Sulfite Induced Autoxidation of Cu ( II ) / Tetra / Penta and Hexaglycine Complexes . Spectrophotometric and Rotating-ring-disk Glassy Carbon Electrode Studies and Analytical Potentialities

A oxidação de complexos de Cu(II) com tetra, penta e hexaglicina, em solução aquosa de tampão borato, pelo oxigênio dissolvido é fortemente acelerada por sulfito. A formação de complexos de Cu(III) com máximos de absorbância em 250 nm (ε = 9000 mol L cm) e 365 nm (ε = 7120 mol L cm) foi também caracterizada usando-se voltametria com eletrodo rotativo disco-anel, na qual componentes anódicos e catódicos foram observados em voltamogramas registrados em solução contendo Cu(II). Voltamogramas, obtidos com várias velocidades de rotação, mostraram que a espécie de Cu(III) gerada eletroquimicamente não é estável em toda a janela de tempo do experimento, e em solução contendo tetraglicina a corrente limite é controlada pela cinética de um equilíbrio envolvendo espécies de Cu(II). O valor calculado da constante de decomposição de primeira ordem foi 4,37x10 s. Experimentos eletroquímicos realizados em solução de Cu(II) após a adição de quantidades relativamente pequenas de sulfito demonstraram que a espécie de Cu(III), formada na reação química, é a mesma que foi coletada no eletrodo anel quando Cu(II) é oxidado no eletrodo disco. A concentração dos complexos de Cu(III) é proporcional à quantidade de sulfito adicionada e os resultados indicaram a possibilidade de desenvolvimento de um método analítico indireto para sulfito, com detecção espectrofotométrica ou amperométrica do produto quimicamente gerado.


Introduction
The present article is a comparative study of the sulfite induced oxidation of Cu(II) complexes with tetra, penta and hexaglycine by dissolved oxygen.The symbols G n (G 4 , G 5 and G 6 ) are used here as general terms for tetraglycine, pentaglycine and hexaglycine respectively.(H -x G n ) -(x+1) refers to a peptide ligand with x deprotonated nitrogens coordinated to the copper ion.The degree of protonation of the copper peptides complexes depends on the medium acidity.The representations Cu II G n and Cu III G n refer to all complexes species present in solution.
The autoxidation of Cu II G 4 and the decomposition of Cu III /G 4 /G 5 in the absence of sulfite were previously studied by Margerum's research group. 1,2 he reaction of O 2 with [Cu II (H -3 G 4 )] 2-, represented by the general reaction (equation 1), is very slow at room temperature and pH 7-10.This reaction is thermodynamically favorable (E o Cu (III/II) G 4 = 0.63 V and E o O 2 /H 2 O = 0.815 V vs NHE).This reaction is catalyzed by initial traces of [Cu III (H -3 G 4 )] -(10 -7 mol L -1 ), with an induction period which becomes smaller by addition of strong oxidants as [Cu III (H -3 G 4 )] -(electrochemically generated).The Cu III G 4 is moderately stable in neutral solution, with a half-life of 5.5 h at 25 °C. 1 The rate of subsequent decomposition of [Cu III (H -3 G 4 )] - is dependent on the pH and the oxygen concentration.The main species in solution (pH 7-10) prior to decomposition is [Cu III (H -3 G 4 )] -.The pK a value of [Cu III (H -3 G 4 )]H and [Cu III (H -3 G 4 )] -are 4.2 and 12.1 (Table 1), respectively. 2fter the completion of the oxygenation and the decomposition of [Cu III (H -3 G 4 )] -, new species were detected in the solution as suggested in the following mechanism (pH 7-9). 1 [Cu III (H -3 G 4 )] -→ R + H + (2) R + O 2 → RO 2 (3) R + [Cu III (H -3 G 4 )] -→ Cu II ((H -1 G 4 )DHP) + [Cu II (H -3 G 4 )] 2-(4) R is a reactive intermediate (either a carbon-centered free radical or a Cu(I) complex), RO 2 is either a peroxy radical or a copper(III) peroxide and G 4 DHP is a dehydropeptide which hydrolyzes to give glycylglycinamide and glyoxylglycine. 1urther studies 2 of the redox decomposition of Cu III G 4 at pH 6-8 showed that the species [Cu II (H -2 G 4 )] -is responsible for the catalysis of the decomposition of [Cu III (H -3 G 4 )] -(equations 5 and 6): The [Cu III (H -2 G 4 )] is expected to have a higher reduction potential than [Cu III (H -3 G 4 )] -because one of the peptidic nitrogen is not coordinated and a faster redox decomposition reaction occurs.Kirschenbaum and Meyerstein 3 also concluded the relative instability of neutral [Cu III (H -2 G 4 )] compared to [Cu III (H -3 G 4 )] -species.
Studies related to Cu III G 5 complex in basic medium (pH 11.6), in experimental condition where [Cu III (H -4 G 5 )] 2- is the predominant species, showed that the initial products, are dehydropeptides which hydrolyze to form amides and corresponding carbonyl species. 4,5argerum et al. 1,2,[4][5][6] studied the oxidation of Cu II G 4 complexes by oxygen in the presence and absence of sulfite, these data could not clearly explain the induction period and the autocatalytic behaviour.The present work brings more information about the reaction of Cu II G 4 with oxygen accelerated by sulfite.The studies were carried out in excess of Cu(II), G 4 and oxygen compared two sulfite concentration, such as a first order formation of Cu III G 4 could be evaluated.Besides we also investigated the Cu II G 5 and Cu II G 6 complexes.
We also did complementary study about the decomposition of Cu III G 4 , electrochemically generated, by using the rotating ring-disk electrode (RRDE).The great advantage of the RRDE is that it can be used to analyse short-live species (unstable intermediate or product).The used RRDE technique is more sensitive and precise for kinetic studies than the one used by Margerum group, 2,4 which consisted of a flow system, not adequate for very short-live product as further discussed.
In the present study, Cu(III) formation was followed when sulfite was added to air saturated solutions of Cu II G 4 , G 5 and G 6 complexes in borate buffer in aqueous medium.These new comparative studies, using spectrophotometric and electrochemical measurements bring a new contribution to better understanding of the mechanism and development of a new alternative analytical method.

Reagents
All reagents used were of analytical grade (Merck or Sigma).The water used to prepare the solutions was purified with a Milli-Q Plus Water system (Millipore).
Stock solutions of sulfite (2.00x10 -2 mol L -1 ) were daily prepared by dissolving Na 2 S 2 O 5 salt in water previously purged with nitrogen.To prepare diluted solutions of sulfite, small volumes of the stock solutions were properly added to air saturated water.
Cu(II)/peptide complex solutions, Cu II G n , were freshly prepared by dissolving an appropriate amount of the peptide in 20.0 mL of borate buffer solution followed by the addition of 0.2 mL of Cu(II) perchlorate.The final pH (7, 8, 9 and 10) was adjusted with 1.0 mol L -1 NaOH or 1.0 mol L -1 HCl solutions.Most of the working solutions (solution A) were: 2.0x10 -3 mol L -1 Cu II G n (with 25 % (5.0x10 -4 mol L -1 ) of peptide in excess over Cu II G n ) in 0.02 mol L -1 borate buffer solution.This working solution contains 2x10 -5 mol L -1 Ni(II).Ni(II) at this level concentration increases the reaction rate and the maximum absorbance at 365 nm.
The borate medium was chosen since the Cu(II) and Cu(III) complexes are stable in this medium.In phosphate buffers the Cu(III) formation is less effective as the anion may displace the peptide ligand.
The final concentration in air saturated solutions after mixing the reactants, with or without sulfite, are indicated in all figures.

Spectrophotometric measurements
Freshly prepared solutions were mixed prior to the experiments.An equal volume (1.0 mL) of sulfite solution was mixed with Cu II G n solution in borate buffer (1mL, solution A) in a Tandem spectrophotometric cell (optical path length = 0.875 cm).
The UV/VIS spectra were recorded on a HP8453 spectrophotometer, which was also used for kinetic measurements.Stopped flow data were acquired with a Pro-K.2000Stopped-Flow Mixing Accessory (Applied Photophysics).
In all experiments, air saturated solutions were employed for which the oxygen concentration can be considered 2.8x10 -4 mol L -1 . 8All UV/VIS spectra were recorded using water as a blank solution, since only the product, Cu III G n , absorbs in 365 nm.
A pHmeter Metrohm model 713 with a glass electrode (filled with saturated NaCl) was used in the pH measurements.The temperature was kept at 25.0 ± 0.1 °C .

Electrochemical measurements
Rotating ring-disk electrochemical experiments were carried out using an analytical rotator (AFMSRX) connected to an AFCBP1 bipotentiostat (Pine Instrument Company), recording current potential curves typically at a 50 mV s -1 potential scan rate with a data acquisition software made available by the manufacturer (ASWCV2, PineChem).The glassy carbon / glassy carbon ring-disk electrodes (AFMT29) had the following dimensions: disk radius = 0.5613 cm, inner radius of the ring = 0.6248 cm and outer radius of the ring = 0.7925 cm.The electrodes were polished using 0.3 µm alumina before using.A platinum wire and a Ag/AgCl (saturated NaCl) were used as counter and reference electrodes respectively.During rotating ring-disk electrode (RRDE) experiments, the disk electrode potential was scanned between the limits 0 and 0.6 V, the ring being maintained at 0.1 V to collect the material generated at the disk.0.1 mol L -1 KNO 3 in borate buffer was used as supporting electrolyte.

Results and Discussion
The Cu II G 4 complex is rapidly oxidized to Cu(III) in the presence of dissolved oxygen,Ni II G 4 (traces) and sulfite, with the simultaneous formation of sulfate.The Cu(III) complex can be followed at 365 nm (ε = 7120 mol -1 L cm -1 ).Anast and Margerum 6 proposed an autocatalytic mechanism, where the rate constant depends on the initial Cu(III) concentration.This species reacts with SO 3 2-to form the SO 3 •radical (equation 7) and the further reaction with O 2 gives SO 5 • -(equation 8).The initiation in the absence of Cu(III) may be due to the disproportionation of Cu(II) to Cu(I) and Cu(III), or Ni(III) and formation of peroxomonosulfate by the reaction (equation 9). 6The peroxomonosulfate can then oxidize Cu(II) to Cu(III) (equations 10 and 11).

Spectrophotometric studies of the sulfite induced autoxidation of Cu II G n
The absorbance changes at 365 nm of an air saturated solution of (1.0x10 -3 mol L -1 ) Cu II G 4 in the absence of sulfite (Figure 1A) indicate that the Cu(III) formation due to the spontaneous oxidation by dissolved oxygen (equation 1) is slow with an induction period of about 10000 s.The length of the induction period depends on the acidity and peptide, for instance at pH 7 (data not reported in the present work), the induction period is longer for G 5 and G 6 complexes, about 5.5 h, while for the G 4 it is about 3.5 h.
In order to evaluate the first fast reaction step, some kinetic measurements were performed on a stopped-flow instrument equipped with an online data acquisition system (Figure 1B, C and D). Figure 1B, C and D shows the absorbance changes at 365 nm after addition of sulfite to a Cu II G 4 solution in borate buffer (pH 9).At the indicated experimental conditions, two maximum peaks appear, one at 365 nm (Cu III G 4 complex) and other around 250 nm (Cu III G 4 and Cu II G 4 complexes).The same spectrum profile was observed for G 5 and G 6 complexes.
At 365 nm only Cu III G n species absorb 1,2,6 and the absorbance before and after sulfite addition is higher, when compared to the one at 250 nm, and better for analytical purposes.
Figure 1B, C and D shows that in the presence of sulfite and oxygen the Cu III G 4 formation is accelerated by sulfite, the induction period still exists (around 3 s) and depends on the initial Cu III G 4 concentration (around 10 -7 mol L -1 , due to the spontaneous oxidation with O 2 (equation 1)) and sulfite.This induction period is an evidence of autocatalytic behavior.
Figure 1B, C and D (data at the first 15 seconds) and Figure 2 (data at longer time) shows that one fast reaction ocurrs at the beginning, with Cu(III) formation, oxygen comsumption and simultaneous oxidation of S(IV) to S(VI) (equations 7-11).In some cases, after the fast first step, this process is followed by a slow absorbance increase (the oxidation of Cu II G n with oxygen still remaining in solution, equation 1, i.e., Figure 2B) and further decrease due to its decompositon (equations 2-4).The formation and decomposition depends on the oxygen concentration and acidity.
Cu(III) formation, calculated neglecting the induction period, can be obtained by the slope of ln(Absorbance) t vs. time, for several initial sulfite concentrations, [SO 3 2-] i .The slope increases with [SO 3 2-] i (Figure 3).According to Coichev and van Eldik 9 and Atkins 14 the later part the kinetics trace exhibits the maximum rate of Cu(III) formation.The curves show an induction period and autocatalytic behaviour.
Some conclusions can be done regarding to the pH dependence shown in the absorbance vs. time profile (Figure 2), which is a result of Cu(III) formation induced by SO 3 2-in the presence of oxygen (equations 7-11) and its further decomposition (equations 2, 4, 5 and 6).The pH dependence might be due to the [Cu II (H -x G n )] (1-x) species present in the solution, as a result of variable degree of protonation of the Cu(II) peptides complexes. 15Besides, shift in the HSO 3 -/SO 3 2-equilibrium (pK a = 6.3 16 ) will lead to an increase in the redox rate constant at pH > 6.According to Anast and Margerum, 6 the maximum net generation of [Cu III (H -3 G 4 )] -was found at pH 8, which also corresponds to the maximum rate of formation of [Cu III (H -3 G 4 )] -, the rate constant value decreasing at both higher and lower pH values.
As can be seen by the pK a values listed in Table 1, at pH range 7 to 10 the Cu(III) complex formed is present only as the triply deprotonated peptide complex ] -, such as, at pH 9-10 the [Cu I I (H -3 G 5 )] 2-and [Cu II (H -3 G 4 )] 2-are the predominant species in solution and the ratio [Cu II (H -2 G 4 )] -: [Cu II (H -3 G 4 )] 2-is about 1:1 at pH 9.This is in agreement with the higher maximum absorbance values obtained at pH 9 (Figure 2).As reported in previous studies, the decomposition of [Cu III (H -3 G 4 )] -, the predominant Cu(III) species at pH 7-9, is catalyzed by [Cu II (H -2 G 4 )] -(pKa = 9.14 17 ) (equations 5 and 6), 2 mainly at lower pH.
Figure 4 shows the effectiveness of Cu(III) complexes formation as the sulfite concentration increases.The maximum generation of Cu(III) peptide is attained at pH 9.
For analytical purposes, it is interesting to note that absorbance values at 365 nm are proportional to the initial SO 3 2-concentration (Figure 4).A flow injection procedure has already been developed for kinetic studies by measuring absorbance values at 365 nm originated from the formation of [Cu III (H -3 G 4 )] -. 18 These kinetic studies showed the catalytic effect of some transition metal ions on the oxidation of S(IV), which are likely to exist in environmental samples. 18The influence of formaldehyde on the kinetics of the S(IV) oxidation process was also addressed.The well-known property of this aldehyde in the stabilization of S(IV) in non complexant medium containing metallic ions being confirmed. 8The pseudofirst-order rate constants of sulfite consumption were determined in the presence of formaldehyde, and lower values were found at higher formaldehyde concentrations (1.0x10 -3 mol L -1 ). 18ectrochemical studies of the oxidation of Cu II G 4 in the absence of sulfite Some rotating ring-disk (RRD) voltammetry studies were carried out to characterize the formation of Cu III G 4 .Figure 5 shows a typical steady state voltammogram, recorded at the disk, for a 5.0x10 -4 mol L -1 Cu II G 4 solution in borate medium (pH 10) also containing 0.1 mol L -1 KNO 3 , at ω = 900 rpm; the corresponding signal at the ring (maintained at 0.1 V) is also presented.At this pH Cu II G 4 complex in the solution is present partially as [Cu II (H -2 G 4 )] - and predominantly as [Cu II (H -3 G 4 )] 2-, some Cu(OH) 2 precipitates, which does not interfere in the electrodes reactions.The Cu III G 4 generated complex is present as [Cu III (H -3 G 4 )] -.The following equations are responsible for the anodic and cathodic processes respectively: Cu III G 4 + eº Cu II G 4 at the ring ( 13) Since the half-wave potentials for both processes (oxidation and reduction) are very similar, it may be concluded that the electrochemical process related to the couple Cu(II)/Cu(III) in a medium containing G 4 is reversible.The half-wave potential found 0.66 V vs. NHE (Figure 5), for all these peptides it is in agreement with the literature (see Table 1).
Figure 6 shows the results of RRDE experiments performed at various rotation speed values in the 100 -4900 rpm range.The current signals obtained at relatively high rotation speeds (> 900 rpm) led to I ring / I disk values (N k collection efficiency) around 0.37, similar results being obtained by using an electroactive species characterized by fast electron transfer with no further chemical steps (Fe(CN) 6 4-) 21 as a probe.Therefore, in the Cu II/III G 4 system no chemical transformation involving the electrogenerated Cu III G 4 is noticeable at this time window (rotation speed > 900 rpm), suggesting that the Cu III G 4 decomposition could not be monitored at such short time.However, proportionally lower collection efficiency values were obtained at slower rotation rates, i.e., 0.26 at 100 rpm, demonstrating the possibility of a following chemical step involving the Cu(III) complex formed at the disk, which decomposition, according to the literature, 1,2     at higher pH values (also in agreement with the data in Figure 2) and relatively high Cu(III) concentration.
The RRDE experiments performed at various rotation speeds have a particular application in mechanisms study of electrochemical processes happening on two electrodes, it is possible the detection of unstable products (such as Cu III G 4 complexes) formed on the disk electrode.The advantage of these electrodes is the transport enhancement of the electroactive species to the ring electrode, leading to higher currents and therefore, to a higher sensibility and reproducibility. 22,23 cording to the literature, 24 for cases where the intermediate undergoes a first-order homogeneous chemical reaction a plot of N k as a function of ω -1 leads to a straight line whose slope contains the information on the first order kinetic constant (k).From the data shown in Figure 6, k was found to be 4.37x10 -3 s -1 at 25°C, ionic strenght 0.12 mol L -1 and pH 10 (borate buffer), from which t 1/2 was calculated as 159 seconds.The observed first-order rate constant reported in the literature 2 at pH 10.07 (0.05 mol L -1 carbonate buffer, 25°C), 1.12x10 -3 s -1 , is smaller.It can be explained by the different buffer composition and also depends on the concentration of oxygen and Cu(II) complexes (equations 3 and 5), which has being found to be responsible for the catalysis of the decomposition of Cu(III). 2,3 esides the studies of Margerum et al. 2,4 were carried out in a flow system, which consisted of an electrochemical flow cell, used to generate Cu III G 4 , followed by spectrophotometric measurements of Cu(III) decomposition.With such experimental design is difficult to evaluate the life time of short-live product (Cu II G 4 ).
In fact the RRDE experiments allowed the determination of first-order constant of Cu III G 4 decomposition, which would not be possible by spectrophotometric measurements.For instance Figure 2A (pH 9) shows the absorbance decrease due to Cu III G 4 decomposition.In this case Cu III G 4 was chemically generated by the reaction of Cu II G 4 with oxygen accelerated by sulfite, with simultaneous consumption of oxygen and sulfite (equation 1).The kinetics of the Cu III G 4 decomposition (equations 2-4) proved to be quite complex, and depends on Cu II G 4 and oxygen concentration.At the conditions in Figure 2A (pH 9) the oxygen concentration decreases during the global process of Cu(III) formation (equations 7-9) and decompositon (equations 2-4), and it will influence on the rate decomposition.Figure 2 also shows that the kinetic profile depends on the ligand (G n ).
Also from Figure 6, it can be concluded that the I vs ω 1/2 dependence does not correspond to a diffusion-controlled process since a clear deviation from a straight line is noticed at more intense hydrodynamic conditions. 25,26 his observation may be associated with the existence of an equilibrium between different Cu II G 4 species in solution which controls the overall electrode process.As is further discussed, from the data in Figure 7, [Cu II (H -3 G 4 )] 2-is the electroactive species in solution, such as the acid-base equilibrium (equation 14) is displaced by the fast oxidation of the electroactive species.Hence, for relatively long time windows the depletion of the electroactive species by the electrochemical oxidation process may be compensated by the conversion of the non-electroactive species near to the electrode surface.At high rotation speeds this replenishment is not fast enough and the current at the disc is pure kinetically controlled and independent of the rate of mass transport.This conclusion is in agreement with the electrochemical studies reported in literature on the electroxidation of copper (II) G 4 and G 5 complexes. 27u II (H -2 G 4 )]º [Cu II (H -3 G 4 )] 2-+ H + ( 14) non-electroactive electroactive The voltammograms obtained at the 7-10 pH range (Figure 7) clearly show the importance of the acid-base equilibrium of the [Cu II (H -x G n )] (1-x) species in solution.The current increases with the pH, that is, at higher pH the [Cu II (H -3 G 4 )] 2-species concentration also increases, such as it can be concluded that [Cu II (H -3 G 4 )] 2-is the electroactive species in solution.This evidence was also obtained from Figure 6.The conclusion that [Cu II (H -3 G n )] 2-must be more reactive than [Cu II (H -2 G n )] -was already discussed from the spectrophotometric results presented in Figure 2. concentration and the amount of sulfite added to the solution when experiments are performed in conditions of excess of both Cu II G 4 and oxygen in respect to sulfite. 18rom an analytical point of view this relationship could be useful for the development of an indirect method for sulfite.At the experimental condition reported in the Figure 4 (pH 9) the estimated detection limit was 2.0x10 -6 mol L -1 .By flow injection analysis this limit was 7.0x10 -6 mol L -1 . 18ing to the advantages of the amperometric detection as its excellent signal to noise ratio 26 (since the capacitive current is virtually zero at constant potential), some experiments were performed at E = 0.1 V, where the Cu III G 4 species is electroactive (Figure 8).Accordingly, Figure 9 shows the results for successive injections of sulfite solution to an aerated solution containing Cu II G 4 .The fast increase in the cathodic current after addition of relatively low amounts of sulfite to the working solution indicates the feasibility of an indirect method for determination of sulfite with electrochemical detection of the Cu III G 4 compound chemically generated.The proposed indirect approach is particularly relevant because the direct electroxidation of sulfite at bare electrode surfaces has been reported to be irreversible. 28Hence, drastic pretreatment procedures combined with very positive potentials 29,30 or the modification of the electrode surface by incorporating catalytic layers 31,32 have been suggested to enhance the sensitivity.In this way, owing to the good results noticed in Figure 9 the proposed method is being fitted for a flow injection configuration and the preliminary results are very promising.

Electrochemical studies of the sulfite induced autoxidation of Cu II G 4
The formation of a Cu III G 4 complex when sulfite is added to a Cu II G 4 solution under aerobic conditions was investigated electrochemically as shows Figure 8, where voltammograms were recorded at the rotating disc electrode (ω = 900 rpm).After the addition of sulfite a voltammogram with an anodic and a cathodic component is obtained, confirming that both Cu II G 4 and Cu III G 4 complexes exist in solution.A comparison of the voltammograms presented in Figures 5 and 8 suggests that the Cu(III) complex formed as a consequence of the induced sulfite oxidation is the same species electrochemically generated from Cu II G 4 .
Preliminary spectrophotometric studies have shown that there is a linear dependence between the Cu III G 4  Interesting mechanistic studies 9 and development of alternative analytical methods [10][11][12][13]33 based on metal ion catalyzed reactions, for determination of S(IV) in food, environmental samples and degraded hexafluoride have been already reported by our group.

Figure 6 .
Figure 6.Dependence of the limiting currents on the square root of the rotation rate in RRDE experiments.Solution: [Cu II G 4 ] = 5.0x10 -4 mol L -1 in 0.05 mol L -1 borate buffer (pH 10) and 0.1 mol L -1 KNO 3 .(A) anodic current measured at the disc, (B) cathodic current measured at the ring (E = 0.1 V).Collection efficiency values N k as function of rotation rate to calculate the first order constant.Slope = k = 4.37x10 -3 s -1