GHz and a modulation frequency of 100 kHz. The irradiations in situ were performed in the EPR cavity using xenon mercury Hanovia lamp and the convenient cut-off filter.

Results The FeNTA UV-visible spectrum is not affected by the protolytic equilibrium whereas the spectrum of a chromium(VI) aqueous solution is strongly pH dependent (Fig. 1). The mixture of FeNTA and chromium(VI) was thermally stable (in the dark and at room temperature). At pH = 7.0, a slight evolution of the UV-visible spectrum appeared due to the slow phenomenon of FeNTA decomplexation: an abatement of 10% was observed after 10 h and no more decomposition was observed for longer time (a few days). This decomplexation was also observed in the absence of chromium(VI). At pH 2.0 or 4.0 the UV-visible spectrum was stable. No significant disappearance of FeNTA and chromium(VI) was observed in the dark.

Table 1 Percentage of absorbed photons by each species in the mixture of FeNTA (3 ¥ 10-4 mol L-1 ) and chromium(VI) (2 ¥ 10-4 mol L-1 ) at different pH, l irr = 313 nm Initial pH % of absorbed photons

FeNTA Cr(VI)

2.0

4.0

5.5

7.0

73 27

73 27

80 20

91 9

Fig. 2 UV-visible absorption spectrum of the mixture FeNTA (0.3 mM) and chromium(VI) (0.2 mM) as a function of irradiation time at 313 nm at (a) initial pH = 7.0; insert: difference of the spectrum (T irr - T 0 ) and (b) initial pH = 2.0; insert: zoom of the spectrum between 520 and 660 nm.



250 nm is observed. And an absorption band centred at 575 nm is observed with the irradiation time corresponding to the formation of chromium(III) (insert of Fig. 2b). It is worth mentioning that the molar extinction coefficient of the bound at 575 nm calculated by assuming chromium(VI) to be entirely reduced into chromium(III) (ecalc. ª 13 M-1 cm-1 ) is consistent with the value reported in the literature for Cr(H2 O)6 3+ .15,16 At pH 4.0, the spectral evolution is more complex likely combining both evolutions. The kinetics of chromium(VI) disappearance were followed by complexometry with diphenylcarbazide (DPC) at pH 7.0 and l irr of 313 nm (Fig. 3). The effect of both FeNTA and light on the transformation of chromium(VI) is confirmed on this graph. Indeed, the degradation of chromium(VI) was negligible when either light or FeNTA was omitted in the system. On the contrary, a fast decrease of chromium(VI) concentration is observed during the first hour of irradiation (75% of disappearance) when FeNTA and light are present. Moreover, the photochemical disappearance of chromium(VI) is faster and more important when the concentration of FeNTA increased and reached a plateau value at FeNTA concentrations higher than 0.6 mM (Fig. 3).

Fig. 4 Kinetics of (a) chromium(VI) (0.2 mM) disappearance and (b) Fe(II) formation at different initial pH values (2.0, 4.0, 5.5 and 7.0). [FeNTA] = 0.3 mM, l irradiation = 313 nm.

Fig. 3 Kinetics of chromium(VI) (0.2 mM) photodegradation in the absence or presence of FeNTA 0.1, 0.3 (with and without light), 0.6 and 0.9 mM. l irr = 313 nm, initial pH = 7.0.

chromium(VI) remains in solution after 240 min of irradiation. It can be attributed to the far less oxidative feature of CrO4 2when compared to HCrO4 - .16 Moreover, in Table 2 the initial rate constants of chromium(VI) as a function of initial pH value are presented. The first order apparent rate constants increase when the pH decreased. This effect of the initial pH was not observed when dealing with the degradation of an organic pollutant like 4-chlorophenol photoinduced by FeNTA.17 The formation of Fe(II) is only observed at initial pH 2.0, very fast during the first 20 min of irradiation and reached a pseudo plateau value near 2.6 ¥ 10-4 M (Fig. 4b), and then decreases for longer irradiation times. The observation that no Fe(II) is detected at the other initial pH values, gives evidence for an efficient reoxidation process of iron in such a system.

Effect of the pH The photochemical reaction (313 nm) in the presence of FeNTA (3.0 ¥ 10-4 M) and chromium(VI) (2.0 ¥ 10-4 M) was followed at different initial pH (2.0, 4.0, 5.5 and 7.0). The kinetics of chromium(VI) disappearance and Fe(II) formation are present in Fig. 4a and 4b respectively. The disappearance of chromium(VI) is very fast and complete for pH values between 2.0 and 5.5. At pH 7.0 the transformation is much slower, but only 15% of 826 | Photochem. Photobiol. Sci., 2010, 9, 823–829

Table 2 First order apparent rate constants for chromium(VI) disappearance at different pH values under solar light or at 313 nm. [chromium(VI)] = 2 ¥ 10-4 M and [FeNTA] = 3 ¥ 10-4 M Initial pH

2.0

4.0

5.5

7.0

kdis ¥ 10-4 (s-1 ) (313 nm) kdis ¥ 10-4 (s-1 ) (solar light)

52 37

27 10

20 7

7.3 3


Table 3 Quantum yields of chromium(VI) disappearance in oxygen free and aerated solution at pH 2.0 and 7.0. [chromium(VI)] = 2 ¥ 10-4 M and [FeNTA] = 3 ¥ 10-4 M, l irr = 313 nm Initial pH

fchromium(VI) , aerated

fchromium(VI) , oxygen free

2.0 7.0

0.15 0.09

0.15 0.08

but with a major difference regarding the efficiency of the photochemical reaction. For instance, only 5 min were needed to completely eliminate chromium(VI) from a deaerated solution. The formation of iron(II), a species only present at pH 2 in aerated solution, was negatively affected by oxygen: the value at the plateau, maximum in the absence of oxygen (3.0 ¥ 10-4 M) which corresponded to the starting iron(III) concentration, decreased to 2.6 ¥ 10-4 and 1.0 ¥ 10-4 M in aerated and oxygen saturated solutions respectively.

Because we are facing various redox or photoredox processes, a pH evolution might be suspected in such systems. H+ consumption increased when the starting pH was decreased from 7.0 (no H+ consumption) to 5.5 and to 4.0. In the last two cases a plateau value was obtained after 20 min of irradiation, period of time directly correlated to the observations made on chromium(VI) disappearance and Fe(II) formation. After 20 min of irradiation the pH increased from 4.0 to 6.5 and 5.5 to 7.2 corresponding to a H+ consumption of 10-4 and 3 ¥ 10-6 M respectively (results not shown).

Isopropanol is known to be an efficient trap for radicals, so chromium(VI) disappearance and Fe(II) formation was studied in the presence of isopropanol (0.01 M). At pH 2.0, chromium(VI) disappearance and Fe(II) formation were both accelerated by the presence of isopropanol whereas at pH 7.0 only a very weak acceleration was observed for chromium(VI) disappearance.

Oxygen effect

Chromium speciation

Oxygen effect was studied at the two extreme pH values 2.0 and 7.0. The effect of oxygen concentration on the quantum yields of chromium(VI) disappearance is presented in Table 3. Contrary to 4-CP disappearance photoinduced by FeNTA complexes, which is strongly inhibited in the absence of oxygen,17 chromium(VI) disappearance is not significantly affected by oxygen concentration. The initial quantum yields of chromium(VI) disappearance are similar in the presence or in the absence of oxygen. The kinetics of chromium(VI) disappearance in oxygen free, aerated and oxygen saturated solution at pH 7 are presented in Fig. 5. The initial rate of chromium(VI) disappearance is only slightly higher in oxygen saturated solution. However, in the second part of the kinetics, the disappearance of chromium(VI) slows down and stops in oxygen saturated as well as in aerated solution while chromium(VI) disappearance is achieved after 100 min of irradiation in the absence of oxygen. Analogous results were obtained at pH 2

EPR spectroscopy upon irradiation is a powerful technique to detect and differentiate the various chromium species susceptible to be involved in the process. Chromium(VI) with electronic structure d0 is diamagnetic (not detectable), chromium(V) (d1 ) is paramagnetic with a strong and narrow signal, chromium(IV) (d2 ) is paramagnetic only at very low temperature and chromium(III) (d3 ) presents a weak and broad signal. EPR spectrum of mixtures of FeNTA (3 ¥ 10-4 M) and chromium(VI) (2 ¥ 10-4 M) were recorded at different pH values upon irradiation (l > 300 nm). At pH 2.0, chromium(III), with a maximum at 575 nm, was unambiguously detected by UV-visible spectrophotometry as already described (Fig. 2b). Characteristic signals of chromium(V)16,18–20 were detected neither at t = 0 nor upon irradiation. The failure to detect any chromium(V) intermediate by EPR spectroscopy was surprising enough when related to a system in which chromium(VI) was involved. This was true at any pH. Actually, chromium(V) signals of thermal and photochemical origin were detected in a previously investigated dichromated system.16,20 At pH 7.0 where the spectral evolution of the system was straightforward (Fig. 2a), chromium(III) formation was monitored by OVAQ complexometric method.14 The results clearly show the absence of chromium(III) during the photochemical process. The disappearance of chromium(VI) together with the absence of chromium(III) formation led us to assign the spectral evolution to the formation of chromium(V) strongly complexed by the starting NTA or its photoproducts. The formation of chromium(V) was also supported by analogy with the spectral evolution of DCPVA(dichromated polyvinylalcohol) films (pH 5.5) with the presence of an isosbestic point at 410 nm and a system in which chromium(V) appeared to be complexed by the functional group present on the macromolecular chain.19 The formation of chromium(IV) in such systems appears to be exceptional. It was observed when dealing with unsaturated compounds such as tryptophan or methacrylamide allowing the addition of chromate anion on the double bound.15,21 Accordingly, the absorbance at 460 and 575 nm, measured from Fig. 2a, (there is no absorbance at t = 0) were plotted as a function of disappeared chromium(VI) (Fig. 6). The linear correlation between the two sets of parameters indicates

Fig. 5 Kinetics of chromium(VI) (0.2 mM) disappearance in aerated, oxygen-free and oxygen saturated solutions. [FeNTA] = 0.3 mM, lirr = 313 nm, initial pH = 7.0.

Isopropanol effect



true solar light in terms of both, the mechanism and the efficiency of the reaction.

Discussion

Fig. 6 Evolution of the absorbance at 460 and 575 nm as a function of chromium(VI) disappearance during the photochemical process. [FeNTA] = 0.3 mM, [Cr(VI)] = 0.2 mM, l irr = 313 nm and initial pH = 7.0.

the formation and the stabilization of a unique reduced chromium species likely chromium(V). It was then possible to calculate the chromium(V) extinction coefficient values (e): they were evaluated to 630 and 159 M-1 cm-1 at 460 and 575 nm respectively. Solar irradiations After investigating the capability of FeNTA to photoinduce (l irr = 313 nm) chromium(VI) reduction, experiments were carried out with solar light. The solar irradiation results are reported in Fig. 7 with the linearisation corresponding to first order kinetics. The apparent k values are summarized in Table 2. It is worth noting that our irradiation set up with artificial light is representative of

In previous work carried out on FeNTA11 or focused on 4-chlorophenol (4-CP) degradation photoinduced by FeNTA,17 the reactive species arising from FeNTA in the excited state were clearly identified. The first step was the intramolecular photoredox process between iron and the ligand. At long wavelength (l > 345 nm) the reaction involves the carboxylate group to form radical RCO2 ∑ together with Fe(II). A sequence of reactions leads to the formation of ∑ OH radical, which is the species responsible for pollutant degradation.17 At shorter wavelength (l > 310 nm) the radical CO3 ∑ - was identified as a major species formed in the photoredox process and responsible for the 4-CP degradation. In that case, the first step is the photoredox process between Fe(III) and a water molecule in coordination around Fe(III).17 According to these previous studies the following radicals are photogenerated at 313 nm: FeNTA + hn → Fe(II) + NTA∑ , ∑ OH, CO3 ∑ - , O2 ∑ - /HO2 ∑

(1)

The formation of CO3 ∑ - , ∑ OH or O2 ∑ - requires the presence of oxygen. When dealing with chromium(VI), its reduction observed under our experimental conditions can be due to various reactions: Cr(VI) + Fe(II) → Cr(V) + Fe(III)

(2)

Cr(VI) + NTA∑ → Cr(V) + NTAox

(3)

Cr(VI) + O2 ∑ - → Cr(V) + O2

(4)

Our results clearly demonstrated that oxygen does not play a crucial role in chromium(VI) reduction. The only reducing species whose formation does not involve oxygen are Fe(II) and the primary organic radical arising from the redox process in FeNTA in the excited state. Therefore, the main reactions responsible of chromium(VI) reduction are the reaction with Fe(II) (reaction (2)) and with the first radical formed on NTA (reaction (3)). Both correspond to an electron transfer. Furthermore, no significant effect was observed in the presence of isopropanol, an efficient trap for hydroxyl radical, gives emphasis to the role played by Fe(II). This conclusion is in agreement with Buerge and Hug’s work related to the reduction of chromium(VI) by iron(II).22 But contrary to what was obtained in such systems, a nice specific spectral evolution was observed when dealing with iron(III) complexes upon irradiation. Even though there is no significant oxygen influence on chromium(VI) reduction, a minor involvement of CO3 - ∑ and ∑ OH, leading to the reoxidation of the metallic cation, cannot be ruled out.

Fig. 7 Kinetics of chromium(VI) disappearance at different initial pH values (2.0, 4.0, 5.5 and 7.0) under solar irradiation.

828 | Photochem. Photobiol. Sci., 2010, 9, 823–829

Fe(II) + CO3 ∑ - → Fe(III) + CO3 2-

(5)

Fe(II) + ∑ OH → Fe(III) + OH-

(6)

Cr(V) + CO3 ∑ - → Cr(VI) + CO3 2-

(7)


Cr(V) + ∑ OH → Cr(VI) + OH-

(8)

However, with FeNTA, alone or in the presence of 4-CP, and whatever the pH Fe(II) formation is observed and the reoxidation of Fe(II) was shown to be negligible.17 The absence of Fe(II) in the investigated system has to be directly correlated to the presence of chromium(VI). The quantum yield, evaluated by monochromatic irradiation (l = 313 nm), of chromium(VI) disappearance (0.09 at pH = 7.0) is independent of oxygen concentration and is very similar to that of FeNTA disappearance (0.11 at pH = 6.0) and of Fe(II) formation (0.09 at pH = 6.0), evaluated under the same conditions, suggesting the low involvement of chromium(V) oxidation by CO3 ∑ - and/or ∑ OH radicals.17 The results show that chromium(VI), a very strong oxidative species, reacts faster with Fe(II) and with NTA∑ than with reactive oxygen species like superoxide anion (O2 ∑ - ). Reaction (2), widely reported in the literature,5,23 was put forward as being of major importance in the biogeochemical cycles of iron and chromium. The importance of this reaction is also supported by the results that no Fe(II) is detected in solution at pH between 4.0 and 7.0. The pre-eminence of reactions (2) and (3) is true at any pH but in more acidic solution (pH 2.0) chromium(V) is quite unstable and is quickly reduced into the chromium(III), as already reported when dealing with dichromated system.16 Chromium(III) formation can result from the stepwise reduction of chromium(V) by the organic radicals and/or Fe(II) analogous to those described for chromium(VI). Cr(V) + Fe(II) →→ Cr(III) … + Fe(III) Cr(V) + NTA∑ →→ Cr(III) … + NTAox Under natural conditions, that can be found in the environment (pH 7.0 or 5.5 and solar light), chromium(VI) is efficiently reduced into chromium species different from non-toxic chromium(III). The reduction of chromium(V) into chromium(III) is a far slower process when chromium is in the presence of strong chelating agents such as polycarboxylate derivatives. The results raise the question about the fate of the reduced intermediate chromium species and more particularly whether we are dealing with the highly carcinogenic chromium(V).

References 1 L. M. Calder, Chromium contamination in the natural and human environments, Wiley Interscience, New York, NY, 1988. 2 S. A. Katz and H. Salem, The Biological and Environmental Chemistry of Chromium, VCH, New York, 1994. 3 J. O. Nriagu, E. Nieboer and Editors, Advances in Environmental Science and Technology, Vol. 20: Chromium in the Natural and Human Environments, 1988.

4 C. D. Palmer and P. R. Wittbrodt, Processes Affecting the Remediation of Chromium-Contaminated Sites, Environ. Health Perspect., 1991, 92, 25–40. 5 S. J. Hug, H.-U. Laubscher and B. R. James, Iron(III) Catalyzed Photochemical Reduction of Chromium(VI) by Oxalate and Citrate in Aqueous Solutions, Environ. Sci. Technol., 1997, 31, 160–170. 6 M. Gaberell, Y. P. Chin, S. J. Hug and B. Sulzberger, Role of dissolved organic matter composition on the photoreduction of Cr(VI) to Cr(III) in the presence of iron, Environ. Sci. Technol., 2003, 37, 4403–4409. 7 Y. G. Zuo and J. Hoigne, Formation of Hydrogen-Peroxide and Depletion of Oxalic-Acid in Atmospheric Water by Photolysis of Iron(III) Oxalato Complexes, Environ. Sci. Technol., 1992, 26, 1014– 1022. 8 B. M. Voelker and B. Sulzberger, Effects of fulvic acid on Fe(II) oxidation by hydrogen peroxide, Environ. Sci. Technol., 1996, 30, 1106– 1114. 9 D. L. Sedlak and J. Hoigne, The Role of Copper and Oxalate in the Redox Cycling of Iron in Atmospheric Waters, Atmos. Environ., Part A, 1993, 27, 2173–2185. 10 J. H. Espenson, Rate Studies on Primary Step of Reduction of Chromium(VI) by Iron(II), J. Am. Chem. Soc., 1970, 92, 1880–1883. 11 S. L. Andrianirinaharivelo, J. F. Pilichowski and M. Bolte, Nitrilotriacetic Acid Transformation Photoinduced by Complexation with Iron(III) in Aqueous-Solution, Transition Met. Chem., 1993, 18, 37– 41. 12 J. G. Calvert and J. N. Pitts, Photochemistry, Wiley, New York, 1966. 13 K. Wrobel, K. Wrobel, P. L. LopezdeAlba and L. LopezMartinez, Enhanced spectrophotometric determination of chromium (VI) with diphenylcarbazide using internal standard and derivative spectrophotometry, Talanta, 1997, 44, 2129–2136. 14 B. Tang, T. X. Yue, J. S. Wu, Y. M. Dong, Y. Ding and H. J. Wang, Rapid and sensitive spectrofluorimetric determination of trace amount of Cr(III) with o-vanillin-8-aminoquinoline, Talanta, 2004, 64, 955– 960. 15 G. Mailhot, J. F. Pilichowski and M. Bolte, Unexpected Dihydroxylation of Methacrylamide Photoinduced by Chromium-(VI) in AqueousSolution, New J. Chem., 1995, 19, 91–97. 16 P. Fageol, M. Bolte and J. Lemaire, Initiation Mechanism of Acrylonitrile Polymerization Photoinitiated by Chromium(VI), J. Phys. Chem., 1988, 92, 239–243. 17 O. Abida, G. Mailhot, M. Litter and M. Bolte, Impact of iron-complex (Fe(III)-NTA) on photoinduced degradation of 4-chlorophenol in aqueous solution, Photochem. Photobiol. Sci., 2006, 5, 395–402. 18 C. Lafond, C. Pizzocaro, R. A. Lessard and M. Bolte, Primary photochemical process in films of dichromated gelatin: a quantitative approach, Opt. Eng., 2000, 39, 610–615. 19 C. Pizzocaro, C. Lafond and M. Bolte, Dichromated polyvinylalcohol: key role of chromium(V) in the properties of the photosensitive material, J. Photochem. Photobiol., A, 2002, 151, 221–228. 20 J. M. Meichtry, M. Brusa, G. Mailhot, M. A. Grela and M. I. Litter, Heterogeneous photocatalysis of Cr(VI) in the presence of citric acid over TiO2 particles: Relevance of Cr(V)-citrate complexes, Appl. Catal., B, 2007, 71, 101–107. 21 M. Bolte and J. Lemaire, HCrO4- Photosensitized Oxidation of Tryptophan Exposed to near-Uv-Visible Light, J. Photochem. Photobiol., A, 1989, 46, 285–294. 22 I. J. Buerge and S. J. Hug, Kinetics and pH dependence of chromium(VI) reduction by iron(II), Environ. Sci. Technol., 1997, 31, 1426–1432. 23 R. J. Kieber and G. R. Helz, Indirect Photoreduction of Aqueous Chromium(VI), Environ. Sci. Technol., 1992, 26, 307–312.

