Kinetic studies of the oxidation of L-ascorbic acid by tris(oxalate)cobaltate in the presence of CDTA metal ion complexes

ARTICLE

Kinetic studies of the oxidation of L-ascorbic acid by tris(oxalate)cobaltate in the presence of CDTA metal ion complexes

Horacio D. Moya^I; Nina Coichev^{*, II}

^IFaculdade de Medicina da Fundação do ABC, FMABC, CP 106, 09060-650 Santo André – SP, Brazil

^IIInstituto de Química, Universidade de São Paulo, CP 26.077 05599-970 São Paulo – SP, Brazil

ABSTRACT

It was investigated the kinetic of the reduction reaction of tris(oxalate)cobaltate by L-ascorbic acid, H₂A, at different acidity. The Co(III) complex was monitored at 600 nm, under pseudo-first-order conditions: [H₂A] = 3.0 x 10^-2 mol L^-1, [Co(C₂O₄)₃]^3- = 3.0 x 10^-3 mol L^-1, in the presence of CDTA (3.0 x 10^-3 mol L^-1) and ionic strength 1.0 mol L^-1, (NaCl) at 25.0±0.1 ºC. The catalytic activity of CDTA metal ion complexes of Fe(III), Ni(II), Cu(II), Cr(III) and Mn(II) was also examined. The pseudo-first-order rate constant was proportional to the concentration of Fe(III)/CDTA, which is the best catalyst, in the range of (1-10) x 10^-5 mol L^-1.

Keywords: ascorbic acid, tris(oxalate)cobaltate(III), CDTA, iron(III), kinetic study

RESUMO

Realizaram-se estudos cinéticos envolvendo a reação de redução de tris(oxalato)cobaltato por L-ácido ascórbico, em diferentes valores de pH. A variação da concentração do complexo de Co(III) foi acompanhada pela absorbância em 600 nm, em condições de pseudo-primeira ordem: [H₂A] = 3,0 x 10^-2 mol L^-1, [Co(C₂O₄)₃]^3- = 3,0 x 10^-3 mol L^-1, na presença de CDTA (3,0 x 10^-3 mol L^-1), em I = 1,0 mol L^-1 (NaCl) e a 25,0±0,1 ºC. Foram investigadas as atividades catalíticas dos complexos de CDTA com Fe(III), Ni(II), Cu(II), Cr(III) e Mn(II). Para Fe(III)/CDTA, o melhor catalisador, os valores de constantes de velocidade observada de pseudo-primeira ordem foram proporcionais à concentração de ferro (1-10) x 10^-5 mol L^-1.

Introduction

The redox reactions of L-ascorbic acid are of fundamental interest in chemistry, biochemistry, pharmacology and several areas of medicine, since it is necessary in human diet in order to synthesize collagen and epinephrine, besides preventing scurvy. L-Ascorbic acid, H₂A, has two acid protons (pK₁ = 4.04 and pK₂ = 11.34), and is a strong reducing agent (E⁰ D/H₂A = 0.390 V vs. N.H.E.) in aqueous solution. This reduction potential depends on the medium acidity, and in some cases the HA^–, HA^•, A^-• radicals may be formed as intermediate species.^1-4

Redox studies of L-ascorbic acid by various metal ion complexes have proposed the formation of a protonated ascorbate free radicals as H₂A^+• or HA^• but not the radical A^-• in the rate-determining step.^3-9

The oxidation studies of L-ascorbic acid by [Co(C₂O₄)₃]^3– were already performed in basic and acid aqueous solutions, and it was pointed out that the redox process in acidic medium (3.2 < pH < 4.7) produced L-dehydroascorbic acid, D (equation 1). L-dehydroascorbic acid, D, also includes the hydration and cyclization of D form, with formation of the bicyclic-L-dehydro species.^8,9 These studies indicated an outer-sphere electron-transfer process and no evidence of stable intermediate species formation.

The redox reaction rate demonstrates a strong dependency on the pH and the oxidation of A^2– showed to be faster than HA^– or H₂A.^8,9

It was found that iron(III) catalyzes the oxidation of L-ascorbic acid by dissolved oxygen, hydrogen peroxide, peroxide-bound chromium and [Co(C₂O₄)₃]^3– in presence of EDTA.^10-14 This last study, performed in universal buffer medium, led to a linear relationship between the observed pseudo-first order constant and the iron(III)/EDTA complex concentration.¹⁴

The catalytic effect of iron(III)/EDTA could be explained by the much faster reactions of iron(III)/EDTA, rather than [Co(C₂O₄)₃]^3– , with L-ascorbic acid.¹⁴ The rate constants for the reduction of several iron(III)/complexes and [Co(C₂O₄)₃]^3– by L-ascorbic acid are reported in the literature and showed the strong influence of acidity and also the effect of the ligand, L, coordinated to Fe(III) (L = H₂O, phen, EDTA, bipy, C₂O₄^2–).^{5,6,8-10,14-17}

The reaction of Fe(II) with [Co(C₂O₄)₃]^3– , which was investigated by some authors, has also to be considered. By keeping [Co(C₂O₄)₃]^3– at 5x10^-2 mol L^-1, which is about ten times over iron(II) concentration, the k_obs value obtained for the redox reaction was 9.49 s^-1 and k = 190 mol^-1 L s^-1.^18-20

In this work the catalytic effect of iron(III) on the oxidation of L-ascorbic acid by [Co(C₂O₄)₃]^3– was studied in the presence of polyaminocarboxylic acid, CDTA (pK₁ = 2.42; pK₂ = 3.54; pK₃ = 5.84 and pK₄ = 9.22), in a universal buffer solution over a large pH region. This study provides information to improve an analytical method for iron(III) at pH = 7.0.^14,17 For comparative studies, the catalytic effect of others transitions metal ions such as: Ni(II), Cu(II), Cr(III) and Mn(II) was also investigated.

Experimental

Reagents and solutions

All reagents used were from AR or CP specification and all solutions were prepared using deionised water obtained from a Nanopure System. Sodium chloride (Merck), stock solution (2.0 mol L^-1), was prepared without further purification or standardization. Sodium hydroxide (Aldrich), 1.768 mol L^-1 was standardized with potassium hydrogenphthalate.²¹

CDTA, trans-1,2-cyclohexyletilenedinitrilotetraacetic acid, C₁₄H₂₂N₂O₈ .H₂O (Merck), stock solution (1.5x10^-2 mol L^-1, pH = 5.8) was prepared by dissolution in NaOH standard solution (pH= 5.8).

Phosphoric, acetic and boric acids (Merck) 1.80 mol L^-1 stock solutions were diluted to 0.180 mol L^-1and then standardized with NaOH solution (0.1768 mol L^-1).

Iron(III) perchlorate (Aldrich) stock solution (6.0x10^-2 mol L^-1) was standardized by complexometric method with EDTA.²¹ The potassium tris(oxalate) cobaltate(III) salt was prepared as described elsewhere.²² It was dissolved in buffer solution just before use in order to get a solution 6.0 x 10^-3 mol L^-1. L-Ascorbic acid (Merck), L-C₆H₈O₆, was also dissolved in buffer solution, just before use, in order to obtain a solution 6.0 x 10^-2 mol L^-1.

Working solutions and spectrophotometric measurements

Several universal buffers solutions were prepared containing H₃PO₄, H₃C-COOH and H₃BO₃ (0.18 mol L^-1 of each). NaOH 1.768 mol L^-1 standard solution was used to adjust the pH of the buffers solutions from 3 to 8.²³ NaCl 2.0 mol L^-1 solution was use to make up the ionic strength 1.0 mol L^-1 in all working solutions.

The working solutions were prepared by mixing equal volumes of the of L-C₆H₈O₆ 6.0 x 10^-2 mol L^-1 and [Co(C₂O₄)₃]^3– 6.0 x 10^-3 mol L^-1 solutions containing CDTA 6.0 x 10^-3 mol L^-1. In the catalytic studies iron(III) was added as Fe(III)/CDTA in the L-C₆H₈O₆ solution just before mixing, once Fe(III) is reduced by L-C₆H₈O₆.

The final solutions concentration of [Co(C₂O₄)₃]^3– , CDTA, L-C₆H₈O₆ and Fe(III)/CDTA, after mixture, were 3.0x10^-3, 3.0x10^-3, 3.0x10^-2 and (1-10)x10^-5 mol L^-1, respectively.

A glass electrode, combined with an Ag/AgCl reference electrode, Metrohm AG Herisau, filled with 3.0 mol L^-1 NaCl and a 654 pHMeter Metrohm instrument were used in the pH measurements at (25.0 ± 0.1) ºC.

Spectrophotometric measurements were performed in a Hewlett Packard 8452A diode-array spectrophotometer using a thermostated Tanden cell (optical path length = 0.875 cm).

Results and Discussion

Data treatment

The redox reaction was followed spectrophotometrically at 600 nm where the major absorbing species is the complex [Co(C₂O₄)₃]^3– (molar absorptivity of 150 ± 10 mol^-1 L cm^-1).¹⁴

A ten times excess of L-C₆H₈O₆ over [Co(C₂O₄)₃]^3– was kept in all experiments in order to have pseudo first order conditions. As the reaction was not affected by dissolved oxygen it was not necessary to eliminate the dissolved air before the kinetic runs.

The kinetic curves were analysed with the OLIS KINFIT set of programs. All the observed rate constants values presented in this work are the mean of at least three determinations and have an average error smaller than 5%.²⁴

The pH influence on the uncatalysed reaction

The variation of [Co(C₂O₄)₃]^3– concentration with time from the reduction by L-C₆H₈O₆, in the absence of any CDTA metal ion complexes showed a pseudo-first-order behaviour.

These experiments were carried out over a large pH range in universal buffer solution containing CDTA, which was added to avoid precipitation of cobalt(II) and cobalt(III) oxalate at high pH. No experiments at pH lower than 3.0 were carried out, due to the [Co(C₂O₄)₃]^3– decomposition and precipitation of CDTA.^25,26

The dependence of k_obs with pH (Figure 1) suggests that the H₂A specie is much less reactive than the HA^– species. In the 6.0 < pH < 7.5 range, where HA^– is the predominant species, the k_obs value is almost constant. It was observed that at pH higher than 10 the oxidised species, D, decomposes rapidly.^1,4,14

The sequence of reactions represented below, may describe the mechanism.^8,9,14

The reduction rate of the tris(oxalate)cobaltate(III) complex concentration is given by the rate law described in equation 8. Using the experimental data obtained at pH range from 3 to 5 the rate law under pseudo-first-order conditions can be written by equation 9, which results in equation 10 (in all equations C_H2A is the total concentration of ascorbic acid).

In this pH range the contribution A^2– is very small (see Figure 1) and the term k_c [A^2–] can be ignored.

By working under limiting conditions, such as k_a is much smaller than k_b because H₂A (equation 4) reacts much slower than HA^– (equation 5), the term 2k_a [H⁺] can be neglected and the equation 10 can be represented as equation 11, as follows:

Taking the equation 11, in the pH region from 3 to 5, a plot of 1/ k_obs vs. [H⁺] provides a linear relationship and from the slope and intercept the values of k_b and K₁ can be, respectively, obtained (Figure 2). The least squares regression (Y= 5465 + 6.382x10⁶ X, r = 0.97) showed some dispersion of the experimental data in spite of the k_obs values have been obtained with the average error smaller than 5%. The values found were k_b = (3.2 ± 1.5) x10^-3 mol^-1 L s^-1 and K₁ = (0.83 ± 0.09) x 10^-4 mol L^-1 and the order of magnitude of these data is in good agreement with the literature: k_b = 4.1x10^-3 mol^-1 L s^-1, K₁ = 1.12x10^-4 mol L^-1, k_b = 7.0x10^-3 mol^-1 L s^-1 and K₁ = 0.71x10^-4 mol L^-1.^9,14

By using the experimental data, which were obtained in the pH range from 6.0 to 8.0, the rate law can be described according to the following equations:

Taking into account the second ionisation step for ascorbic acid K₂ = 4.6 x 10^-12 mol L^-1, the k_c = 8.5 mol^-1 L s^-1 was obtained from the slope of the k_obs vs. 1/[H⁺] plot. The k_c values found in the literature were 20 mol^-1 L s^-1 and 10.8 mol^-1 L s^-1, in universal buffer solution and in ionic strength kept with NaClO₄ and NaCl, respectively.^8,14,23

The iron(III) catalytic effect

The variation of the [Co(C₂O₄)₃]^3– concentration by the reduction with L-C₆H₈O₆, in the presence of Fe(III)/CDTA complex, also revealed a pseudo-first-order rate behaviour. The observed rate constant is proportional to the Fe(III) concentration and dependent of the pH medium (Figures 1 and ³).

It can be also noted that at pH lower than 5.0, where H₂A and HA^– species are present, the reaction is slower even in the presence of Fe(III).

The k_obs values depend linearly on the iron concentration (^{Figure 3}). According to equation 14, where the intercept, k_unc, is k_obs for the uncatalysed reaction and the slope, k_spe, is the specific rate constant for iron(III) catalysed reaction. At pH = 5.0 it was obtained k_unc. = 2.2x10^-4 s^-1 and k_spe = 24 mol^-1 L s^-1 but at pH = 7.5, k_unc and k_spe were found to be 3.2x10^-4 s^-1 and 93 mol^-1 L s^-1, respectively (^{Figure 3}). In similar studies with Fe(III)/EDTA, at pH 7.0, it was obtained k_spe = 145 mol^-1 L s^-1and the k_obs and k_spe had maximum values, where HA^– is the predominant species in solution.¹⁴

Likewise Fe(III)/EDTA, when Fe(III)/CDTA is acting as a catalyst, the mechanism can be described by the sequence of reactions from 15 to 19, where the electron transfer reaction involving Fe(II) is much faster than the reaction of Co(III) complex with ascorbic acid, as it was described in the literature.^5,6,9,14-17

The catalytic effect of Fe(III)/CDTA (log b = 30.1, k_spe = 93 mol^-1 L s^-1, pH = 7.5) is smaller than Fe(III)/EDTA (log b = 23.8, k_spe = 145 mol^-1 L s^-1, pH = 7.5).^14,27 The Fe(III)/EDTA complex, which is seven coordinated with one coordination site occupied by a labile water molecule, will account for the efficient formation of a inner-sphere complex, represented by equation 15.^10,28 The catalyst activities order for Fe(III)/L (L = phen, bipy, H₂O, EDTA, CDTA) reaction with L-C₆H₈O₆ depends on the nature of ligand and is in agreement with the reduction potential values for the complexes involved.^5,6,9,14-17

Influence of others transition metal ions

The catalytic effects of other CDTA metal ion complexes were also investigated. At pH = 7.5 (Universal buffer) the observed rate constant was 1.7x10^-4 s^-1 without catalyst ([Co(C₂O₄)₃]^3- , CDTA and L-C₆H₈O₆ were 3.0x10³, 3.0x10^-3 and 3.0x10^-2 mol L^-1, respectively). Cu(II)/CDTA (k_obs = 2.3x10^-4 s^-1), Mn(II)/CDTA (k_obs = 2.4x10^-4 s^-1) and Ni(II)/CDTA (k_obs = 2.4x10^-4 s^-1) exhibit no significant catalytic effect. Cr(III)/CDTA showed some activity (k_obs = 3.8x10^-4 s^-1) but ten times lower than the Fe(III)/CDTA (k_obs = 3.9x10^-3 s^-1). For Cr(III)/CDTA it can be assume that Cr(III) is present as free ion due to the inertia in the Cr(III) complexes formation.

Conclusions

The reduction of iron(III) complexes and tris(oxalate) cobaltate by L-ascorbic acid depends on the pH and the nature of the ligand. Iron(III) complexes react much faster than [Co(C₂O₄)₃]^3– , which may explain the catalytic effect of iron(III) ion. Considering the dependence of k_obs with the pH, it was possible to calculate the rate constants of the reaction involving the species H₂A, HA^– and A^2– (equations 4-6), the order of reactivity was H₂A < HA^– < A^2–. The present studies showed that Fe(III)/CDTA increases the catalytic activity of iron(III). As suggested by the literature,^1-3 the addition of EDTA or CDTA to complex metal ions and stabilize ascorbic acid solutions from oxidation by oxygen may not work properly.

Acknowledgments

The authors acknowledge the financial support from Brazilian agencies: FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo), CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) and NEPAS (Núcleo de Ensino, Pesquisa e Assessoria a Saúde da FMABC).

Supplementary Information

Supplementary data are available free of charge as PDF file at http://jbcs.sbq.org.br.

Received: July 1, 2005

Published on the web: February 22, 2006

FAPESP helped in meeting the publication costs of this article.

• 1. Tur'yan, Y.; Kohen, R.; J. Electroanal. Chem. 1995, 380, 273.
• 2. Roig, M.G.; Rivera, Z.S.; Kennedy, J.F.; Int. J. Food Sci. Nutr. 1993, 44, 59.
• 3. Davies, M.D.; Polyhedron 1992, 11, 285 and references therein.
• 4. Willians, N.H.; Yandell, J.K.; Aust. J. Chem 1982, 35, 1133.
• 5. Bänsch, B.; Martinez, P.; Uribe, D.; Zuluaga, J.; van Eldik, R.; Inorg. Chem 1991, 30, 4555.
• 6. Bänsch, B.; van Eldik, R.; Martinez, P.; Inorg. Chim. Acta 1992, 201, 75.
• 7. Martinez, P.; Zuluaga, J.; Uribe, D.; van Eldik, R.; Inorg. Chim. Acta 1987, 136, 11.
• 8. Martinez, P.; Zuluaga, J.; Kraft, J.; van Eldik, R.; Inorg. Chim. Acta 1988, 146, 9.
• 9. Kimura, M.; Yamamoto, M.; Yamabe, S.; J. Chem. Soc. Dalton Trans. 1982, 2, 423.
• 10. Taqui Khan, M.M.; Martell, A.E.; J. Am. Chem. Soc. 1967, 89, 4176.
• 11. Taqui Khan, M.M.; Martell, A.E.; J. Am. Chem. Soc. 1967, 89, 7104.
• 12. Grinstead, R.R.; J. Am. Chem. Soc. 1960, 82, 3464.
• 13. Ghosh, S.K.; Gould, E.S.; Inorg. Chem. 1989, 28, 1538.
• 14. Fornaro, A.; Coichev, N.; J. Coord. Chem. 1999, 46, 519.
• 15. Subba Rao, P.V.; Saradamba, G.V.; Ramakrishna, K.; Mohana Rao, K; Subbaiah, K.V.; Indian J. Chem 1989, 28A, 1060.
• 16. Pelizzetti, E.; Mentasti, E.; Pramauro, E.; Inorg. Chem. 1978, 17, 1181.
• 17. Ohashi, K; Sagawa, T; Goto, E; Yamamoto, K.; Anal. Chim. Acta 1977, 92, 209.
• 18. Barrett, J.; Baxendale, J.H.; Transactions of the Faraday Society 1956, 52, 210.
• 19. Haim A.; Sutin, N.; J. Am. Chem. Soc 1966, 88, 5343.
• 20. Cannon, R.D.; Stillman, J.S.; J. Chem. Soc., Dalton Trans. 1976, 5, 428.
• 21. Skoog, D.A.; West, D.M.; Holler, F. J.; Fundamentals of Analytical Chemistry, 7^th ed., Saunders College Publishing: Orlando, 1996, p. 870.
• 22. Booth, H. S.; Inorganic Synthesis, 1^st ed., McGraw-Hill Book Company: New York, 1939, Vol. I, p. 36.
• 23. Lurie, L.; Handbook of Analytical Chemistry, 2^nd ed., Mir: Moscow, 1975, p. 488.
• 24. Olis Kinfit Routines, On-Line Instruments Systems, Inc. Jefferson: GA, 1989.
• 25. Hin-Fat, L.; Higginson, W.C.E.; J. Chem. Soc. (A) 1970, 2836.
• 26. Aggett, J.; Odell, A.L.; J. Chem. Soc. (A) 1968, 1415.
• 27. Beck, M. T.; Gorog, S.; Journal of Inorganic & Nuclear Chemistry 1960, 12, 353.
• 28. Dellert-Ritter, M.; van Eldik, R.; J. Chem Soc., Dalton Trans 1992, 1037 and references therein.

Publication Dates

• Publication in this collection
  17 Apr 2006
• Date of issue
  Apr 2006

History

• Received
  01 July 2005