CrV ... - MAFIADOC.COM

1 downloads 0 Views 666KB Size Report
Kinetic studies show that the redox reaction between Glucur and CrVI proceeds through a mechanism combining one- and two-electron pathways for the reduction ...... 53 O. Pestovsky, A. Bakac and J. H. Espenson, J. Am. Chem. Soc., 1998,.
View Article Online / Journal Homepage / Table of Contents for this issue

www.rsc.org/dalton | Dalton Transactions

PAPER

Redox and complexation chemistry of the CrVI /CrV /CrIV -D-glucuronic acid system†‡

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

Juan Carlos Gonz´alez,a Silvia Garc´ıa,a Sebasti´an Bell´u,a Juan Manuel Salas Peregr´ın,b Ana Mar´ıa Atria,c Luis Federico Sala*a and Sandra Signorellaa Received 31st July 2009, Accepted 17th November 2009 First published as an Advance Article on the web 5th January 2010 DOI: 10.1039/b915652f When excess uronic acid over CrVI is used, the oxidation of D-glucuronic acid (Glucur) by CrVI yields D-glucaric acid (Glucar) and CrIII as final products. The redox reaction involves the formation of intermediate CrIV and CrV species, with CrVI and CrV reacting with Glucur at comparable rates. The rate of disappearance of CrVI , and CrV increases with [H+ ] and [substrate]. The experimental results indicated that CrIV is a very reactive intermediate since its disappearance rate is much faster than CrVI /CrV and decreases when [H+ ] rises. Even at high [H+ ] CrIV intermediate was involved in fast steps and does not accumulate in the reaction. Kinetic studies show that the redox reaction between Glucur and CrVI proceeds through a mechanism combining one- and two-electron pathways for the reduction of intermediate Cr(IV) by the organic substrate: CrVI → CrIV → CrII and CrVI → CrIV → CrIII . The mechanism is supported by the observation of free radicals, CrO2 2+ (superoxoCrIII ion) and CrV as reaction intermediates. The EPR spectra show that five-co-ordinate oxo-CrV bischelates are formed at pH ≤ 4 with the uronic acid bound to CrV through the carboxylate and the a-OH group of the furanose form. Five-co-ordinated oxo-CrV monochelates are observed as minor species in addition to the major five-co-ordinated oxo-CrV bischelates. At pH 7.5 the EPR spectra show the formation of a CrV complex where the cis-diol groups of Glucur participate in the bonding to CrV . In vitro, our studies on the chemistry of CrV complexes can provide information on the nature of the species that are likely to be stabilized in vivo. In particular, the EPR pattern of Glucur-CrV species can be used as a finger print to identify CrV complexes formed in biological systems.

Introduction Compounds of CrVI represent a potential environmental hazard because of their mammalian toxicity and carcinogenecity.1–4 The observation of CrV and CrIV intermediates in the selective oxidation of organic substrates by CrVI and their implication in the mechanism of Cr-induced cancers1,5–7 has generated a considerable amount of interest in the chemistry and biochemistry of this element.8–13 We are investigating the possible fate of CrVI and CrV in biological systems by examining reactions of CrVI with lowmolecular-weight neutral14–26 and acid27–29 saccharides. Naturally occurring acid saccharides are suitable ligands for stabilization of CrV , since they possess 2-hydroxycarboxylato and vic-diolato sites for potential chelation of CrV . Glucur is an essential saccharide in plant and animal physiology. It is one of the metabolites involved in the formation of polygalacturonic acid present in pectins of the

plant kingdom.30 In animal physiology it is vital to the elimination of xenobiotic substances.31 The determination of the ability of Glucur to reduce or stabilize high oxidation states of Cr, will contribute to unravelling its potential role in the biochemistry of this metal. It is then necessary to determine the ability of Glucur to bind and reduce CrVI and CrV under conditions where interference from other reaction products is negligible as well as to assign the coordination modes of Glucur in the CrV species formed in solution. In vitro studies on the chemistry of CrV complexes can provide information on the nature of the species that are likely to be stabilized in vivo. In this work, we report the study of the redox reaction of CrVI and CrIV with an excess of Glucur in acidic medium and the ability of Glucur to coordinate CrV generated in the Glucur/CrVI reaction.

Results a

Instituto de Qu´ımica de Rosario-CONICET, Universidad Nacional de Rosario, UNR, Suipacha 531, S2002LRK, Rosario, Argentina. E-mail: [email protected]; Fax: +54 341 4350214; Tel: +54 341 4350214 b Departamento de Qu´ımica Inorg´anica, Facultad de Ciencias, Universidad de Granada, Fuentenueva s/n, 18071, Granada, Spain c Facultad de Ciencias Qu´ımicas y Farmac´euticas, Universidad de Chile, Casilla 233, Santiago, Chile † Work first presented at the 10th FIGIPAS Meeting in Inorganic Chemistry, Palermo, July 1–4, 2009. ‡ Electronic supplementary information (ESI) available: Additional Figures S1–S6 and Table S1. See DOI: 10.1039/b915652f

2204 | Dalton Trans., 2010, 39, 2204–2217

Detection of CrV by EPR spectroscopy EPR spectroscopy is the most specific and sensitive technique to detect micromolar concentrations of CrV species. So, the presence of oxo-CrV -Glucur complexes can be identified with great sensitivity by EPR spectroscopy. With a modulation amplitude of 0.4 G, mixtures of CrVI /Glucur showed narrow isotropic spectra signals at room temperature. This spectrum consisted of a major signal, 86%, (from 52 Cr) centered at giso = 1.9783, This journal is © The Royal Society of Chemistry 2010

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

flanked by a weak quartet from the 53 Cr isotope (I = 3/2, natural abundance 9.5%) and by a minor signal, 14%, at giso = 1.9760, Fig. 1. Both signals were composed of more than one species, as illustrated in Fig. 1 (inset). Changing the [Glucur] did not affect the spectral resolution. [CrVI ] used in EPR experiments was lower in order to avoid relaxation phenomena, which would result in spectra broadening and loss of resolution. The intensity of the EPR signal and the period over which CrV species could be observed were dependent on the [H+ ]. In EPR spectra, the ligand superhyperfine structures (shfs) of a few acid saccharides are already known: oxo-CrV with lactic and tartaric acid.32 In addition, oxo-CrV complexes with 2-hydroxy acids (ehba = 2-hydroxy-2ethylbutanoic and hmba = 2-hydroxy-2-methylbutanoic acids), without carbinolic protons coupled to the CrV electron spin, have been fully characterised.2

Table 1 Experimental and simulated X-band EPR spectra of (a) [CrVI ] = [GSH] = 0.48 mM; [Glucur] = 1.0 M, pH = 1.0, Mod. Amp = 0.4 G, (b) [CrVI ] = [GSH] = 2.48 mM, [Glucur] = 0.42 M, pH = 7.5 (hepes 0.5 M), Mod. Amp. = 0.25 G, (c) [Glucar] = 0.12 M, [CrVI ] = [GSH] = 0.29 mM, pH = 4.5 (buffer HAc/Ac 0.1 M), Reaction time 43 min, Mod. Amp. = 0.20 G, (d) [CrVI ] = [GSH] = 3.63 mM, [lactone] = 0.63 M, Mod. Amp. = 0.25 G I = 1.0 M; T = 20.0 ◦ C, Resolution = 1024. Isotropic spectra were simulated using pest WinSim program pH giso

Multiplicity 1 Haiso /104 cm-1

1–4 1.9783/1.9782 1.9754/1.9763 7.5 1.9798/1.9796 4.5 1.9785/1.9783 3.0 1.9798 1.9802

t/t d/d t/t t/t q q

Species

0.58/0.64 oxo-CrV -Glucur 0.69/0.70 0.75/1.0 0.70/0.78 oxo-CrV -Glucar 0.66/0.58/0.98/0.64 oxo-CrV -lactone 0.77/1.00/0.73/1.00

t = triplet, d = doublet, q = quintet

contribute to the EPR signal observed in the reaction mixtures. Fig. 2, shows experimental and simulated spectra and Table 1 summarizes the spectroscopic EPR parameters for all these oxoCrV species. Taking into account the results shown in Table 1, it is evident that neither a significant amount of Glucar nor glucurono-6,3-lactone were present in the Glucur/CrVI mixtures in acid media. In the Glucur/CrVI reactions at pH 1–4, the ultimate fate of the chromium was a CrIII species and a typical broad CrIII EPR signal centered at g ~ 1.98 was always observed during the later stages of the measurements (Fig. S1, ESI‡).

Fig. 1 Experimental X-band EPR spectra of oxo-CrV -Glucur species [Glucur] = 1.0 M, pH = 1.0; [CrVI ] = 0.48 mM; F = 9.76526 MHz; Mod. Amp = 0.4 G, Sweep Width = 100.0 G, T = 20.0 ◦ C, I = 1.0 M.

When a large excess of Glucur over CrVI was used (≥2000 : 1), either in strongly acidic conditions (0.2–0.4 M HClO4 ) or in the 1–4 pH range, the EPR spectra were composed of two triplets at giso1 = 1.9783 and giso2 = 1.9782 and two doublets at giso3 = 1.9754 and giso4 = 1.9763. The presence of the weak 53 Cr hyperfine peaks with Aiso = 17.2(1) ¥ 10-4 cm-1 , Fig. 1, demonstrated that each of these components originated from Cr. These conclusions were also confirmed by performing additional EPR measurements on solutions from the reaction of Na2 53 CrO4 with excess Glucur at several pH values. At room temperature and pH 1–4, the reaction of CrVI with GSH (1 : 1 ratio) in the presence of 2000-times molar excess of Glucur produced CrV EPR spectra identical to those obtained by direct reaction of CrVI with Glucur. In contrast, under similar conditions in hepes buffer (pH 7.5) different spectra signals of CrV -Glucur were observed. This could be fitted with two triplets at giso5 = 1.9798 and giso6 = 1.9796. The EPR spectral parameters of oxo-CrV -Glucar and oxo-CrV glucurono-6,3-lactone species were also determined: Glucar was formed during the oxidation of Glucur by CrVI and the lactone came from Glucur (acid form) via lactonization. Both could This journal is © The Royal Society of Chemistry 2010

Fig. 2 Experimental and simulated X-band EPR spectra of (a) [CrVI ] = [GSH] = 0.48 mM; [Glucur] = 1.0 M; pH = 1.0; Mod. Amp = 0.4 G; (b) [CrVI ] = [GSH] = 2.48 mM; [Glucur] = 0.42 M; pH = 7.5 (hepes 0.5 M); Mod. Amp. = 0.25 G; (c) [Glucar] = 0.12 M; [CrVI ] = [GSH] = 0.29 mM; pH = 4.5 (buffer HAc/Ac 0.1 M); Mod. Amp. = 0.20 G; (d) [CrVI ] = [GSH] = 3.63 mM; [lactone] = 0.63 M; Mod. Amp. = 0.25 G; I = 1.0 M; T = 20.0 ◦ C; Resolution = 1024. Isotropic spectra were simulated using WinSim program.

Detection of CrVI esters Chromate esters were investigated by differential UV-Vis spectra of Glucur/CrVI mixtures. The mixtures exhibited an absorption band with l max = 377 nm consistent with that ascribed to oxoCrVI -ester.33 At pH 6.0, the redox reaction of CrVI with Glucur Dalton Trans., 2010, 39, 2204–2217 | 2205

View Article Online

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

proceeds very slowly, with negligible reduction of CrVI . Thus, at this pH the ester formation step can be distinguished clearly from the electron transfer reaction. Spectra obtained within 2.0 min after mixing revealed a distinctive absorption band at 377 nm, Fig. 3. Varying the concentration of Glucur between 0.18–0.72 M at pH 6.0 showed that the absorbance at 377 nm increased with increasing [Glucur] probably as a result of a shift towards the ester in the esterification equilibrium.

and CrVI in HClO4 was demonstrated by conversion to CrO2 2+ upon reaction with dioxygen.34–38 In appropriate experimental conditions, such as high [O2 ] – 1.26 mM – and low [CrVI ] – 6.5 ¥ 10-4 mM – the reaction of CrII (if any) with O2 to give CrO2 2+ can compete successfully with: a) the reaction of CrII with CrVI and b) the autocatalytic consumption of CrO2 2+ by CrII . If CrII is an intermediate species in the redox reaction, CrO2 2+ should be detected.35,39 Periodic scanning of the O2 -saturated solution (1.26 mM) of the Glucur/CrVI reaction mixture in 0.1 M HClO4 showed two absorptions bands at 290 and 247 nm characteristic of CrO2 2+ , Fig. 4. These results indicated that CrII forms in the Glucur/CrVI reaction and can be taken as evidence that CrIV is involved in the redox mechanism of the reaction between Glucur and CrVI . Inset Fig. 4 shows evolution of [CrO2 2+ ] with time. Absorbance values were calculated in the following way: Abs(CrO2 2+ ) = Abs247 - Abs350 ¥ (e1 )-1 ¥ (e2 )

Fig. 3 UV-Vis difference spectra of Glucur/CrVI mixtures at pH 6.0, showing the increasing band at 377 nm with increasing [Glucur]. (a) 0.18, (b) 0.36 and (c) 0.72 M Glucur. [CrVI ] = 0.6 mM, I = 1.0 M, T = 33 ◦ C.

Intermediacy of Cr

II

It is known that CrIV oxidizes alcohols as a two-electron oxidant to yield CrII and the oxidized organic product. The fact that CrII is involved in the oxidation mechanism of several alcohols by CrIV

where e1 and e2 represent the molar absorption coefficient of CrVI at 350 nm and 247 nm respectively. In these experimental conditions e1 = 1550 M-1 cm-1 and e2 = 1900 M-1 cm-1 . Taking into account that the molar absorption coefficient of CrO2 2+ at 24735 nm is 7000 M-1 cm-1 , the maximum concentration of CrO2 2+ (tmax = 26.0 min) was 4.7 ¥ 10-4 mM (yield 72%). When the absorbance at 350 nm was negligible, 0.150 mM of Fe2+ was added to bring about this reaction: CrO2 2+ + 3Fe2+ + 4H+  Cr3+ + 3Fe3+ + 2H2 O The spectrum of the reaction mixture was recorded every 2.0 min. Each spectrum was subtracted from the corresponding one prior to Fe2+ addition. What remains is a difference spectrum between the final solution and that after the addition of Fe2+ . As shown in Fig. S2 (ESI‡), there is a negative absorbance difference around 290 nm, consistent with the presence of CrO2 2+ .

Fig. 4 CrO2 2+ formation from the reaction between Glucur and CrVI [Glucur] = 0.13 M, [HClO4 ] = 0.1 M, [O2 ] = 1.26 mM, [CrVI ] = 6.5 ¥ 10-4 mM, I = 1.0 M, T = 25 ◦ C and 10.0 cm quartz cell. Spectra 1–4 and 5–15 were recorded every 2.0 and 10 min respectively. Inset: evolution of [CrO2 2+ ] with time.

2206 | Dalton Trans., 2010, 39, 2204–2217

This journal is © The Royal Society of Chemistry 2010

View Article Online

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

Rate studies Reaction Glucur + CrIV . The reaction of Glucur with CrIV , CrO2+ , generated in situ in acid media and O2 -saturated solutions was followed by CrO2 2+ formation at 290 nm. UV-Vis spectra of mixtures of Glucur/CrIV confirmed that the increase of the absorbance at 290 nm was the result of the increment in the [CrO2 2+ ]. The spectra obtained revealed l max at 247 and 290 nm and the Abs247 /Abs290 ratio was 2.2, which is characteristic of CrO2 2+ . At the end of the Glucur/CrIV reaction, Fe2+ was added and the spectra showed a negative absorbance difference at 290 nm confirming the identity of the CrO2 2+ . The monotonic increase in absorbance was found to follow first-order kinetics. The experimental rate constants, k4 , were calculated by nonlinear least-square fit of absorbance–time data using 80% of the total experimental values, eqn (1): Abst = Abs• + (Abs0 - Abs• )e

(-k4 t)

(1)

where Abs• and Abs0 are absorbance at infinite time and initial absorbance, respectively. Fig. 5 shows spectral scanning of Glucur/CrIV mixtures and the right fitting of the absorbance data vs. time at 290 nm employing eqn (1). It is important to notice that – in eqn (1) – CrO2 2+ is considered as a final redox product: in our experimental conditions CrO2 2+ forms very fast and then decays slowly (~ = 20 times slower than the Glucur/CrIV reaction) to CrIII through a Glucur-independent pathway. At fixed T, [H+ ] and I, the variation of [Glucur] between 1.0–2.5 mM did not affect the decay rate of CrO2 2+ indicating that Glucur does not react with CrO2 2+ . An independent prepared40 solution of CrO2 2+ decays at the same rate as CrO2 2+ formed in the Glucur/CrIV reaction if experimental conditions are the same. Table 2 summarizes values of k4 for various concentrations of Glucur in HClO4 . Experimental conditions were chosen so that the Glucur/CrIV reaction can compete successfully with CrIV

disproportionation into CrVI and CrIII . In the absence of – or very low concentration of – Glucur, disproportionation of CrIV was evident by the appearance of a typical CrVI spectrum which has a characteristic band at 350 nm (Fig. S3). Disproportionation should be avoided because the CrVI formed absorbs at 290 nm. CrVI was not detected using the concentrations of Glucur shown in Table 2. As represented in Fig. 6, plots of k4 vs. [Glucur] gave good straight lines from which slope values of k4H were determined [eqn (2)]. The bimolecular rate constant, k4H , varied linearly with [H+ ]-1 with a positive intercept kI IV = 47.9 ± 2.7 s-1 M-1 and slope kII IV = 11.5 ± 0.5 s-1 eqn (3) (inset Fig. 6). k4 = k4H [Glucur]

(2)

k4H = kI IV + kII IV [H+ ]-1

(3)

The rate constant for CrIV disappearance is given by eqn (4): k4 = (kI IV + kII IV [H+ ]-1 )[Glucur]

(4)

Reaction of Glucur + CrVI . Absorbance curves versus time at 350 nm of the Glucur/CrVI mixtures exhibited a monotonic decrease which cannot be described by a single exponential decay. These kinetics profiles were appropriately described by the set of consecutive first-order reactions of Scheme 1. It is known that CrV species absorb strongly at 350 nm and may superimpose CrVI absorbance yielding the wrong interpretation of spectrophotometric absorbance decay values, especially when the CrVI and CrV decay rates are of the same order or [CrV ] is appreciable.41 Considering the CrV absorption superimposition, the absorbance at 350 nm, at any time during the redox reaction, is given by eqn (5): Abs350 = eVI [CrVI ] + eV [CrV ]

(5)

Combining eqn (5) with rate expressions41 derived from Scheme 1 yields

Fig. 5 Formation of CrO2 2+ from the reaction 1.0 mM Glucur, [H+ ] = 0.30 M, [O2 ] = 1.26 mM, I = 1.0 M, T = 15 ◦ C, [CrIV ] = 0.07 mM. Inset: absorbance at 290 vs. time. Fitted lines were calculated using eqn (1) and Origin 6.0 Program.

This journal is © The Royal Society of Chemistry 2010

Dalton Trans., 2010, 39, 2204–2217 | 2207

View Article Online

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

Table 2 Observed pseudo-first-order rate constants (k4 ) for different [HClO4 ] and [Glucur]a 103 ¥ [Glucur]/M

1.00

[HClO4 ]/M

10 ¥ k4 /s

0.10 0.15 0.20 0.40 0.60 0.80

1.7 ± 0.08 1.4 ± 0.07 1.1 ± 0.05 0.85 ± 0.04 0.70 ± 0.03 0.65 ± 0.03

a

b

1.26

1.50

2.00

2.50

2.0 ± 0.10 1.6 ± 0.08 1.3 ± 0.06 1.0 ± 0.05 0.83 ± 0.04 0.81 ± 0.04

2.4 ± 0.12 2.1 ± 0.10 1.6 ± 0.08 1.2 ± 0.06 1.0 ± 0.05 0.95 ± 0.05

3.3 ± 0.16 2.6 ± 0.13 2.0 ± 0.10 1.5 ± 0.07 1.3 ± 0.06 1.2 ± 0.04

3.9 ± 0.04 3.4 ± 0.04 2.6 ± 0.04 1.9 ± 0.04 1.7 ± 0.04 1.5 ± 0.04

-1

T = 15 ◦ C; [CrIV ]0 = 0.07 mM I = 1.0 M. b Mean values from multiple determinations. Rate constants were obtained using Sigma Plot 8.0 Program.

Abs350 = Abs0 e-2k6 t + k6 eV [CrVI ]0 (e-k5 t - e-2k6 t )/(2k6 - k5 )

(6)

V

Fig. 6 Effect of [Glucur] on k4 at 15 ◦ C, I = 1.0 and [H+ ] (a) 0.10, (b) 0.15, (c) 0.20, (d) 0.40, (e) 0.60 and (f) 0.80 M. Inset: linear dependence of k4H on [H+ ]-1 .

In eqn (6), e refers to the molar absorptivity of oxo-CrV Glucur at 350 nm (eV 2.1 ± 0.20 ¥ 103 M-1 cm-1 see experimental section). Parameters k6 and k5 refer to the rate of disappearance of CrVI and CrV , respectively, and were evaluated from a nonlinear iterative computer fit using eqn (6). It must be noted that in eqn (6), k6 appears in the numerator of the pre-exponential term and 2k6 appears in the denominator and in the exponential terms because, according to the proposed reaction (Scheme 1), only half of the CrVI reaches CrIII through a CrV intermediate. The calculated kinetics parameters, k6 and k5 , for various [Glucur] at fixed [HClO4 ] are summarized in Table 3. In the range of protons employed in this work, plots of k6 vs. [Glucur] gave good straight lines from which values of kH6 were determined, Fig. 7 and eqn (7). The bimolecular rate constant, k6H , varied with the [H+ ] with quadratic dependence as shown in inset Fig. 7, according to eqn (8). k6 = k6H [S]

(7)

k6H = kS6 [H+ ]2

(8)

k6 = kS6 [H+ ]2 [S]

(9)

where kS6 = (1.52 ± 0.15) ¥ 10 s M -2

Scheme 1

-1

-3

Table 3 Observed pseudo-first-order rate constants (k5 and k6 ) for different [HClO4 ] and [Glucur]a [Glucur]/M

0.180

0.240

0.20 0.40 0.60 0.80 0.96

0.1 ± 0.01 0.46 ± 0.05 0.97 ± 0.10 1.5 ± 0.15 2.6 ± 0.26

0.13 ± 0.01 0.63 ± 0.06 1.2 ± 0.12 2.2 ± 0.19 3.4 ± 0.34

[Glucur]/M

0.180

0.240

0.425

0.19 ± 0.02 0.82 ± 0.08 1.6 ± 0.16 2.8 ± 0.28 4.1 ± 0.41

0.22 ± 0.02 1.0 ± 0.1 1.7 ± 0.18 3.3 ± 0.33 5.3 ± 0.33

0.28 ± 0.03 1.1 ± 0.1 2.3 ± 0.23 3.7 ± 0.37 6.3 ± 0.63

0.300

0.362

0.425

1.5 ± 0.07 2.3 ± 0.11 3.0 ± 0.15 3.6 ± 0.18 4.4 ± 0.22

1.6 ± 0.08 2.6 ± 0.13 3.5 ± 0.17 4.2 ± 0.21 5.1 ± 0.25

3

-1

10 ¥ k5 /s

[HClO4 ]/M

a

0.362

103 ¥ k6 b /s-1

[HClO4 ]/M

0.20 0.40 0.60 0.80 0.96

0.300

1.0 ± 0.05 1.6 ± 0.08 1.9 ± 0.10 2.2 ± 0.11 2.6 ± 0.16

1.2 ± 0.06 1.7 ± 0.08 2.1 ± 0.10 2.6 ± 0.13 3.1 ± 0.16

b

1.3 ± 0.07 2.1 ± 0.10 2.5 ± 0.12 3.0 ± 0.15 3.8 ± 0.20

T = 33 ◦ C; [CrVI ]0 = 0.6 mM; I = 1.0 M. b Mean values from multiple determinations. Rate constants were obtained using Sigma Plot 8.0 Program.

2208 | Dalton Trans., 2010, 39, 2204–2217

This journal is © The Royal Society of Chemistry 2010

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

Fig. 7

Effect of [Glucur] on k6 at 33 ◦ C, I = 1.0 and [H+ ]: (a) 0.96, (b) 0.80, (c) 0.60, (d) 0.40, (e) 0.20. Inset: dependence of k6H on [H+ ].

At constant [H+ ], plots of k5 vs. [Glucur] exhibited a linear dependence on [Glucur] with positive intercept, eqn (10) and Fig. 8, from which the bimolecular rate constants, k5H , were calculated. A plot of k5H vs. [H+ ] revealed a linear dependence, eqn (11) and inset Fig. 8. k5 = kH0 + k5H [S]

(10)

k5H = kS5 [H+ ]

(11)

k5 = k5H + kS5 [H+ ] [S]

(12)

the 0.91 M Glucur/CrVI mixture grows and decays at 11 ◦ C, Fig. 9(a). This signal was monitored with higher modulation amplitude (2.0 G) by EPR and higher [CrVI ]0 than those used in the kinetics studies. The higher values were required in order to obtain good intensity/noise ratios and avoid the splitting of the signal hence making measurements of the signal areas more accurate.

where kH0 = (6.4 ± 0.6) ¥ 10-4 s-1 and kS5 = (1.07 ± 0.11) ¥ 10-2 s-1 M-2

Fig. 8 Effect of [Glucur] on k5 at 33 ◦ C, I = 1.0 and [H+ ]: (a) 0.96, (b) 0.80, (c) 0.60, (d) 0.40, (e) 0.20. Inset: dependence of k5H on [H+ ].

The rate constants for the CrVI and CrV disappearance are then given by eqn (9) and (12), respectively. The rate constants k6 and k5 were independently obtained by EPR spectroscopy. In 0.4 M HClO4 , the CrV EPR signal of This journal is © The Royal Society of Chemistry 2010

Fig. 9 EPR signal and absorbance vs. time for the oxidation of Glucur by CrVI , I = 1.0 M, T = 11 ◦ C, [Glucur] = 0.91 M, [H+ ] = 0.4 M, (a) [CrVI ] = 9.1 mM, modulation amplitude = 2.0 G; (b) [CrVI ] = 0.5 mM, l = 350 nm.

Dalton Trans., 2010, 39, 2204–2217 | 2209

View Article Online

The EPR data were fitted using eqn (13) derived from Scheme 1 for the total CrV in the mixture at any time.

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

h = Ak6 (e-k5 t - e-2k6 t )/(2k6 - k5 )

(13)

where A depends on the spectrometer settings (gain, power, modulation, etc.). Fig. 9(b) shows the absorbance changes at 350 nm of the Glucur/CrVI under the same experimental conditions. The k5 and k6 values obtained by EPR or UV-vis were the same using eqn (13) or (6). The absorption UV-vis and EPR spectra of the Glucur/CrVI mixtures evidenced that the final redox species are not CrIII (ac) . All mixtures showed bands at 419 and 563 nm with intensities higher than those attributable to CrIII (ac) . Fig. S4 depicted the spectra obtained from 0.8 M Glucur in 0.4 M HClO4 mixture at different times (900 s–24 h) : the maximum absorbance decayed and, at the same time, shifted to the values expected for CrIII (ac) . Also, kinetic data at 570 nm, Fig. 10(a), indicated that the absorbance grew to higher intensities. This unexpectedly higher absorbance decayed in a very short time (ª 500 s). After that, the absorption values decreased very slowly, see Fig. S4, and reached CrIII (ac) after 24 h. The higher intensity of the CrIII at 570 nm may be due either to the presence of underlying absorption of CrV species or to a CrIII –Glucur intermediate which then decomposed to the final product.

Fig. 10 (a) Experimental absorbance vs. time curves of Glucur/CrVI reactions at 570 nm. Experimental data were fitted with eqn (15). (b) Calculated concentration of Cr species using eqn (7–12). [Glucur] = 0.80 M, [HClO4 ] = 0.4 M, [CrVI ] = 6.0 mM, I = 1.0, T = 33 ◦ C.

The absorbance data vs. time at 570 nm, Fig. 10(a), could be adequately adjusted considering the presence of CrV intermediates, besides CrIII . In this way, the total absorbance at 570 nm at any time is given by eqn (14). Abs570 = eV [CrV ] + eIII [CrIII ] 2210 | Dalton Trans., 2010, 39, 2204–2217

(14)

Combining eqn (14) with rate expressions derived from Scheme 1 for [CrIII ] and [CrV ] present in the mixture, we arrived at eqn (15). Abs570 = eIII [CrVI ]0 {1 - e-2k6 t + (eV - eIII )k6 (e-2k6 t - e-k5 t )/(k5 (15) 2k6 )} Good fitting of experimental data was obtained by means of eqn (15) with eV = 54 M-1 cm-1 and eIII = 22 M-1 cm-1 , Fig. 10(a). Values of eV and eIII employed were reasonable for CrV and CrIII complexes with saccharides.2 Values of k6 and k5 were in agreement with those calculated through either eqn (15) or (7)–(10). The kinetics profile simulated using k5 and k6 showed that the time of maximum intensity (tmax ) of Abs570 is very close to the time calculated for [CrV ]max in the reaction mixture (up to 19% of the total Cr in the solutions), Fig. 10(b). These results indicated that the intermediate CrV species should be responsible for the growth and decay of the absorbance at 570 nm in a short time, then CrIII species decomposed slowly to CrIII (aq) .

Discussion Characterization of CrV -Glucur species Presence of oxo-CrV -Glucur species can be identified with great sensitivity by EPR spectroscopy, where strong and narrow isotropic signals were observed at room temperature in X-band spectra. Assignment of the probable structures of the oxo–CrV species in solution have, therefore, been made on the basis of the isotropic EPR parameters giso , Aiso and 1 Haiso together with any proton shfs. An empirical relationship between the nature and number of donor groups and the EPR spectral parameters of CrV complexes has been established, and five-coordinated CrV species have higher giso and lower 53 Cr Aiso values than the corresponding six-coordinated species.2,42 It is known that the five-membered oxoCrV chelates are more stable than the six-membered ones42–44 and, in addition, Glucur is well suited for stabilization of five-membered CrV chelates since it has 2-hydroxycarboxylato and vic-diolato sites for potential chelation. The EPR spectra of the intermediate oxoCrV species from the reaction Glucur/CrVI , at any [H+ ] – pH ≤ 4 (Fig. 2(a)), consist of two triplets with giso1 = 1.9783 and giso2 = 1.9782 and two doublets with giso3 = 1.9754 and giso4 = 1.9763. Calculated giso1 , giso2 and Aiso = 17.3(1) ¥ 10-4 cm-1 are consistent with those for five-co-ordinated oxo-CrV complexes having two carboxylate and two alcoholate donor groups.45 The shfs of the signals with giso1 and giso2 is that expected for two equivalent carbinolic protons coupled to the CrV electronic spin. In order to yield the most favored five-membered CrV bischelates, these carbinolic protons must belong to the 2-hydroxocarboxylate donor sites of two Glucur molecules. In this way Glucur acts as a bidentate molecule binding to CrV , which gives [CrV O(O6 ,O5 glucofuranuronate)2 ]- . The formation of such a CrV bischelate is only possible with Glucur bound to CrV in the furanosic form. However, in the pH range 2–8, Glucur exists in aqueous solution as an equilibrium between the pyranosic and furanosic isomers, in about 9 : 146,47 ratio, with the b-anomer being the most stable one. Thus, the furanosic form which is the less stable, is the one preferred in complexation. This finding is similar to that found in oxo-CrV with D-galacturonic acid.32 Similarly, NMR studies of MoO4+ and WO4+ with D-galacturonic acid provided evidence This journal is © The Royal Society of Chemistry 2010

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

Fig. 11 Proposed structures of CrV complexes with Glucur at pH 1–4 and 7.5.

for complexation involving the carboxylate and the adjacent carbinolic group as the major species in solution.46,47 The different 1 H aiso and giso for the two triplets correspond to the two geometric isomers of the CrV bischelate ([Cr(O)(O5 ,O6 -glucofuranuronate)2 ](I and II in Fig. 11). The doublet shfs for the species with giso3 1.9754 indicated that, in this oxo-CrV -Glucur complex, only one carbinolic proton is coupled to the CrV electronic spin. This species can be assigned to the oxo-CrV monochelate [Cr(O)(OH2 )2 (O6 ,O5 -glucofuranuronate)](III in Fig. 11) with one Glucur molecule acting as a bidentate ligand through the a-hydroxocarboxylate moiety of the furanosic form of Glucur and two molecules of H2 O. The species with giso4 1.9763 and doublet shfs could correspond to a different monochelate where Glucur coordinates through the 1,2-cis-diolate group of the one molecule of the pyranosic form of Glucur and two molecules of H2 O complete the co-ordination sphere, [CrV O(OH2 )2 (O1 ,O2 -glucopyranuronato)]- (IV in Fig. 11). At pH 7.5, CrV was generated by reaction between GSH and CrVI . In these experimental conditions Glucur trapped CrV yielding oxo-CrV –Glucur species that remained in solution for a long time. At this pH, EPR spectra were composed of two triplets with giso5 1.9798 and giso6 1.9798, Fig. 2(b). EPR parameters indicated that triplets can associate with two bischelates – geometric isomers – where two molecules of Glucur coordinate through the 1,2-cisdiolate group of the pyranosic form of Glucur, [CrV O(O1 ,O2 glucopiranuronato)2 ]3- ,V and VI in Fig. 11. At pH ≤ 4 all EPR spectra in mixtures Glucur/CrVI were constituted of the same kind of species when either very large (170– This journal is © The Royal Society of Chemistry 2010

2000 : 1) Glucur to CrVI ratios are used or upon addition of Glucur to an equimolar GSH/CrVI mixture. These results eliminated the possibility of one of the four oxo-CrV species being formed by chelation to the organic oxidation product, Glucar. This was also confirmed by experiments where CrV was generated by reaction of GHS with CrVI and then stabilized by addition of Glucar at pH 4.5. EPR spectra of oxo-CrV -Glucar species show giso and/or shfs different from those observed for the oxo-CrV -Glucur species (Fig. 2(c)). Triplets with giso = 1.9785 and 1.9783 can be assigned to the two geometric isomers of [CrV O(O1 ,O2 -Glucar)]- . The results obtained indicated that Glucur can coordinate CrV in two different ways depending on the pH: at pH < 4 through the 2-hydroxycarboxylato and at pH 7.5 using 1,2cis -diolate groups. Several molecules possessing both the vicdiol (or cyclic cis-diol) and 2-hydroxycarboxylate groups yield exclusively the CrV -diolate2 complexes at pH 7.5, whilst the CrV 2-hydroxycarboxylate2 chelates are the major species formed at pH ≤ 5.14,15,48 Similar behavior is also observed with the CrV /quinic acid system, where at pH 7.5, CrV binding to the cyclic-cis-diolato group is preferred.42 Oxidation of Glucur by CrVI /CrV The reduction of CrVI by Glucur is strongly dependent on pH. The redox reaction is fast when [H+ ] > 0.1 M but is very slow at pH > 1. Therefore, the 0.2–0.96 M [H+ ] range was chosen to study the kinetics of Glucur/CrVI redox reaction. The timedependent UV/vis spectra of the reaction mixture showed that Dalton Trans., 2010, 39, 2204–2217 | 2211

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

the absorbance at 350 nm and 420–470 nm decreased with time, while absorbance at 570 nm increased, without an isosbestic point (Fig. S5). The lack of an isosbestic point indicated that there are two or more competing reactions at any time and that in the reduction of CrVI to CrIII intermediate chromium species are present in appreciable concentrations. Formation of CrV and/or CrIV intermediates by the redox reaction of CrVI with the substrate has been observed previously for a number of substrates.2,11,32,48 The fact that CrO2 2+ was detected in the reaction of Glucur with CrVI , coupled with the observation of relatively long-lived oxo-CrV species in the EPR experiments, and the successful trapping of organic radicals using acrylamide, indicated that both CrIV and CrV intermediate species were formed in the Glucur/CrVI reaction. Besides, the CrV-IV intermediates detected suggested that the Glucur/CrVI reaction occurred through both one and two electron pathways. However, under conditions employed in the kinetics measurements the reaction between CrIV and Glucur was very fast as demonstrated previously. Using the experimental rate laws, eqn (4) and (12), the CrIV reaction with Glucur (at 15 ◦ C) was at least 2.1 ¥ 103 times higher than the CrV –CrVI reaction at 33 ◦ C. Therefore, CrIV should be involved in fast steps and does not accumulate in the Glucur/CrVI mixtures. The simulated profiles of the Cr species concentration indicated that the maximum and minimum [CrV ] were 0.145 mM (24%, tmax = 160 s) and 0.042 mM (7.0%, tmax = 1920 s) respectively, Fig. S6. The time dependence of the reaction absorption data at several wavelengths can be fitted with the sequence proposed in Scheme 1. Besides, the curves for EPR area vs. time can be fitted with eqn (13) and the first-order rate constants obtained from these measurements agree perfectly with those calculated from the electronic spectroscopy data. By taking account of a) the absence of an isosbestic point in Glucur/CrVI mixtures, Fig. S5, b) the detection and the yielding of CrO2 2+ in Glucur/CrVI and Glucur/CrIV mixtures, c) the detection and characterization of oxo-CrV species with Glucur by EPR, d) the positive test of organic radicals by polymerization of acrylamide, d) the kinetic results of Glucur redox reaction with CrVI , CrV and CrIV , e) the detection of an intermediate CrVI ester at high pH, and f) the formation of Glucar as the only organic reaction product, we are now able to propose a possible mechanism, Scheme 2. In the [H+ ] range under study, CrVI exists49 as HCrO4 - and this species is proposed as the reactive form of CrVI , which is also in agreement with the first order dependence of the reaction rate on [CrVI ]. It is known that oxidation of alcohols, aldonic and uronic acids by CrVI is preceded by the formation of a chromate ester.32,33,50,51 The observation of the absorbance bands characteristic of chromate oxy-esters around 377 nm a few minutes after mixing Glucur and CrVI under conditions where the redox reaction is slow, indicates that at least three intermediate CrVI complexes are formed rapidly prior to the redox steps. Thus, the first step of the mechanism proposed in Scheme 2 involved the formation of Glucur-CrVI monochelates VII–IX with Glucur acting as a bidentate ligand, which is also consistent with the first order dependence of the reaction on [CrVI ]. Several co-ordination modes are possible for the CrVI -Glucur species. Since the oxidation of Glucur occurs only on the anomeric hydroxyl group, the complex IX is the only redox active intermediate. Complex IX 2212 | Dalton Trans., 2010, 39, 2204–2217

must also be in rapid equilibrium with any other linkage isomer (VII, VIII). The slow step proposed in Scheme 2 for the CrVI consumption involves the intramolecular two-electron transfer within the active CrVI -Glucur species (IX) to yield CrIV and Glucar (detected by HPLC), eqn (17), and requires two protons, so that the redox reaction Glucur/CrVI should be favored in acid medium, such as observed. The rate law for the CrVI consumption derived from eqn (16)–(17) in Scheme 2 is given by eqn (23), where [CrVI ]T refers to the total [CrVI ] in the reaction mixture. -d[CrVI ]/dt = k6 K VI [H+ ]2 [Glucur][CrVI ]T /(1 + K VI [Glucur]) (23) If K VI [Glucur]  1, eqn (23) agrees perfectly with the experimental rate law (eqn (9)). The formation of CrIV was consistent with the observation of CrO2 2+ , the product of the reaction of CrII with O2 , which is taken as evidence of the CrIV formation.35,36 After the slow redox step, CrIV is predicted to react with excess substrate to yield CrIII and radical or CrII and oxidize substrate through two alternative fast steps, eqn 18, 19a and 19b. The first parallel step was supported by the observed polymerisation of acrylamide when it was added to the reaction mixture, while the steps 19a and 19b are supported by the formation of CrO2 2+ . As stated above, CrIV does not accumulate in the reaction mixtures because it is involved in fast steps. The yield of CrO2 2+ increases as the [CrVI ]0 decreases and reaches a limiting value of 72%. This result suggests that part of the CrVI is reacting through a pathway that does not involve CrII . The yield of CrO2 2+ is expected to approach 100% if the reaction should take place exclusively through the CrVI → CrIV → CrII pathway.35 CrV was produced by fast reaction of CrII with CrVI eqn 20.32,35 The [CrVI ] used in the kinetic experiments is higher than that used for the detection of CrO2 2+ , therefore CrVI can successfully compete with O2 for CrII . Alternative generation of CrV was the rapid reaction of the organic radical (R∑ ) with CrVI , eqn 21.32,50,51 The reaction of R∑ with CrVI should be very fast and prevents the detection of the organic radical in the EPR measurements. Under the conditions used in the kinetic experiments, high [CrVI ] and low [O2 ], eqn 22 can be neglected.52–54 The kinetic data indicated that CrV formed in the fast steps can further oxidize Glucur through two competitive slow steps to yield CrIII and Glucar as the final redox products. Additionally, on the basis of the EPR and kinetics results, in Scheme 2, it is proposed that CrV reacts with Glucur to form an oxo-CrV -Glucur monochelate (K V 1 ) in rapid equilibrium with the oxo-CrV -Glucur2 bischelate (K V 2 ). CrV -Glucur-monochelate intermediate yielded the redox products in the presence of another Glucur molecule through an acid catalyzed step, (k5 ) or directly without the participation of organic substrate or protons (k). The CrV -Glucur and CrV Glucur2 complexes in eqn (23) represent several linkage isomers – the structure of the major ones (I–IV) was discussed above – but the selectivity of the oxidation of Glucur requires that the complex with the anomeric hydroxyl group bound to CrV should be the redox active CrV species (IV in Fig. 11). The rate law for the disappearance of CrV takes the form of eqn (24), which is in total agreement with the experimental rate law (eqn (12)): -d[CrV ]/dt = {k + k5 [H+ ][Glucur]}[CrV ]T

(24)

This journal is © The Royal Society of Chemistry 2010

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

Scheme 2

Oxidation of Glucur by CrIV The bimolecular rate constants were inversely proportional with [H+ ] in the range 0.2–0.6 M. The rate law was resolved using the known acid–base equilibrium between the carboxylic acid (HS) and the conjugate base (S- ). k4 = k1 [HS] + k2 [S- ] = {(k1 [H+ ] + k2 K a )/(K a + [H+ ])}[HS]T (25)

(26)

The rate law for the disappearance of CrIV takes the form of eqn (26), which is in total agreement with the experimental rate law, This journal is © The Royal Society of Chemistry 2010

Experimental section Materials

since [H+ ]  K a ,55 thus, -d[CrIV ]/dt = d[CrO2 2+ ]/dt(k1 + k2 K a [H+ ]-1 )[HS]T

eqn (4). The rate constant kI IV is much larger than the rate constant for oxidation of HS possibly due to the increased equilibrium constant for formation of the precursor complex from oppositelycharged ions.

acid (Sigma, 99.0%, m.p. 165 ◦ C), D-glucaric acid (Sigma, 98.0%), glutathione (reduced form) (Sigma, 98.0–100.0%), potassium dichromate (Mallinckrodt), sodium perchlorate monohydrate (Fluka 98.0%), 53 Cr2 O3 (Atomic Energy Research D-Glucuronic

Dalton Trans., 2010, 39, 2204–2217 | 2213

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

Establishment, Harwell, England, 95.0%), oxygen (99.99%), argon (99.9%), perchloric acid (A. C. S. Baker), acrylamide (Sigma, 99.0%), ehba = 2-ethyl-hydroxybutanoic acid (Aldrich 99.0%), sodium hydroxide (Cicarelli, p. a.), D-glucofuranurono-6,3-lactone (Sigma 99.0%), diphenylpicryllhydrazyl (dpph), ethyl acetate (Douglas, PA), silver nitrate (Cicarelli, PA), sodium thiosulfate (Cicarelli, PA), sodium dichromate dihydrate (Sigma 99.5%), Zn (Cicarelli, PA), HgCl2 (Cicarelli, PA), H2 SO4 (Cicarelli, PA), NaH2 PO4 (Sigma, HPLC), acetone (Anedra 99.5%), ethanol (Cicarelli, PA), were used without further purification. 4-(2-Hydroxyethyl)-1-piperazinethanesulfonic acid: hepes (Sigma ultra, 99.5%) and acetic acid (Cicarelli, PA) were used to prepare buffers and then employed to adjust the pH value of some solutions to 7.5 and 4–4.5 respectively. Aqueous solutions were prepared in milliQ deionized water (HPLC quality). [CoIII (NH3 )5 Cl]Cl2 56 and Na[CrV O(ehba)2 ]·H2 O57 were synthesized according to the method described in the literature. For experiments performed in the 1–4 pH range, the pH of the solutions was adjusted by addition of HClO4 or NaOH. In experiments performed at constant ionic strength (I = 1.0 M) and different hydrogen ion concentrations, mixtures of sodium perchlorate solution and perchloric acid solutions were used. The concentration of stock solutions of perchloric acid was determined by titration employing standard analytical methods. Caution: CrVI compounds are human carcinogens, and CrV complexes are mutagenic and potential carcinogens.58 Contact with skin and inhalation must be avoided. Acrylamide is carcinogenic and must be handled carefully.59 Methods D-Glucuronic acid stability and D-glucuronic acid/glucurono-6,3-

lactone equilibrium. The stability of Glucur under experimental conditions – used in the kinetics studies – was monitored by HPLC. The chromatograms were obtained by means of a KNK500A chromatograph provided with a 7125 HPLC pump. The separation was carried out on an Aminex HPX-87H column (300 ¥ 7.8 mm, Bio-Rad Laboratories) using 4.5 mM H2 SO4 as eluent and a flow rate of 0.6 mL min-1 , at 33 ◦ C. The effluent was monitored with a UV detector (115 UV Gilson, l = 220 nm). Chromatograms, recorded after incubation of the standard sample in 3.0 M HClO4 (higher than the highest [H+ ] used in the kinetic measurements) at 33 ◦ C during 5.0 h, show two peaks. The retention times (Rt ) were: Rt1 = 7.43 min and Rt2 = 10.63 min, these Rt were identical to those of the freshly prepared sample of the Glucur and Glucofuranurono-6,3-lactone respectively. These results are consistent with the fact that Glucur exists in acid and lactone form in aqueous solution. Since lactone can compete with Glucur to reduce CrVI , it was necessary to quantify the lactone form during the kinetic studies. The HPLC technique showed that the maximum quantity of lactone formed, in the experimental conditions used in the kinetic studies, was 7.2 mM (4.2% based on total [Glucur]), Table S1. In addition, lactone reacted with CrVI 4.3 times slower than Glucur under the same experimental conditions. Consequently, both the small quantity of lactone formed and its low reactivity prevent it from competing with the Glucur-CrVI reaction. Product analysis. In Glucur/CrVI mixtures under the conditions used in the kinetic measurements (excess of Glucur over 2214 | Dalton Trans., 2010, 39, 2204–2217

CrVI ), Glucar was identified as the only reaction product by HPLC. The chromatograms were obtained on a Varian Polaris 200 chromatograph provided with a cc Star 9000 HPLC pump. The separation was carried out on an anion-exchange Spherisorb S Sax HPLC column (250 ¥ 4.6 mm) using 20 mM NaH2 PO4 with 5% of ethanol as eluent and a flow rate of 0.5 mL min-1 , at 44 ◦ C (Zeltec column heater). The effluent was monitored with a UV detector (Prostar 325 UV-vis detector, l = 210 nm). The [H+ ] of the standard and the reaction mixture samples was adjusted to 0.2 M by addition of HClO4 and the samples then filtered through a 0.2 mm membrane prior to injection into the chromatographic system. Standard solutions of Glucur, Dglucuro-6,3-lactone and Glucar were prepared individually in 0.20 M HClO4 and the chromatographic Rt were determined separately. In aqueous solution, Glucur shows two peaks, Rt1 = 8.34 min and Rt2 = 6.47 min corresponding to the acid form and the lactone respectively. Besides the peaks from Glucur and its lactone, in reaction mixtures containing 30 times excess of Glucur over CrVI in 0.20 M HClO4 , a different peak at Rt3 = 10.75 min was detected. Rt of this new peak is coincident with that of a standard solution of Glucar-1,4-lactone. Furthermore, co-chromatography of a CrVI -Glucur reaction mixture with added Glucar-1,4-lactone resulted in the increase of the peak at 10.75 min. Reaction mixtures with Glucur to CrVI ratios lower than 30 : 1 yielded more complex chromatograms because of further oxidation of Glucur by CrVI . For the Glucur to CrVI ratios used in the kinetic studies (higher than 30 : 1), neither carbon dioxide nor formic acid were detected as reaction products. Polymerization test Polymerization of acrylamide was investigated during the reaction of Glucur with CrVI as a test for free organic radical generation. In a typical experiment, a solution of CrVI (1.5 mL, 0.03 M) was added to a solution containing 2.16 mmol of Glucur and 3.7 mmol of acrylamide at pH 3.3. When [CrVI ] became negligible, the mixture was diluted with 5.0 mL of methanol (at 0 ◦ C) which led to the precipitation of a white polymer of acrylamide. Control experiments showed that no polymerization of acrylamide takes places under the experimental conditions with either K2 Cr2 O7 or Glucur alone. Possible reactions of CrV and CrIV with acrylamide were tested with Na[CrV O(ehba)2 ] and [CrIV O(ehbaH)2 ].60 No precipitation occurred on mixing the CrV or CrIV complexes with acrylamide under the conditions used in the CrVI + Glucur reaction. Spectrophotometric measurements In situ generation of oxo-CrIV , CrO2+ . Aqueous CrO2+ was generated by rapid oxygen-oxidation of CrII employing the following procedure. Zn/Hg amalgam was prepared by stirring a mixture of Zn (10.0 g, previously washed with 3.0 M HCl during 3.0 min) and HgCl2 (0.3 M in 0.1 M HCl) for about 30 min. Afterwards, the excess of HgCl2 was eliminated and the resulting amalgam was washed three times with 1.0 N H2 SO4 and finally with distilled water. The amalgam was added to a solution of 6.0 mM Cr(ClO4 )3 in 0.2 M HClO4 (100 ml) and left to stir with H2 bubbling. After 3.0 h, Cr(ClO4 )3 was quantitatively reduced. The [CrII ] was determined by treating a reaction aliquot with an aqueous This journal is © The Royal Society of Chemistry 2010

Published on 05 January 2010. Downloaded by Universidad de Granada on 21/02/2014 17:19:11.

View Article Online

solution of [CoIII (NH3 )5 Cl]Cl2 , under anaerobic atmosphere (Ar); the mixture was then poured into concentrated HCl and the [CoII ] content was analyzed by measuring the absorbance of CoCl4 2- at 692 nm.56 For the in situ generation of CrO2+ , a deoxygenated solution of CrII was injected into an acidic aqueous solution of Glucur saturated with oxygen (1.26 mM). In a typical experiment, 100 ml of 6.0 mM Cr2+ were injected into a septum-capped spectrophotometric quartz cell, with path length of 1.0 cm, filled with 2.3 ml of an O2 -saturated solution containing 1.0–2.5 mM Glucur and appropriate concentrations of HClO4 and NaClO4 at 15 ◦ C. Under these conditions, CrO2+ immediately formed and then reacted with Glucur. At very low CrII /O2 ratios (