Selective oxidative degradation of azo dyes by hydrogen peroxide catalysed by manganese(II) ions

Abstract

Manganese(II) ions catalyse the oxidative degradation of Calmagite (H₃CAL) dye in aqueous solution at 20 ± 1 °C in the pH range 7.5–9.0 using hydrogen peroxide (H₂O₂) as oxidant by a mechanism that involves strong complexation to the Mn^II centre. It is proposed that [Mn^III(CAL)(O₂H)]⁻ i.e. a dye coordinated hydroperoxyl (O₂H⁻) Mn^III complex is formed and bleaching of the dye is initiated by an electron-transfer to Mn^III, with the binding of H₂O₂ being the rate determining step. At pH 9.0 in (bi)carbonate, HCO₃⁻, H₃CAL is rapidly bleached via the in situ formation of coordinated peroxycarbonate (HCO₄⁻); a TOF (TOF = moles of dye bleached per mole of manganese per hour) of ∼5000 h⁻¹ can be achieved. The bleaching of the related azo dyes Orange II and Orange G is different because, unlike Calmagite, they lack an o,o-dihydroxy motif so are unable to complex strongly to Mn^II and no oxidation to Mn^III occurs. At pH 8.0 (phosphate buffer) Orange II and Orange G are not bleached but bleaching can be achieved at pH 9.0 (HCO₃⁻ buffer); the rate determining step is dye coordination and it is proposed bleaching is achieved via an outer-sphere oxygen atom transfer. Mechanisms for dye bleaching at pH 8.0 and pH 9.0 are proposed using data from EPR, UV/VIS and ESI-MS. Mn^II/H₂O₂/HCO₃⁻ form a potent oxidising mixture that is capable of removing stubborn stains such as curcumin.

---

Introduction

Higher valence manganese complexes have been implicated in a wide variety of natural, biomimetic and synthetic processes involving both substrate oxidation and reduction including photosynthetic water oxidation, dioxygen activation and reduction to hydrogen peroxide (H₂O₂), epoxidation and bleaching. The use of H₂O₂ as the terminal oxidant in oxidation processes is favoured due to its environmentally benign nature, the byproduct in its use being water.¹ From a green chemistry viewpoint it is the ideal oxidant with an atom efficiency (47%), bettered only by dioxygen (O₂, 50%), but exhibiting better reaction kinetics. Large scale uses of H₂O₂ include the bleaching of paper in the wood pulp industry, as a detergent in the form of perborates and percarbonates in washing powders, and in the chemical treatment of water systems.²

The treatment of wastewaters from dye manufacturers and the textile processing industries is a challenge³ with azo dyes which contain the –N=N– moiety being particularly problematic as they account for 70% of textile dyestuffs and are also chemically stable.⁴ It has been estimated that 128 tonnes of dye materials are released daily into the environment and by the nature of their use they are designed to be resistant to environmental degradation.^4,5 Thus they are non-biodegradable under aerobic conditions and under anaerobic conditions produce hazardous intermediates.⁶ The oxidative degradation of Orange II (sodium 4-[(2-hydroxy-1-naphthyl)azo]benzene sulfonate, Scheme 1), a model azo dye substrate, has been described by van Eldik and co-workers using H₂O₂ with either Mn^II salts⁷ or Mn^II(bipy)₂ complexes⁸ in carbonate (HCO₃⁻) buffer in the pH range 8–10, and Mn^II salts using commercial peracetic acid⁹ at pH 9.5.

Scheme 1 Azo-hydrazone tautomerism of Calmagite, Orange II and Orange G in aqueous solution.

In the use of Mn^II salts, van Eldik found that only the HCO₃⁻ buffer system was able to oxidatively degrade Orange II and he proposed a mechanism involving peroxycarbonate (HCO₄⁻) that is formed when H₂O₂ reacts with HCO₃⁻ (Scheme 2).

Scheme 2 Mechanism for the formation of peroxycarbonate from carbonate and hydrogen peroxide.

HCO₄⁻ ions are reported to be several orders of magnitude more reactive than H₂O₂ towards nucleophilic substrates¹⁰ because carbonate (CO₃²⁻) is a better leaving group than hydroxide (OH⁻).¹¹ In the presence of Mn^II, HCO₄⁻ is complexed to produce a Mn^II-η²-peroxycarbonate complex which van Eldik proposes then breaks down to Mn^IV=O that is the active catalyst in the oxidative degradation of Orange II through the transfer of an oxygen atom. Van Eldik suggests this transfer is assisted by the weak binding of Orange II to Mn^II and that this complexation stabilises the Mn^II-η²-peroxycarbonate pre-catalyst (Scheme 3).

Scheme 3 Proposed mechanism for the oxidative degradation of O II via the in situ formation of Mn^IV=O species.

We have previously described the complete degradation of Calmagite (3-hydroxy-4-(2-hydroxy-5-methylphenylazo)naphth-alene-1-sulfonic acid) under ambient conditions using in situ generated H₂O₂ catalysed by Mn^II ions at pH 8.0 in N-2-hydroxyethylpiperazine-N′-3-propanesulfonic acid, EPPS, buffer.¹² Under these conditions Orange II and Orange G (3-hydroxy-4-(phenylazo)naphthalene-1,5-disulfonate, disodium salt) were not bleached and it was suggested that this was because these substrates lacked an additional hydroxyl group ortho to the azo group (Scheme 1) thus resulting in weaker binding to Mn^II. The presence of Tiron (1,2-dihydroxybenzene-3,5-disulfonate disodium salt, monohydrate, Na₂TH₂·H₂O), with [Tiron]/[dye] ∼ 15 was required for efficient in situ generation of H₂O₂. The mechanism for oxidative degradation of Calmagite was proposed to involve a Mn^III–OOH species within a highly oxidizing Mn^III Tiron–quinone complex, with no evidence of Mn^IV=O species by EPR. While the in situ generation of H₂O₂ from O₂ for dye bleaching is highly attractive, the presence of hydroxylamine (NH₂OH) in stoichiometric amounts as the reducing substrate and Tiron as an essential co-ligand makes the application of this system unsuitable for wastewater treatment. We were therefore interested to use simple Mn^II salts in the presence of added H₂O₂ to investigate the oxidative degradation of Calmagite, Orange II and Orange G as model substrates in the pH range 7.5–9.0 using both phosphate and carbonate buffer systems. These substrates were chosen because they represent a range of Mn^II binding abilities viz. Calmagite ≫ Orange II ∼ Orange G. In this study we wish to detect the differences in the mechanisms of dye bleaching at pH 8.0 (phosphate) and pH 9.0 (carbonate) and to determine if we could detect Mn^IV=O species in the latter system by EPR, as previously reported by van Eldik.⁷ The use of H₂O₂ provides a cheap and more atom economic oxidant compared to peracetic acid (and other peracids) and hypochlorite (which also releases chlorine into the environment). There are a wide range of metal systems known to catalyse the degradation of azo dyes but these are associated with elaborate ligand systems which add expense and chemical demands in the treatment of wastewater. The use of simple Mn^II salts in these systems is viewed as beneficial in terms of economics and atom efficiency, thus realising the goals of green chemistry.

Results and discussion

Calmagite

Fig. 1 is a visible spectrum scan (400–700 nm) showing the change in absorbance of an aqueous solution of Calmagite (H₃CAL henceforth referred to as ‘CAL’, 0.100 mM, λ_max ∼ 540 nm) at pH 8.0 ± 0.1 (phosphate buffer) and 20 ± 1 °C in the presence of added H₂O₂ (50.0 mM) and aqueous Mn²⁺ (5.00 μM) as catalyst (initial molar ratio Mn : CAL : H₂O₂ = 1 : 20 : 10 000).

Fig. 1 The change in absorbance spectrum of CAL at 5 min intervals from t = 0 to t = 20 min at 20 ± 1 °C with [MnCl₂·4H₂O] at 5.00 μM. Initial [CAL] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 8.0 ± 0.1 (50 mM, phosphate buffer). Insert: the colour of CAL before (left) and after (right) the reaction.

There is an initial rapid decrease in [CAL] (t = 0–5 min) followed by smaller decreases [CAL] with time. At the end of the 20 min period there is no trace of the purple colour in the reaction solution (Fig. 1 (insert)). Previous studies on the in situ generation of H₂O₂ and CAL dye bleaching were carried out at pH 8.0 using the non-coordinating, biological, EPPS buffer; the use of phosphate buffer here seems not to have any detrimental affect on the ability of the Mn^II to catalyse this reaction suggesting that phosphate can easily be displaced by CAL in coordination sites around Mn^II. In this coordinated state it would be expected that the azo form of CAL would dominate over the hydrazine tautomer (Scheme 1) with electron density concentrated on the oxygen atom.¹² The effect of the variation of [Mn^II] between 0 and 100 μM on the oxidative degradation of CAL is shown in Fig. 2 and ESI1.†

Fig. 2 The change in [CAL] with time monitored at 540 nm and at 20 ± 1 °C with [MnCl₂·4H₂O] varied from 0, and with added ethylenediaminetetraacetic acid (EDTA, 1.00 mM), to 100 μM. Initial [CAL] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 8.0 ± 0.1 (50 mM, phosphate buffer).

In the absence of added Mn^II there is only slow bleaching of CAL and this is stopped almost completely in the presence of ethylene diamine tetra-acetate (EDTA⁴⁻, 1.00 mM), and is presumably due to the presence of adventitious Mn^II. [Mn^II] as low as 0.500 μM catalyses the bleaching of CAL. At 5.00 μM and lower concentrations of Mn^II, the rate of bleaching is independent of the [CAL] ([CAL]/[Mn^II] ≥ 20 initially) suggesting that coordination of CAL to Mn^II is fast. There is a first order dependency on [Mn^II] (ESI1†) with the pseudo first order rate constant calculated to be 3.9 × 10⁻² s⁻¹ and with a TOF of ∼140 h⁻¹. There are small increases in the rate of CAL bleaching when the pH is raised from 7.5 to 8.0 and then 8.5. This is consistent with an increase in the [HO₂⁻] (pK_a H₂O₂ = 11.67)¹⁴ with relative [HO₂⁻] at pH 7.5, 8.0 and 8.5 of 1 : 3.3 : 10 respectively and this suggests that the coordination of H₂O₂ may be the rate determining step. Unsurprisingly, the rates of reaction and TOF observed here are much faster than that observed previously using in situ generated H₂O₂ (from O₂) where the equilibrium concentration of oxidant is lower and competition for binding sites around the manganese greater due to the presence of reducing substrate hydroxylamine (NH₂OH, 100 mM) and ligand Tiron (1.50 mM).¹³ A first order dependency of CAL bleaching on [H₂O₂] (ESI2†) and [CAL] (ESI3†) was observed but the rate of bleaching was found to be independent of the [phosphate]. To gain insight into the chemical species formed during the reaction cycle, we employed electron paramagnetic resonance (EPR) spectroscopy. For semi-integer spin systems, e.g. Mn^II S = 5/2, the highest EPR sensitivity (i.e. higher transition probability) is achieved with the external field and the microwave field perpendicular to each other. For integer spin systems, e.g. Mn^III S = 2, the transition probability drops to zero when the external field and the applied magnetic field are perpendicular, but is largest when they are parallel. The EPR spectrum (X-band, perpendicular mode) of a solution of Mn^II (100 μM) at pH 8.0 (phosphate buffer) containing H₂O₂ (50.0 mM) revealed the characteristic six-line spectrum (ESI4†) of the high-spin ion (S = 5/2) coupled to the manganese nucleus (⁵⁵Mn, I = 5/2). When the EPR experiment was repeated in both perpendicular and parallel modes after the addition CAL (0.100 mM), the six line spectrum (perpendicular mode) was still present (Fig. 3) but there was no evidence for the presence of Mn^III species (parallel mode) nor was there any signal at ∼150 mT (g ∼ 4.5) in perpendicular mode indicative of Mn^IV=O (S = 3/2).

Fig. 3 X-band EPR spectra of MnCl₂·4H₂O (100 μM), H₂O₂ (50.0 mM) and CAL (0.100 mM) at pH 8.0 (50 mM, phosphate buffer) in both perpendicular (9.65 GHz) and parallel (9.41 GHz) modes. EPR conditions: 10 K, 2 mW microwave power, modulation amplitude 0.5 mT.

On the other hand, Mn^III was observed in the EPR (parallel mode) when an aqueous solution of MnCl₂·4H₂O (0.10 mM) was added to aqueous Tiron (∼0.30 mM) at pH 8.0 (50 mM EPPS) to form [Mn^III(T)₂]⁵⁻ in situ by aerial oxidation¹⁵ (Fig. 4). The measured signal is faint and does not display any resolved hyperfine structure, nonetheless its resonance position (g ∼ 8.1) is consistent with previously reported Mn^III signals.¹⁶

Fig. 4 X-band EPR spectra of MnCl₂·4H₂O (100 μM) and Tiron (∼0.30 mM) at pH 8.0 (50 mM, phosphate buffer) followed by the addition of H₂O₂ (50.0 mM) and then CAL (0.100 mM). EPR conditions: parallel mode, 9.41 GHz, 10 K, 2 mW microwave power, modulation amplitude 0.5 mT.

When H₂O₂ (50 mM) was added the Mn^III signal was still observable, whereas when CAL (0.100 mM) was introduced the signal disappeared. This suggests that in the presence of excess CAL (and H₂O₂) the equilibrium [Mn^III] is low as it is reduced to Mn^II due to a one-electron transfer from bound CAL, which in turn initiates the oxidative degradation of the dye. Returning to the added H₂O₂ system (without Tiron), Fig. 5 shows the results of a semi-quantitative EPR study at pH 8.0 (phosphate buffer) in which the height of the peak at lowest magnetic field (∼315 mT) was used as a proxy for the [Mn^II] present.

Fig. 5 X-band EPR spectra of MnCl₂·4H₂O (100 μM), H₂O₂ (50.0 mM) and CAL (0.100 mM) at pH 8.0 (50 mM, phosphate buffer) in perpendicular (9.45 GHz) mode. EPR conditions: 10 K, 2 mW microwave power, modulation amplitude 0.7 mT.

When only H₂O₂ was added there was no observable change in [Mn^II], indicating that this state is quite stable despite the presence of an excess of a strong oxidant. When the experiment was repeated with CAL but no H₂O₂, there was a large decrease in Mn^II signal, possibly due to the formation of the Mn^III–CAL complex. Like the Mn/Tiron system, this d⁴ state is presumably stabilised by the CFSE now gained by the coordination of a relatively strong-field ligand. With both CAL and H₂O₂ present, there is a ∼50% reduction in Mn^II signal, and the assumed ∼50% Mn^II/Mn^III present is due to the presence of both reduced and oxidised Mn species as coordinated CAL is 1−e⁻ oxidised. None of the EPR data shows any evidence of radical species and when CAL bleaching was repeated in the presence of an excess of the radical scavenger 2,4,6-tri-tert-butylphenol there was no reduction in the rate of oxidative degradation suggesting that the mechanism does not involve radical species e.g. ˙OH and ˙O₂H. Fig. 6 shows an UV/VIS spectrum of a mixture of Mn^II with H₂O₂ (pH 8.0, phosphate buffer) exhibiting an absorption peak at ∼240 nm due to the formation of a Mn^II–hydroperoxo complex; this peak is substantially broadened and red shifted on the addition of CAL and the peak decreases only slowly over a period of 30 min. The change in the UV part of the spectrum on addition of CAL is presumably due to the binding of CAL to the metal centre and subsequent oxidation of Mn^II to Mn^III and the very small changes in the UV region thereafter suggests that as CAL is oxidatively decomposed there is fast coordination of remaining CAL to the metal centre, with rapid mineralisation of the initial oxidation products of the dye.

Fig. 6 UV/VIS spectrum of a CAL bleaching solution at 20 ± 1 °C before and after the addition of CAL with then repeat scans until t = 30 min. [MnCl₂·4H₂O] was 5.00 μM and the initial [CAL] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 8.0 ± 0.1 (50 mM, phosphate buffer).

Taken together, the kinetic, UV/VIS and EPR data are consistent with a mechanism for CAL bleaching that involves the oxidative degradation of coordinated CAL in the presence of coordinated HO₂⁻ at a Mn^III centre (eqn (1)–(4)).

Mn^II + H₃CAL ⇌ [Mn^II(CAL)]⁻ + 3H⁺ (1)

2[Mn^II(CAL)]⁻ + ½O₂ + 2H⁺ ⇌ 2[Mn^III(CAL)] + H₂O (2)

[Mn^III(CAL)(O₂H)]⁻ → Mn^II + CAL oxid. products (4)

Step (4) can be considered to be fast given the lack of observable Mn^III in the EPR parallel mode and the persistence of observable Mn^II in EPR perpendicular mode. The role of H₂O₂ as the terminal oxidant in this system is interesting in that it could be used simply to reduce Mn^III back to Mn^II in a catalase-like process, but the system is probably more complicated with coordinated peroxide taking an active role in the initiation of CAL breakdown. The presence of excess H₂O₂ in solution is probably responsible for further oxidation of the initial oxidation product resulting in complete the mineralisation of the dye.¹³ Oakes et al. have also proposed that specific complexation of CAL is a requirement for its oxidative degradation when using peroxosulfate as terminal oxidant and suggested that this may occur via a one-electron transfer from Mn^III involving an inner-sphere mechanism.¹⁷ The rate of loss of CAL is given by eqn (5).

−d[CAL]/dt = k₁[Mn(CAL)(O₂H)] (5)

When there is an excess of H₂O₂ in solution and the [CAL] is large, the rate of CAL bleaching can be expressed as eqn (6).

−d[CAL]/dt = k_obs[Mn^II] (6)

We were interested to investigate the effect of Tiron on CAL dye bleaching at pH 8.0 (50 mM EPPS) as this ligand was previously found to be essential not only for the in situ generation of H₂O₂ but also for CAL bleaching.⁹ In the absence of added Na₂CO₃ with [Tiron⁴⁻] at 1.50 mM (Mn : Tiron⁴⁻ : CAL : H₂O₂ = 1 : 30 : 2 : 500), there is a clear lag period during which time CAL bleaching is inhibited. However, this lag period is reduced when sodium carbonate was added and there were step-wise reductions as the [Na₂CO₃] is increased to 50.0 mM (Fig. 7).

Fig. 7 The change in [CAL] with time monitored at 535 nm and at 20 ± 1 °C in the presence of Tiron with [Na₂CO₃] varied from 0 to 50 mM and with [MnCl₂·4H₂O] at 50.0 μM. Initial [CAL], [H₂O₂] and [Tiron] were 0.100 mM, 50.0 mM and 1.50 mM respectively. The pH was 8.0 ± 0.1 (50 mM, N-2-hydroxyethylpiperazine-N′-3-propanesulfonic acid, EPPS, buffer).

In the absence of added Na₂CO₃ these results are consistent with the need for CAL to be bound to Mn^II for bleaching to occur and that Tiron⁴⁻ can initially compete effectively for binding sites around the metal. When Na₂CO₃ was added the in situ formation of peroxycarbonate (HCO₄⁻) is itself able to compete with Tiron for binding sites around Mn^II ([HCO₄⁻]/[Tiron] ∼ 6.7–33) and this leads to a progressive reduction in the lag period as the [HCO₄⁻] is increased.

Fig. 8 shows a comparison of the rates of CAL bleaching at pH 8.0 ± 0.1 (50 mM phosphate) and at pH 9.0 ± 0.1 (50 mM carbonate, λ_max ∼ 608 nm) with [Mn^II] at 0.500 μM.

Fig. 8 The change in [CAL] with time monitored at 535 nm at pH 8.0 ± 0.1 (50 mM, phosphate buffer) and 608 nm (50 mM, carbonate buffer) at 20 ± 1 °C with [MnCl₂·4H₂O] at 0.500 μM. Initial [CAL] and [H₂O₂] were 0.100 mM and 50.0 mM respectively.

The TOF at pH 9.0 is calculated to be ∼5000 h⁻¹ which is ∼35 times larger than at pH 8.0. These higher bleaching rates cannot be solely accounted for by the higher [HO₂⁻] at this pH (∼10 times higher than at pH 8.0) and it would appear that the binding of HO₂⁻ at pH 9.0 is much faster than the binding of H₂O₂ in phosphate buffer at pH 8.0 and this leads to the faster formation of [Mn^III(CAL)(O₂H)]⁻ and can account for the faster bleaching rate under these conditions (Scheme 4).

Scheme 4 Proposed mechanism for the oxidative degradation of CAL at pH 8.0 (phosphate, black) and pH 9.0 (carbonate, red).

The reaction at pH 9.0 was much more sensitive to adventitious Mn^II with slow bleaching in the absence of added Mn^II; this activity was again effectively reduced to zero in the presence of EDTA⁴⁻ (1.00 mM). Thus while carbonate/peroxide mixtures resulting in the in situ formation of HCO₄⁻ may be viewed as strongly oxidising, accelerated rates under ambient conditions may only be observed in the presence of metal ions.

The addition of Tiron to CAL bleaching at pH 9.0 (EPPS buffer) with varying [Na₂CO₃] is shown in Fig. 9.

Fig. 9 The change in [CAL] with time monitored at 608 nm and at 20 ± 1 °C in the presence of Tiron with [Na₂CO₃] varied from 0 to 50 mM and with [MnCl₂·4H₂O] at 50.0 μM. Initial [CAL], [H₂O₂] and [Tiron] were 0.100 mM, 50.0 mM and 1.50 mM respectively. The pH was 9.0 ± 0.1 (50 mM, N-2-hydroxyethylpiperazine-N′-3-propanesulfonic acid, EPPS, buffer).

In general the lag periods were found to be smaller than at pH 8.0 and the rates of dye bleaching faster with saturation at ∼50 mM Na₂CO₃ consistent with the maximum stoichiometric amount of HCO₄⁻ that can be formed in solution at pH 9.0 (Scheme 2).

Orange II and Orange G

Fig. 10 and 11 show repeat UV/VIS scans (300–700 nm) of aqueous solutions of Orange II (O II, 0.100 mM, λ_max ∼ 484 nm) and Orange G (O G, 0.100 mM, λ_max 486 nm) respectively at pH 9.0 ± 0.1 (carbonate buffer) and at 20 ± 1 °C in the presence of added H₂O₂ (50.0 mM) and aqueous Mn²⁺ (5.00 μM) as catalyst.

Fig. 10 The change in [O II] from t = 0 to t = 10 min and then at 5 min intervals to 45 min monitored at 484 nm and at 20 ± 1 °C with [MnCl₂·4H₂O] at 5.00 μM. Initial [O II] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 9.0 ± 0.1 (50 mM, carbonate buffer). Insert: the colour of O II after attempted oxidative degradation at pH 8.0 (left, no change) and at pH 9.0 (right, completely bleached).

Fig. 11 The change in [O G] from t = 0 to t = 60 min at 5 min intervals to 45 min monitored at 486 nm and at 20 ± 1 °C with [MnCl₂·4H₂O] at 5.00 μM. Initial [O G] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 9.0 ± 0.1 (50 mM, carbonate buffer). Insert: the colour of O G after attempted oxidative degradation at pH 8.0 (left, no change) and at pH 9.0 (right, completely bleached).

For these dyes not only is there a decrease in the absorbance of the broad colour chromophore centered at ∼484 nm (and shoulder at ∼500 nm in the case of O II) but also a decrease in the absorbance of the naphthalene ring absorbance at ∼320 nm, suggesting complete mineralisation of the dyes. These results are different to that obtained by Nunes et al. using manganese(II) hydroxybenzyl-pyridyl-diamine complexes in the oxidative degradation of methyl orange where there was an increase in the absorbance at ∼330 nm attributed to the formation of the amine oxide.¹⁸ When the bleaching of O II and O G was attempted at pH 8.0 (phosphate or EPPS, Fig. 10 and 11, inserts), there was no change in the colour of the dyes over a period of ∼30 min. Van Eldik also reported that O II was not bleached in any other of a range of buffer systems (including phosphate) in the pH range 8–9.⁷ Fig. 12 and 13 show the effect of the variation of [Mn^II] between 0 and 100 μM on the bleaching of O II and O G respectively.

Fig. 12 The change in [O II] with time monitored at 484 nm and at 20 ± 1 °C with [MnCl₂·4H₂O] varied from 0 to 100 μM. Initial [O II] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 9.0 ± 0.1 (50 mM, carbonate buffer).

Fig. 13 The change in [O G] with time monitored at 486 nm and at 20 ± 1 °C with [MnCl₂·4H₂O] varied from 0 to 100 μM. Initial [O II] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 9.0 ± 0.1 (50 mM, carbonate buffer).

In the absence of added Mn^II there is slow adventitious Mn^II bleaching of the dyes and this is virtually stopped in the presence of EDTA⁴⁻ (1.00 mM). [Mn^II] as low as 1.00 μM catalyses the bleaching of the dyes and there is an unexpected ∼½ order dependency on [Mn^II] (ESI5†) for both dyes with pseudo first order rate constants for O II and O G calculated to be 1.4 × 10⁻² s⁻¹ and 6.2 × 10⁻³ s⁻¹ respectively giving TOFs of ∼50 h⁻¹ and ∼20 h⁻¹ respectively. These TOFs is ∼100 times lower than that for CAL bleaching at pH 9.0 using carbonate buffer and this reflects the lower binding constants of O II and O G to Mn^II compared to CAL. Again, rather unexpectedly there are similar fractional order dependencies on [H₂O₂] (ESI6†) and HCO₃⁻ (ESI7†) of ∼½ and ∼3/4 for both dyes. These fractional order dependencies suggest that the catalytic system is complicated and merit further investigation. A UV/VIS spectrum of solution an O G bleaching solution is shown in Fig. 14. Before the addition of O G there is a peak centered at ∼245 nm presumably due to the Mn^II-η²-peroxycarbonate complex. This peak is broadened and red shifted on the addition of O G and there is then a decrease in the absorption of this peak over a period of 60 min. At the end of 60 min (when all the O G would be exhausted), the UV/VIS spectrum is almost identical to the spectrum containing Mn^II and H₂O₂ only indicating that the change in spectrum is reversible and is due to the weak binding of O G to the metal centre. A very similar spectrum was obtained with O II. These changes are presumably due to the formation and subsequent decomposition of a manganese–peroxycarbonate species.

Fig. 14 UV/VIS spectrum of a O G bleaching solution at 20 ± 1 °C before and after the addition of O G with then repeat scans until t = 60 min. [MnCl₂·4H₂O] was 5.00 μM and the initial [O G] and [H₂O₂] were 0.100 mM and 50.0 mM respectively. The pH was 9.0 ± 0.1 (50 mM, carbonate buffer).

The EPR spectrum of a solution of Mn^II (100 μM) at pH 9.0 (HCO₃⁻ buffer) exhibits the usual six-line pattern, which appears to be broadened and exhibit distinctive anisotropy as compared to its counterpart in phosphate buffer, presumably because of the coordination of the HCO₃⁻ anion (ESI4†). On addition of H₂O₂ (50.0 mM) the six-lines pattern sharpens up but reduces, however there is no evidence for a signal at g ∼ 4.5 (∼150 mT) characteristic of Mn^IV=O. This six line spectrum is again prominent on the addition O II (0.100 mM), yet there is no evidence for the formation of Mn^IV=O species at 150 mT (g ∼ 4.5) in perpendicular mode nor Mn^III in parallel mode at 100 mT (ESI8†), nor any decrease in the [Mn^II] in a bleaching solution containing O II, as judged from the height of the peak at lowest magnetic field at ∼315 mT (ESI9†). These results are at odds with the mechanism proposed by van Eldik who implicated Mn^IV=O species for O II bleaching in carbonate buffer at pH 8.5.⁷ However, van Eldik did not carry out any EPR studies in his work and so the involvement of Mn^IV=O in these systems is somewhat speculative. A similar set of EPR spectra were obtained when O II was replaced by O G. This work suggests that Mn^IV=O is either not formed or is a very transient species in this system. When O II and O G bleaching were repeated in the presence of an excess of 2,4,6-tri-tert-butylphenol there was no reduction in the rate of dye bleaching. This implies that the mechanism does not involve radical species. The rate determining step in the oxidative degradation of O II and O G is the coordination of the dye substrate to Mn^II. Analogous to Scheme 4 for CAL in the presence of carbonate, a mechanism can be proposed for the bleaching of O II and OG that in the last step involves an oxygen atom transfer to dye substrate (S) from coordinated HO₂⁻ (Scheme 5).

Scheme 5 Proposed mechanism for the oxidative degradation of O II and O G via the in situ formation of Mn^II–O–O–H species.

The final step in the oxidation of dye substrate is similar to the mechanism proposed by van Eldik but in this outer coordination sphere process there is no requirement for the formation of Mn^IV=O species. The fractional order dependencies on Mn^II, H₂O₂ and HCO₃⁻ may be explained by the complex series of equilibria in the formation of HCO₄⁻ and its consumption in the release of CO₂ and in the formation of Mn^II–OOH. This mechanism does require a change in oxidation state of Mn^II and fits the EPR data obtained. Manganese is effectively acting as a binding centre enabling transfer of an oxygen atom from bound oxidising agent to bound dye substrate. The differences in the rate of dye bleaching and TOF values between CAL, O II and O G at pH 9.0 (carbonate buffer) may be explained by the differences in their complexation to the manganese centre in the stabilisation of the in situ formed dye–manganese–peroxycarbonate species. The better a dye is able to bind to Mn^II and stabilise this species (with oxidation to Mn^III with CAL) the greater the rate of oxidative degradation (Table 1).

Table 1 Summary of the rate constants (k) and turnover frequencies (TOF) at 20 ± 1 °C for the manganese(II) catalysed oxidative degradation of dyes using hydrogen peroxide. In all cases the initial [Dye] was 0.100 mM, [buffer] was 50 mM, [H₂O₂] was 50.0 mM and [MnCl₂·4H₂O] was 5.00 μM

Fig.	Dye	pH/buffer	k/s^−1 (TOF/h^−1)^a
a TOF = turnover frequency = moles of dye oxidised per mole of MnCl2·4H2O per hour.
1/ESI1	Calmagite	8.0/phosphate	3.9 × 10^−2 (140)
8	Calmagite	9.0/carbonate	1.4 × 10^0 (5000)
12/ESI5	Orange II	9.0/carbonate	1.4 × 10^−2 (50)
13/ESI5	Orange G	9.0/carbonate	6.2 × 10^−3 (20)

---

We have determined the stability constants for the complexation of Mn^II with O II and O G based on the reduction in the wavelength maxima for 0.0500 mM concentrations of the dyes on the addition of 10.0–60.0 μM concentrations of MnCl₂·4H₂O (ESI10†). The log stability constants for O II and O G were calculated to be identical at 4.31 which is very close to a previously reported value 4.5⁷ for O II. The log stability constant for the coordination of Mn^II to CAL has a value of 11.5.¹⁹ Therefore the binding constant of CAL for Mn^II is ∼10⁷ greater than that for O II and O G and this can account for its ability to be rapidly bleached by an inner-sphere electron transfer from Mn^III. With O II and O G their weak binding ability to Mn^II means that Mn^III is most likely not formed they cannot be bleached at pH 8.0 (phosphate buffer); at pH 9.0 (carbonate buffer) oxidative degradation occurs through a more difficult outer-sphere oxygen atom transfer. The lower bleaching rates observed with O G compared to O II could be due to the presence of two sulfonate groups in its structure that reduce is binding ability still further by reducing the electron density at the phenolic oxygen anion and possibly the lone pairs of electrons on the two nitrogen atoms.

Conclusions

This study has shown that a simple Mn^II salt is able to efficiently activate H₂O₂ and H₂O₂/HCO₃⁻ mixtures to oxidatively degrade model dyes under ambient conditions. Dyes which strongly complex with the metal e.g. CAL may be degraded by a one-electron oxidation reactions at both pH 8.0 (phosphate) and pH 9.0 (carbonate). O II and O G are not bleached at pH 8.0 (phosphate) but can be degraded at pH 9.0 (carbonate) through the in situ formation of [S⋯Mn^II(OOCO₂)] and oxidation occurs through an oxygen atom to the dye substrate from [S⋯Mn^II(OOH)]. Direct evidence for the formation of peroxycarbonate (HCO₄⁻) and Mn^II-η²-peroxycarbonate [Mn^II(OOCO₂)] previously suggested by van Eldik has proved elusive. The 5 nm difference in the λ_max for solutions of Mn^II in H₂O₂ at pH 8.00 (phosphate, ∼240 nm) and pH 9.00 (carbonate, ∼245 nm) is too small to make a conclusive assignment. The identification of these species in solutions remains an important goal for future work but nevertheless the formation of these (or similar) species do fit with the mechanism proposed for the carbonate system at pH 9.0. The differences in the rates of bleaching between CAL (pH 8.0, phosphate) and O II (and O G, pH 9.0 carbonate) can be compared (ESI11†) through successive additions of fresh portions of CAL and O II to the reaction solutions, which also show only small reductions in the rates of dye bleaching at each step, suggesting that the integrity of the catalyst is preserved during these five repeated cycles. Work in these laboratories has shown that even difficult to remove stains such curcumin (curry stain) may be degraded using H₂O₂/HCO₃⁻ mixtures at pH 9.0 which provide a potent oxidising medium in the presence of added Mn^II. The tailoring of dye bleaching to the nature of the oxidising reagent has important implications in laundry formulations in the need to remove stain molecules while not bleaching fabric dyes. One application of these findings is the development of simple colour tests for the selective detection of H₂O₂ vapours e.g. in peroxide-based explosives.²⁰

Experimental section

Manganese(II) chloride tetrahydrate (AnalaR, 99.5%, BDH), Tiron (1,2-dihydroxybenzene-3,5-disulfonate, disodium salt, monohydrate, Sigma-Aldrich), Calmagite, Orange G (Sigma-Aldrich), Orange II (Fluka), hydrogen peroxide (30% w/w, Sigma-Aldrich), N-2-hydroxyethylpiperazine-N′-3-propane-sulfonic acid (EPPS, Sigma-Aldrich), sodium hydrogen carbonate (AnalaR, Sigma-Aldrich), potassium dihydrogen phosphate (AnalaR, BDH), sodium hydroxide pellets (semiconductor grade, 99%, Sigma-Aldrich) and 2,4,6-tri-tert-butylphenol (Sigma-Aldrich) were used as received. De-ionised water (ELGA Purelab) was used in all experiments and plastic spatulae were used to transfer solid reagents. (Bi)carbonate solutions were made up in boiled water (to remove dissolved CO₂).²¹

Catalytic oxidations of the dyes were carried out at pH 7.5–9.0 and 20 ± 1 °C in the presence of an excess of H₂O₂ over dye (∼500) so that rates were not influenced by changes in [H₂O₂]. In experiments used for subsequent kinetic analysis the dye : Mn ratio was high (≥20) in order that absorbance readings reflected the concentration of uncomplexed dye thus simplifying the analysis. A 500 mL four-necked Pyrex container was used that was specially designed to be narrow in the section containing the reaction solution so as to prevent excessive frothing of the dyes. The reaction solution was circulated using a sipper pump (JENWAY) via silicone rubber tubing through a 1 mm glass flow cell housed in a JENWAY 6315 (scanning) UV/VIS spectrophotometer and then back to the reaction vessel. The flow was at a rate such that there was no significant temperature loss between the reaction vessel and the flow cell. The remaining neck on the reaction vessel was left open for the addition of aqueous manganese(II) catalyst; this also ensured that there was no build up of pressure inside the container. In a typical experiment, potassium dihydrogen phosphate (1.00 M, adjusted to pH 8.0 using NaOH(aq.), 5.00 mL), aqueous Calmagite (1.00 mM, 5.00 mL) and hydrogen peroxide (0.500 M, 5.00 mL) were added to deionised water (∼30 mL). The pH of the solution was adjusted to 8.0 ± 0.1 using freshly prepared (CO₂-free) semi-conductor grade NaOH(aq.) and the solution made up to exactly 50.0 mL. The now dark purple reaction solution was transferred to the reaction vessel and the stirrer and pump turned on. MnCl₂·4H₂O (5.00 mM, 0.500 mL) was added and readings were commenced. The changes in absorbance of the solution at 540 nm (ε = 10, 500 L mol⁻¹ cm⁻¹) were recorded electronically at 10 second intervals. Dilute aqueous HCl was used to adjust the pH of carbonate buffer solutions to pH 9.0. EPR studies were carried out using a Bruker EMX spectrometer at X-band (∼9.5 GHz), equipped with a Bruker ER 4116DM dual mode resonator (9.41 GHz parallel mode, 9.65 GHz perpendicular mode) and with an Oxford Instruments ESR900 cryostat for measurements at cryogenic temperatures 4–10 K. The EPR measurements in Fig. 5 were performed using an X/Q-band Bruker Elexsys E580 spectrometer (Bruker BioSpin GmbH, Germany) equipped with a closed-cycle cryostat (Cryogenic Ltd, UK) and an X-band split-ring resonator module with 2 mm sample access (ER 4118X-MS2). Baseline spectra samples containing only buffer were used for baseline correction; all the spectra presented have been baseline-subtracted. EPR samples, 100 or 10 μL, contained 100 mM aqueous MnCl₂·4H₂O mixed with aqueous buffer (either phosphate or carbonate), H₂O₂ and dye substrate and were flash-frozen and stored in liquid nitrogen (77 K) prior to EPR investigation. The lag time between mixing and freezing was 20 seconds. Experimental conditions are reported in the figures captions.

Note added after first publication

This article replaces the version published on October 31, 2016, which contained errors in the Acknowledgements section.

Acknowledgements

The authors would like to thank the EPSRC for the use of their National EPR Facility - University of Manchester and Dr Alistair Fielding for coordinating the use of the facility and Ms Rownok Jahan for running UV/vis spectra as part of a 3rd year BSc project.

Notes and references

1. G. Strukul, Catalytic Oxidations With Hydrogen Peroxide As Oxidant, Kluwer Academic Publishers, 1993; R. Hage, J. E. Iburg, J. Kerschner, J. H. Koek, E. L. M. Lempers, R. J. Martens, U. S. Racherla, S. W. Russell, T. Swarthoff, M. R. P. Vanvliet, J. B. Warnaar, L. Vanderwolf and B. Krijnen, Nature, 1994, 369, 637; R. Noyori, M. Aoki and K. Sato, Chem. Commun., 2003, 1977; D. Chandra and A. Bhaumik, Ind. Eng. Chem. Res., 2006, 45, 4879.
2. J. M. Campos-Martin, G. Blanco-Brieva and J. L. Fierro, Angew. Chem., Int. Ed., 2006, 45, 6962.
3. Pollution, Prevention and Abatement Handbook, World Bank Group, Washington D.C., 1998, ISBN 0-8213-3638-X.
4. H. Zollinger, Color Chemistry: Synthesis, Properties and Applications of Organic Dyes and Pigments, Wiley, Weinheim, 3rd edn, 2003.
5. E. S. Beach, R. T. Malecky, R. R. Gil, C. P. Horwitz and T. J. Collins, Catal. Sci. Technol., 2011, 1, 437.
6. G. L. Baughman and E. J. Weber, Environ. Sci. Technol., 1994, 28, 267.
7. E. Ember, S. Rothbart, R. Puchta and R. van Eldik, New J. Chem., 2009, 33, 34.
8. S. Rothbart, E. Ember and R. van Eldik, Dalton Trans., 2010, 39, 3264.
9. S. Rothbart, E. Ember and R. van Eldik, New J. Chem., 2012, 36, 732.
10. D. Swern, Organic Peroxides, Wiley, New York, 1970, p. 313.
11. V. Nadtochenko and J. Kiwi, J. Chem. Soc., Faraday Trans., 1997, 93, 2373.
12. J. Griffiths, J. Soc. Dyers Colour., 1971, 801(1972), 106.
13. T. S. Sheriff, S. Cope and M. Ekwegh, Dalton Trans., 2007, 5119; T. S. Sheriff, S. Cope and D. S. Varsani, Dalton Trans., 2013, 42, 5673.
14. M. G. Evan and N. Uri, Trans. Faraday Soc., 1949, 45, 224.
15. T. S. Sheriff, J. Chem. Soc., Dalton Trans., 1992, 1051.
16. K. A. Campbell, M. R. Lashley, J. K. Wyatt, M. H. Nantz and R. D. Britt, J. Am. Chem. Soc., 2001, 123, 5710.
17. J. Oakes, P. Gratton and I. Weil, J. Chem. Soc., Dalton Trans., 1997, 3805.
18. C. A. Wegermann, P. Strapasson, S. M. M. Romanowski, A. Bortoluzzi, R. R. Ribeiro, F. S. Nunes and S. M. Drechsel, Appl. Catal., A, 2013, 454, 11.
19. O. Abollino, C. Sarzanini and E. Mentasti, Talanta, 1994, 41, 1107.
20. T. S. Sheriff, S. Miah and K. L. Kuok, RSC Adv., 2014, 4, 35116.
21. G. E. Delroy and E. J. King, Biochemistry, 1945, 39, 245.

---

Footnote

† Electronic supplementary information (ESI) available: ESI1 log–log plot of initial rate of CAL bleaching vs. [Mn^II]; ESI2 log–log plot of initial rate of CAL bleaching vs. [H₂O₂]; ESI3 log–log plot of initial rate of CAL bleaching vs. [CAL]; ESI4 X-band EPR spectra of MnCl₂·4H₂O (100 μM) at pH 8.0 or pH 9.0 and after the addition of H₂O₂; ESI5 log–log plot of initial rate of O II and O G bleaching vs. [Mn^II]; ESI6 log–log plot of initial rate of O II and O G bleaching vs. [H₂O₂]; ESI7 log–log plot of initial rate of O II and O G bleaching vs. [HCO₃⁻]; ESI8 X-band EPR spectra of MnCl₂·4H₂O (100 μM), H₂O₂ (50.0 mM) and O II (0.100 mM) at pH 9.0 (50 mM, carbonate buffer). ESI9 X-band EPR spectra of MnCl₂·4H₂O (100 μM) and with H₂O₂ (50.0 mM) or O II (0.100 mM) or both H₂O₂ (50.0 mM) and O II (0.100 mM) at pH 9.0 (50 mM, carbonate buffer). ESI10 change in absorbance maxima of O II and O G on addition of MnCl₂·4H₂O; ESI11 the change in absorbance of CAL at pH 8.0 (phosphate buffer) and O II at pH 9.0 (carbonate buffer) with time after successive additions of dye with added MnCl₂·4H₂O. See DOI: 10.1039/c6ra23067a

---