Introduction

The purification of water is of great current concern, particularly the removal of toxic wastes, as reflected by the increasing number of studies in this area (Legrin et al. 1993; Mills et al. 1993).

Chlorinated phenols are among the most widespread pollutants of wastewater and natural waters. They are widely used as fungicides and herbicides, in addition to their application in wood protection, their production during the bleaching of paper pulp and their formation as by-products of water chlorination (Juris et al. 1988; Burrows et al. 1998; Silva et al. 1996 and references therein). Apart from their biological activity, these pollutants can be highly toxic (Halmann 1996), and, once in an aquatic environment, are quite persistent and resistant to chemical, photochemical and biological degradation (Pandiyan et al. 2002). Further, they may be transformed into more toxic compounds under the action of natural factors. These problems have led to the demand for limiting their use and for developing methods of treating contaminated waters in which they are present. To tackle this, techniques such as incineration and adsorption have commonly been adopted. However, these methods often transfer the pollutant from one phase to another, such that permanent disposal is still a problem. Light (photon) energy can be used as an alternative to break down such pollutants to harmless products, and both catalysed and uncatalysed photochemical processes have been used for their decomposition. The uncatalysed processes involve direct photolysis or UV photooxidation in the presence of O3 or/and H2O2 as oxidants (Benitz et al. 2001). Photocatalytic oxidation processes, such as those involving polyoxymetalates (POM) (Mylonas et al. 1996) and titanium dioxide (D’Oliveria et al. 1993), have gained wide use in the treatment of hazardous organic wastes. Although efficient dechlorination of chlorinated systems has been demonstrated using aqueous dispersions of platinized TiO2 (Horikoshi et al. 2003), this photocatalytic process suffers from the economic disadvantage that TiO2 does not absorb in the visible region and so UV lamps are required for irradiation. Thus, alternative routes are desirable.

Because of its attractive spectral and redox properties, Ru(bpy) 2+3 is among the most widely studied systems in transition metal photochemistry (Juris et al. 1988). In a previous report, we have shown that this complex provides a viable alternative method for destroying the toxic organic pollutant 4-chlorophenol (Silva et al. 1996), as it does not require UV light and does not result in a transfer of the pollution problem from one place to another. However, since the photodegradation pathways of different chlorophenols may not be identical, we have extended the study to the photolysis of various mono and polychlorophenols in deaerated aqueous solutions in the presence of Ru(bpy) 2+3 /S2O 2−8 . The photodegradation occurs by electron transfer initiated by the quenching of the ruthenium complex excited state either by S2O 2−8 (forming the corresponding strongly oxidising intermediates Ru(bpy) 3+3 and SO −.4 ) or chlorophenol (with the formation of Ru(bpy) +3 and chlorophenol+/.).

Upon prolonged irradiation, it is suggested that photomineralization of the chlorophenols occurs. As Ru(bpy) 2+3 is regenerated in the reaction, the scheme corresponds to an overall catalytic process. The photodechlorination kinetics associated with these processes are also reported and compared at acidic and alkaline pH. Pseudo-first order kinetic behaviour describes the rapid photodechlorination behavior quite well under our experimental conditions. The kinetics were also studied as functions of Ru(bpy) 2+3 and S2O 2− 8 concentrations for several irradiation times.

For the case of the Ru(bpy) 2+3 /S2O 2−8 sensitized photodegradation of pentachlorophenol, we also report a detailed characterization of the reaction products.

Experimental

Reagents and instruments

2, 3 and 4-chlorophenol (2-CP, 3-CP and 4-CP, respectively), 2,4-dichlorophenol (DCP), 2,4,6-trichlorophenol (TCP) and pentachlorophenol (PCP), methanol (for HPLC) and sodium sulphate were p.a. reagents from Riedel-de-Häen. Tris(2,2′-bipyridyl)ruthenium II dichloride hexahydrate was p.a. from Aldrich. The other compounds were p.a. reagents from Merck. All reagents were used as received. All the experiments were carried out in oxygen-free aqueous solutions at room temperature (±20°C). For high performance liquid chromatography (HPLC) studies, experiments were carried out using MilliQ system millipore deionized water.

Typical procedure

Freshly prepared aqueous solutions of chlorophenols (1.1 × 10−4 to 5.0 × 10−4 M) in the presence of Ru(bpy) 2+3 /S2O 2−8 (concentrations ranging from 5.0 × 10−7 to 6.7 × 10−6 M for the Ru(bpy) 2+3 complex and 6.7 × 10−4 to 8.0 × 10−3 M for S2O 2−8 ) were purged with argon for 15 min and irradiated in 1 cm quartz cells. These used an Applied Photophysics semi micro reactor equipped with a cut-off filter (λ > 350 nm).

The time evolution of the disappearance of chlorophenols upon irradiation in the presence of Ru(bpy) 2+3 /S2O 2−8 , intermediates and photoproducts formation was followed by a Gilson HPLC apparatus using an analytical cartridge [C18 water column (4.6 × 150 mm) with a Gilson 151 UV/vis detector. The retention times were compared with those of standards. The mobile phase consisted of a methanol/water mixture (40/60;v/v). The UV/vis absorption spectra of the irradiated and non-irradiated Ar purged solutions were measured in 1 cm quartz cells using a Shimadzu 2100 spectrophotometer. The luminescence spectra of all Ar purged solutions were monitored at the ruthenium complex maximum absorbance (around 455 nm) and the emission intensities were monitored at the wavelength of maximum emission (around 620 nm) in 1 cm quartz cells, on a Fluorlog 3.22 Isa Horiba spectrofluorimeter. The pH was measured potentiometrically using a micro-Crison pH 2000. Cl− concentrations were also determined potentiometrically with a micro-Crison 2002 (mV), using a Ag/AgCl Methrom reference electrode and an Orion 94–117B chloride selective electrode. The 13C NMR spectra of unphotolysed and photalysed solutions were obtained at 25 ± 0.5°C, on a Varian unity 500 spectrometer operating at 126.68 MHz with off resonance decoupled. The chemical shifts were measured using TMS as the internal reference. The presence of SO 2−4 in the photoproducts was confirmed by qualitative analysis.

Results and discussion

Photodegradation studies

At room temperature, the UV absorption spectra corresponding to transitions of the chlorine and hydroxyl group substituted benzene rings of aqueous solutions of monochlorophenols were observed at 275 and 225 nm, while those for the polychlorophenols occurred around 290 and 235 nm. Aqueous Ru(bpy) 2+3 solutions showed visible absorption and emission spectra with maxima at 455 and 620 nm respectively, in agreement with those of the literature (Juris et al. 1988).

The absorption spectra of mixtures of Ru(bpy) 2+3 and S2O 2−8 and/or chlorophenol were seen to be equal to the sum of the component spectra, showing that, under our experimental conditions, complexation between the sensitizer and the quenchers was negligible in the ground state. Oxygen-free Ru(bpy) 2+3 solutions showed almost no decrease in the steady-state luminescence intensity of the metal complex in the presence of chlorophenol in pH < chlorophenol pKa, in agreement with the observation of Rajagopal et al. (1992) on the lack of quenching by other phenol derivatives in neutral solutions. However, a decrease in the emission intensity was observed in alkaline solutions (NaOH ca 10−2 M).

The quenching of 3Ru(bpy) 2+3 luminescence by several chlorophenols was studied at pH = 12, with at least five different quencher concentrations. The Stern-Volmer plots of the luminescence intensity measurements versus [CP] were linear with correlation coefficients greater than 0.997, indicating the absence of static quenching. Since the quenching of 3Ru(bpy) 2+ 3 by the chlorophenolate ions through electronic energy transfer should be an energetically unfavourable process, and as the chlorophenol ions are electron donors, a reductive quenching mechanism is probable. The corresponding bimolecular quenching rate constants were obtained from the slope of the Stern-Volmer plots, K SV, and τ0, the unquenched lifetime of 3Ru(bpy) 2+3 (Juris et al. 1988). The experimental values, kq = 1.0–2.0 × 109 M−1s−1, which depend on the chlorophenols, are slightly higher than the value reported for phenol under similar conditions (Rajagopal et al. 1992), in agreement with a dominant reductive quenching mechanism.

A similar study was performed for the quenching of 3Ru(bpy) 2+3 by S2O 2−8 and the bimolecular quenching rate constant values, corresponding to oxidative quenching, were of the order 109 M−1 s−1, close to previously reported values (Juris et al. 1988 and references therein).

Upon photolysis of oxygen-free aqueous solutions of mono and polychlorophenols in the presence of Ru(bpy) 2+3 /S2O 2−8 with near visible light, changes were seen in the UV/visible spectra (Fig. 1a, b for acidic pH and Fig. 2a, b for alkaline pH).

Fig. 1
figure 1

a UV/visible absorption spectra of pentachlorophenol. [PCP] = 1.1 × 10−4 M, [Ru(bpy) 2+3 ] = 2.7 × 10−4 M, [S2O 2−8 ] = 6.7 × 10−3 M, before (1) and after irradiation times: 2–5, 3–10, 4–20, 5–30, 6–60, 7–120, 8–240 and 9–360 min. λirr=350 nm. Acidic pH. b Difference in absorption spectra from a between systems after and before irradiation

Full size image
Fig. 2
figure 2

a UV/visible absorption spectra of pentachlorophenol. [PCP]=1.1 × 10−4 M, [Ru(bpy) 2+3 ] = 2.7 × 10−4 M, [S2O 2−8 ] = 6.7 × 10−3 M, before (1) and after irradiation times: 2–5, 3–10, 4–20, 5–30, 6–60, 7–120, 8–240 and 9–360 min. λirr = 350 nm. Alkaline pH. b Difference absorption spectra from a between systems after and before irradiation

Full size image

From Figs. 1 and 2, it is clear that PCP (ε250nm=1,950 M−1cm−1, ε321nm=935 M−1cm−1) is transformed into various photoproducts. Among the possible photoproducts, the major components detected by HPLC are BQ (R t=3.2); 1,2-HQ (R t=2.1); and 1,4-HQ (R t=1.8), in addition to the observation of the peak of PCP (R t=5.2). The 13C NMR spectra (peaks at 189–198 ppm and 172–175 ppm) are also consistent with the formation of benzoquinone and dihydroxybenzenes, while the UV/visible spectral deconvolution shows that benzoquinone and dihydroxybenzenes absorb in a similar region to mono and polychlorophenols themselves. Although this latter observation makes kinetic studies based on UV/visible absorption changes difficult, it is possible to obtain valid data by assuming that bands near 246, 275 and 289 nm correspond to BQ (ε246nm=7,715 M−1cm−1), 1,2-HQ (ε275nm=2,160 M−1cm−1) and 1,4-HQ (ε289nm=2,515 M−1cm−1) maximum, respectively. As can be seen in Table 1, photodegradation of PCP depends on the initial Ru(bpy) 2+3 and S2O 2−8 concentrations and on the irradiation time.

Table 1 UV/visible absorbance changes occurring on the photodegradation of pentachlorophenol. λirr=350 nm [PCP]=1.1 × 10– 4 M
Full size table

From this table and Figs. 1a, b and 2a, b it can also be seen that at longer irradiation times not only is the formation of intermediates enhanced, but also secondary reactions resulting from the degradation of these intermediates become important, as seen by absorptions at 310, 270 and 246 nm. These may result in the photomineralization of PCP at longer irradiation times.

This is confirmed by HPLC, although some traces of unknown photoproducts are still present. The UV/visible absorbance near 450 nm, corresponding to the Ru(bpy) 2+3 MLCT band is almost identical after photolysis to that of non-irradiated solutions, showing that there is no overall complex degradation and that this is a catalytic process. The weak absorbance near 450 nm upon irradiation times between 20–120 min may result from the absorbance of a chlorophenol intermediate formed via ring contraction (Bonnichon et al. 2001), while the appearance of the new band at 520 nm can be attributed to an intermediate ruthenium species, Ru(bpy) +3 (Juris et al. 1988). However, at longer irradiation times, a decrease in this absorbance at 520 nm and in the intermediate band at 450 nm is observed, supporting the catalytic process and the PCP photomineralization.

The quantification of Cl− in the photoproducts was obtained by potentiometric analysis as a function of the initial Ru(bpy) 2+3 and S2O 2− 8 concentrations at various irradiation times (Fig. 3a, b). In acidic media, the photodegradation of PCP is also accompanied by a pH decrease of the irradiated solutions (Table 1).

Fig. 3
figure 3

a Pentachlorophenol dechlorination as a function of the irradiation time in acidic pH. λirr = 350 nm. [PCP] = 1.1 × 10−4 M. Filled diamond [Ru(bpy) 2+3 ] = 1.0 × 10−6 M, [S2O 2−8 ] = 3.3 × 10−4 M; filled triangle [Ru(bpy) 2+3 ] = 2.7 × 10−6 M, [S2O 2−8 ] = 3.3 × 10−4 M; filled square [Ru(bpy) 2+3 ] = 1.0 × 10−6 M, [S2O 2−8 ] = 3.3 × 10−3 M. b Pentachlorophenol dechlorination as a function of the irradiation time at alkaline pH. λirr=350 nm [PCP]=1.1 × 10−4 M. Filled diamond [Ru(bpy) 2+3 ] = 1.0 × 10−6 M, [S2O 2−8 ] = 3.3 × 10−4 M; filled triangle [Ru(bpy) 2+3 ] = 2.7 × 10−6 M; [S2O 2−8 ] = 3.3 × 10−4 M; filled square [Ru(bpy) 2+3 ] = 1.0 × 10−6 M, [S2O 2−8 ] = 6.7 × 10−4 M

Full size image

The results of continuous photolysis can be described according to the following reactions (1–7) initiated by the formation of the 3MLCT excited state of Ru(bpy) 2+3 (1)

$${\text{Ru}}({\text{bpy}})_{3}^{{2 + }} \mathop{\rightarrow}\limits^{\rm hv} {}^{\ast}{\text{Ru}}({\text{bpy}})_{3}^{{2 + }}. $$
(1)

Then S2O 2−8 quenches the 3MLCT excited state of Ru(bpy) 2+3 , (2)

$$ {}^{\ast}{\text{Ru}}({\text{bpy}})_{3}^{{2 + }} + {\text{ S}}_{2} {\text{O}}_{8}^{{2 - }} \to {\text{Ru}}({\text{bpy}})_{3}^{{3 + }} + {\text{ SO}}_{4}^{{. - }} + {\text{ SO}}_{4}^{{2 - }}. $$
(2)

The anion radical SO .-4 can either oxidize a second Ru(bpy) 2+3 species (3) by thermal reaction or react with PCP by electron transfer (4), addition (5) and/or hydrogen abstraction (6)

$$ {\text{Ru}}({\text{bpy}})_{3}^{{2 + }} + {\text{ SO}}_{4}^{{. - }} \to {\text{Ru}}({\text{bpy}})_{3}^{{3+}} + {\text{ SO}}_{4}^{{2-}}, $$
(3)
$$ {\text{PCP}} + {\text{SO}}_{4}^{{.-}} \to {\text{PCP}}^{{+.}} + {\text{SO}}_{4}^{{2-}}, $$
(4)
$$ {\text{PCP}} + {\text{ SO}}_{4}^{{. - }} \to {\text{PCP}} - {\text{SO}}_{4}^{{. - }}, $$
(5)
$$ {\text{PCP}} + {\text{SO}}_{4}^{{. - }} \to {\text{PCP}}^{.} + {\text{ H}}^{ + } + {\text{ SO}}_{4}^{{2-}}. $$
(6)

The ruthenium(III) complex is also a strong oxidant (Juris et al. 1988) and can oxidize PCP according to reaction (7),

$$ {\text{PCP }} + {\text{ Ru}}({\text{bpy}})_{3}^{{3+}}\to {\text{PCP}}^{{{\text{ + }}{\text{.}}}} + {\text{Ru}}({\text{bpy}})_{3}^{{2+}}. $$
(7)

PCP quenches the 3MLCT of Ru(bpy) 2+3 by a reductive quenching process, reaction (8), as is observed, at pH 12

$$ {\text{PCP}}^{ - } + {}^{\ast}{\text{Ru}}({\text{bpy}})_{3}^{{2+}}\to {\text{PCP}}^{.} + {\text{Ru}}({\text{bpy}})_{3}^{ + }. $$
(8)

The transient species formed during PCP degradation by reaction with SO ·−4 and/or the ruthenium complexes lead to the formation of the cation radical PCP+·, and/or its conjugated base, the phenoxyl radical PCP·, is also formed in the reaction (9)

$$ {\text{PCP}}^{{ + .}} \to {\text{PCP}}^{.} + {\text{H}}^{ + }. $$
(9)

The occurrence of reactions 6 and 9 explains the observed pH decrease (Table 1). The intermediates can also eliminate HCl and yield the detected photoproducts, benzoquinone and the dihydroxybenzenes.

There are also indications that ketene (resulting from ring contraction by Wolff rearrangement, Decand et al. 1994), biphenyls (Guyon et al. 1982) and some halogenated compounds (Burrows et al. 1998) could also be formed. As the solutions are continuously irradiated, the ruthenium(II) complex can be re-excited, oxidized by further peroxydisulphate and/or reduced by PCP to give a catalytic cycle (Fig. 4), until S2O 2−8 completely disappears to form sulphate ion and/or PCP is completely mineralized. The presence of SO 2−4 in the photoproducts was qualitatively confirmed by precipitation with Ba2+.

Fig. 4
figure 4

Scheme of the photocatalytic degradation of pentachlorophenol in the presence of the pair Ru(bpy) 2+3 /S2O 2−8

Full size image

Kinetic studies

Kinetic studies of PCP degradation upon irradiation were made by studying Cl− release (Fig. 3). The Cl− concentration expected for complete PCP degradation was greater than what was observed experimentally. This suggests that some chlorinated intermediates were formed and were not completely degraded with an irradiation time of 3 h. However, in spite of this, it was still possible to obtain good pseudo-first order kinetic conditions. The experimental pseudo-first order rate constants obtained were k 1 = 2.7 × 10−6 min−1 (χ = 0.993) and k 2 = 1.9 × 10−4 min−1 (χ = 0.995) when the S2O 2−8 and Ru(bpy) 2+3 initial concentrations change, at acidic and alkaline pH, respectively. The sensitized PCP photodechlorination rate law was found to be proportional to the Ru(bpy) 2+3 concentration [or that of *Ru(bpy) 2+3 ] and S2O 2−8 initial concentrations,

$$ {\text{V}} = {\text{d}}[{\text{Cl}}^{ - } ]/{\text{d}}t = {\text{k}}{\left[ {{\text{Ru}}({\text{bpy}})_{3}^{{2 + }} } \right]}{\left[ {{\text{S}}_{2} {\text{O}}_{8}^{{2 - }} } \right]}. $$
(10)

Assuming that PCP photodechlorination is also first order in the PCP (1.1 × 10−4 M) concentration under our experimental conditions, the PCP photodechlorination rate constant was k = 5.3 × 103 M−2 min−1 (acidic pH) and k = 3.8 × 104 M−2min−1 (alkaline pH) and the photodechlorination yields obtained were 21 and 36% at acidic and alkaline pH, respectively.

We have previously shown that the system Ru(bpy) 2+3 /S2O 2−8 acts as an excellent catalyst for the photodegradation/photodechlorination of the monochlophenols 2, 3, and 4, photodechlorination yield ≅80%, k ≅ 102 M−2  min−1, also of the polychlorophenols 2,4-dichlorophenol, photodechlorination yield ≅ 60%, k ≅ 103 M−2 min−1 and 2,4,6-trichlorophenol, photodechlorination yield ≅56%, k ≅ 103 M−2 min−1, at acidic pH and slightly higher at alkaline pH, assuming that all chlorine atoms were released during photolysis. The observed increase in the photodechlorination rate constants when the number of chlorine atoms and the pH increase can be explained by the increasing acidic character of the chlorophenols (pKa = 9.43–6.23), in general, and pentachlorophenol, in particular (pKa = 4.7), with the dissociated species being more reactive towards oxidants such as the ruthenium complexes and the anion radical SO −4 . In contrast, the photodechlorination yield decreases with increasing benzene substitution or decreasing pH.

Conclusions

The system Ru(bpy) 2+3 /S2O 2−8 acts as an excellent catalyst for the photodegradation of mono and polychlorophenols using near visible light. In the present study, we have shown that this system is very efficient in the photodegradation/photodechlorination of pentachlorophenol. The main organic products among the intermediates formed during photolysis appear to be benzoquinone (or its chloro derivatives) and dihydroxy benzenes (and its respective chloro derivatives), while the inorganic ions Cl−, SO 2−4 and H+ are also formed. The results suggest that the photodegradation of PCP sensitized by Ru(bpy) 2+3 /S2O 2−8 is an oxidative process, initiated by SO .−4 and/or Ru(III,II) formed in the quenching of the excited state of the metallic complex by S2O 2−8 or PCP-. The reaction mechanism probably involves both direct electron transfer and addition–elimination processes. The intermediate species may react with each other or another PCP molecule, leading to dechlorination and the observed pH changes in acidic solutions. As the Ru(bpy) 2+3 complex is regenerated in the overall reaction of PCP, photodechlorination in the presence of the pair Ru(bpy) 2+3 /S2O 2−8 makes the scheme a catalytic process. The kinetic studies carried out under pseudo-first order conditions, show that PCP photodechlorination is a fast reaction and is first order in both Ru(bpy) 2+3 and S2O 2−8 initial concentrations. If the ruthenium complex can be immobilized on an appropriate solid support, the system has considerable potential for general applications for decomposition of chloroaromatic pollutants using sunlight.