2.1 Reagents and chemicals

Pentachlorophenol 86% p.a. (Aldrich) was used without further purification, hydrogen peroxide reagent grade 30% (Merck), analytical grade reagents or calibration standard supplied by Merck or Mallinckrodt were used for HPLC measurements. All solutions used in the degradation studies were prepared using double distilled water. The catalyst was prepared using a commercial TiO₂ Degussa P25 and Mn(NO₃)₂.4H₂O (Merck).

2.2 Catalyst

Catalyst MnO₂/TiO₂ was prepared using TiO₂ Degussa P25 as support with 1% w/w MnO₂, by impregnation at 35°C with an aqueous Mn(NO₃)₂ solution using the incipient wetness method. The solid was dried at 120°C for 12 h and calcined in air at 500°C for 4 h.²⁵(Villaseñor, 2002). Its characterization was by evaluated specific surface area and porosity in an automatic Gemini 2370 Micromeritics system, from the N₂ adsorption isotherm at −196°C in the relative pressure range of 0.05-0.995. X-ray analysis for the catalyst was made on a Bruker D8 Advance instrument with Linear LynxEye detector, Bragg-Brentane geometry, Cu λ = 1.5406 A, 40 KV – 30 mA power, and 0.1 mm fixed optics. Temperature Program Reduction (TPR) experiments were carried out in a TPR/TPD (Temperature Program Desorption) 2900 Micromeritics system provided with a thermal conductivity detector. The reducing gas was a mixture of 5% H₂/Ar (40 cm³/min) and a heating rate of 10 K/min was used. The TPD of ammonia was carried out using an Ar flow of 50 cm³/min as carrier gas. Ammonia pulses were dosed in order to saturate the catalyst surface at 100°C; the sample was cooled to room temperature, and once the base line was restored the temperature was increased linearly (10 K/min) up to 500°C.²⁴ The zero point charge (zpc) was obtained by measuring the zeta-potentials as a function of pH suspensions.²⁶ The measurements were carried out in a Zeta-Meter Inc. (Model ZM-77) using 20 mg of 2 μm catalyst particles ultrasonically suspended in 200 mL of 1×10⁻³ mol/L KCl solution.

2.3 Degradation experiments

The following processes were studied: UV processes (UV and UV/H₂O₂), ozonation processes (O₃, O₃/H₂O₂, O₃/UV), and catalytic ozonation. The degradation process was performed in a Pyrex glass reactor.²⁷ It was charged with 90 mL of aqueous solution of PCP with an initial concentration of 30 mg/L. The MnO₂/TiO₂ catalyst (2.2 g/L) was kept in suspension with magnetic stirring. The external chamber of the reactor was kept at 20°C or 40°C by recirculation of water. For the ozonation processes, the ozone was generated by an OZOCAV ozonizer²⁵ and fed at an oxygen flow of 50 mL/min, reaching an ozone concentration of 22.11 mg/L. The H₂O₂ was used at 1.5×10⁻⁴ mol/L for each reaction.

For the experiments performed under UV light, the solution were irradiated with a General Electric UV lamp (HR 250 DX 37/40, λ ≤ 254 nm) across the quartz reactor windows. The lamp was placed at the top of the window 5 cm over the sample.²⁷ The reactions were ran at pH 5, 7 and 9.

2.4 Analytical methods

Prior to the analysis, the samples were extracted from the reactor at several reaction times (0-60 min) and filtered through 0.20 μm membrane (Millipore). The concentrations of PCP and its degradation intermediates were measured with an HPLC system coupled to a Perkin Elmer Series 200 chromatograph with a UV-VIS detector. All the intermediates were identified by HPLC by comparing the retention time of the standard solution and comparing the GC-MS with the mass spectra of the intermediate compound in the database of NIST (National Institute of Standards and Technology). The following conditions were used for aromatic compounds; a Merck-Chromolith Performance RP-18 column (4.6 mm×100 mm). The mobile phase was an acetonitrile:H₃PO₄ (7×10⁻³ mol/L) mixture (40:60), with a flow rate of 1.5 mL/min, and a wavelength detection at 215 nm. The analyses of the acid compounds were made with a Transgenomic ORH-801 column (6.5 mm × 300 mm) with 5×10⁻³ mol/L sulfuric acid as eluent (0.8 mL/min) at 200 nm.

Evolution of CO₂ was monitored according to previously described method.²⁷ Chloride samples were quantified by a potentiometric method²⁹ using an ORION specific chloride electrode.

2.5 Ecotoxicity assay

The ecotoxicity assays were performed using Daphnia magna (microcrustacean) and two plant species, Lactuca sativa (lettuce) and Panicum millaceum (millet), as dicotyledon and monocotyledon representatives, respectively.^28,29,30

Several colonies of D. magna were maintained at 20°C. All the tests were performed in the dark at constant temperature of 20 ± 2°C. Acute toxicity was assessed by recording the effects of the compounds tested on the motility of D. magna. Daphnids less than 24 h old were used for the assays. The effective media concentration (EC₅₀) was determined as the concentration estimated to immobilize 50% of the daphnids after 24 h and 48 h of exposure with the 30 mg/L PCP solutions; the data was analyzed with Probit model statistical method.^31,32 The toxicity of the samples was reported as percent of immobilized daphnides with respect to a control (water). To compare the effects with the initial PCP concentration, the tests were performed using a same sample volume, higher than EC₅₀ of PCP (0.75 mg/L at 48 hours). In all the experiments, a 300 pL of degradation process samples and PCP solution (30 mg/L) was completed to 10 mL with water.

Lettuce (Lactuca sativa) and millet (Panicum millaceum) seeds were selected as phytotoxicity bioindicators because they germinate very quickly and sprouts show rapid growth. Ten seed per plants were placed on a filter paper inside a vial and 200 pL of undiluted degradation samples were added. After a period, growth 3 days of light and 3 days of darkness at 25°C, stems and roots length were recorded and expressed as the percentage length with respect of control plants treated with water. The statistical significance of differences between groups was determined by one-way ANOVA (p-value ≤ 0.05) and compared with the control group by the Dunnett-test.^{28,29 30}

3.1 Uncatalyzed Reactions

PCP degradation at a concentration of 30 mg/L was studied across the time in the absence of the catalyst. The pH of the solution was adjusted to pH 5, pH 7 and pH 9 with a phosphate buffer, and the process was monitored at 20°C and 40°C. Assuming pseudo-first-order kinetics, the rate constant was between 0.4-0.5 ± 0.1 min⁻¹ in the presence only of ozone at 20°C in all pH tested; at 40°C, a slight increase in the rate constant of 0.7 ± 0.1 min⁻¹ was achieved. At 10 minutes of uncatalyzed ozonation reaction, 96 ± 2% degradation of PCP at both temperatures was reached.

The Fig. 1 shows the degradation of PCP to pH7 at 20°C in the presence of O₃, O₃/UV/H₂O₂ and UV/H₂O₂. In order to optimize the reaction system, UV, H₂O₂ and combinations of both were included; remarkably, in the absence of ozone a slow degradation of PCP was observed, with rate constants between 2×10⁻³ – 6×10⁻³ min⁻¹. On the other hand, no significant impact on PCP decomposition at both temperatures was observed under these conditions. The purpose of introducing UV radiation in the ozonation processes was to yield more free radicals for the higher ozonation rate, as was found in the decomposition of polyethylene glycol by O₃ with UV radiation.³³ These small differences found in the rate reaction for PCP degradation with the addition of H₂O₂ or UV irradiation, indicate that O₃ is the main agent involved in the process.

Figure 1 Degradation of PCP at pH 7 during 60 minutes in the presence of O₃ (Δ), O₃/UV/H₂O₂ (○) and UV/H₂O₂ (□). 

In this study, chloride and CO₂ were found as mineralization products. The release of Cl after 60 min of ozonation was 4.2×10⁻⁴ mol/L at pH 9 and 2.2×10^{− 4} mol/L at pH5. However, without O₃, the chloride production was lower. The formation of CO₂ was ~ 1.1×10⁻⁵ mol/L with O₃ and ~ 4×10⁻⁶ mol/L in the absence of ozone. The ozonation process led to an increased mineralization effect compared to the photochemical processes and oxidation by peroxide. The importance of O₃ in the PCP degradation is evident according to the reactions in the experiments, Fig. 1.

In water, ozone can oxidize contaminants by direct selective reactions, resulting in the addition to C=C bonds to or through a chain reaction mechanism that produces free hydroxyl radicals, which are stronger oxidants than molecular ozone.³⁴ The production of hydroxyl radicals by ozone in aqueous solution may be taken placed through chain reactions as shown in equations 1-4,¹⁹ or by traces of other substances, such as transition metal cations, that leads to an indirect attack on organic compounds, which is faster than a direct attack by molecular ozone. ²⁴

During the uncatalyzed ozonation process of PCP, intermediates formation, namely tetrachloro-1,2-hydroquinone (TCl-1,2-HQ) and tetrachloro-1,2-benzoquinone (TCl-1,2-BQ), at pH 5 and tetrachloro-1,4- hydroquinone (TCl-1,4-HQ) and tetrachloro-1,2-hydroquinone (TCl-1,4-BQ) at pH 9, were identified (Fig. 2). The major intermediate identified in the first minutes of uncatalyzed ozonation was TCl-1,2-HQ at pH 5, which may be produced as a result of the attack of either ozone or a hydroxyl radical on the carbon in ortho position of the PCP molecule. TCl-1,2-BQ is the oxidation product of TCl-1,2-HQ, which produces a reddish solution. At the same time, dehalogenation products such as phenols and mono-, di-and/or trichlorophenols were formed. It is likely that these later compounds are products of electron transfer reactions between the hydroxyl radical and PCP, which generate in the first instance cation radicals such as those shown in Schemes 1 and 2, as proposed previously by Legrini and co-workers.³⁵ These radicals may then react to produce the corresponding dehalogenation products. Moreover, the insertion of a hydroxyl group generates dehalogenation products, resorcinol and hydroquinone, as shown in the scheme 2. First, the dehalogenation of PCP produces tetrachlorocatechol, which also undergoes dehalogenation and addition of hydroxyl groups affording pyrogalol and catechol.

Sheme 1 

Sheme 2 

Figure 2 Degradation of PCP (—) at (a) pH 5 and (b) pH 9 with the evolution of Cl⁻ (- - - -) and the presence of the main products of degradation found under these conditions (TCl-HQ- - - -and TCl-BQ.......). 

The subsequent oxidation of these compounds causes the breaking of the benzene ring and the formation of low molecular weight carboxylic acids such as maleic, oxalic, trichloroacetic and formic acids. The formation of trichloroacetic acid can be explained by the proposal made by Anipsitakis and co-workers³⁷ from 1,1,3,3-tetrachloroacetone (scheme 3):

Sheme 3 

In addition to the formation of several quinones, the final steps in this oxidation process was the conversion of these carboxylic acids to complete mineralization products, carbon dioxide, chloride, and water.

It is noteworthy that TCl-1,2-BQ at pH 5 and 1,4-BQ at pH 9, were respectively formed as the main PCP degradation intermediates (Fig. 3 and 4). According to the pK_a of PCP, at pH 5 the phenolate ion reaches 67%, while at pH 9 it is 99.9% of the phenolate ion. The negative charge on the oxygen atom can hinder the attack of hydroxyl in an ortho position, favoring the addition in para position at pH 9. Under alkaline conditions, a greater degree of dehalogenation products like phenol, pyrogalol or resorcinol at higher concentrations than under acidic conditions was observed. The products of oxidation of these compounds at both experimental pH conditions included the formation of maleic, oxalic and formic acids, however at pH 9 there was generation of additional organic acids such as tartaric and malonic. For the mineralization products, chloride evolution was greater at basic pH than under acidic conditions. The percentage of dechlorination at 60 min of reaction reached 75% of the theorical value.

Figure 3 Degradation pathway proposal with the main intermediaries detected during the ozonation process of PCP at pH 5. 

Figure 4 Degradation pathway proposal with the main intermediaries detected during the ozonation process of PCP at pH 9. 

According to these results, we propose degradation pathways (Fig. 3 and 4) with the main intermediaries detected during the uncatalyzed ozonation process of PCP at pH 5 and pH 9.

3.2 Catalyst Characterization

The preparation of the catalyst MnO₂/TiO₂ was reported previously.²⁵ Table 1 summarizes the characterization of the catalysts studied. Some of these results have been previously reported by our group²⁴.

The textural properties of the catalyst MnO₂/TiO₂ 1% were evaluated by nitrogen sorption experiments. The specific surface area and average pore radius obtained from nitrogen adsorption were 43 m²/g and 5.0 nm, respectively. The specific surface area of the solids obtained from nitrogen adsorption isotherms at 77 K showed the expected trends. The addition of manganese oxide to the support produces a slight decrease in the surface due to surface coverage.

Surface acidity obtained from the TPD of ammonia showed a single, wide peak centered at 533 K for TiO₂. MnO₂/TiO₂ catalyst showed two peaks, the first one was broad and centred at 533 K while the second presented a higher intensity centred at 790 K. Table 1 presents surface acidity estimated from TPD data. The MnO₂/TiO₂ catalyst displays the highest value. If it is considered that the bulk MnO₂ possess a surface acidity of 0.095 meq g⁻¹ and the MnO₂ loading is only 1wt%, only a slight increase should be expected. However, the obtained surface acidity of MnO₂/TiO₂ is approximately 50% higher than the pure support, which is indicative of MnO₂'s high surface coverage.

The ZPC values are also displayed in Table 1. Taking into account that the bulk TiO₂ displays a ZPC of 6.1, a decrease in 0.5 pH units after the deposition of MnO₂ on TiO₂ is observed. Since the ZPC of bulk MnO₂ is 3.4, an estimate of surface coverage indicates that approximately 20% of the TiO₂ surface is coverage with 1wt% of MnO₂, indicating a high dispersion of the oxide phase.

The Fig. 5 shows the XRD (X-Ray Diffraction) of MnO₂/TiO₂ 1% w/w. The catalyst has anatase (PDF file 00-021-1272) and rutile (PDF file 00-0211276) phases, characteristic of titanium oxide; the manganese oxide phase was not be detected, probably due to the low amount of this oxide in the catalyst.

Figure 5 XRD of MnO₂/TiO₂ catalyst. 

3.3 Catalytic Ozonation

The catalytic ozonation reactions of PCP were studied at 20°C and at different pH. At pH 5 the rate constant was 0.7 ± 0.1 min⁻¹, whereas at pH 7 and 9 the k_obs was 0.4 ± 0.1 min⁻¹. The slightly higher value of the rate constant at pH 5 can be explained by a major adsorption of phenolate anions on the slightly positive catalyst surface (ZPC = 5.6), allowing a better contact between PCP and oxidant species on the surface of the catalyst.³⁷. Nevertheless, adsorption of PCP in the dark is rather small along all the pH values employed. The average amount of PCP adsorbed on TiO₂ was 3.7 % and 4.7 % for TiO₂/MnO₂.

Fig. 6 shows the diminution of PCP concentration over time during the first 30 minutes of reaction at pH 5. Under the experimental conditions employed in this work degradation of PCP is relatively fast, and just a slight difference can be observed.

Figure 6 Degradation of PCP at pH 5 during 30 minutes with TiO₂/MnO₂ (□) and without TiO₂/MnO₂ (Δ). 

Although the nature of the intermediates found is the same than in the case of homogeneous ozonation, the effect of the use of TiO₂/MnO₂ is expressed in the change of the proportions of these intermediates. Differences were observed in chloride and CO₂ evolution, which were higher than in a catalyzed presence, indicating a different route to mineralization. Theoretically if the PCP is completely degraded, it would generate 5.6×10⁻⁴ mol/L of chlorides, when comparing the production of chlorides at different pH, it is observed that the highest amount of chlorides formed is at pH9 producing 5.0×10^’5mol/L at the end of the reaction. The greatest degradation of PCP occurs in the first few minutes of the reaction, where the highest CO₂ formation is observed. During ozonation reactions at pH 5 and pH 9, CO₂ was generated reaching concentrations of 7×10⁻⁶ mol/L and 2.8×10⁻⁶mol/L respectively, while in the catalytic ozonation reactions, an increase in the CO₂ generation was observed at both pH values, reaching 9.5×10⁻⁶ mol/L at pH 5, and 6×10⁻⁶ mol/L at pH 9 after 40 min of degradation.

3.4 Toxicity assays

One of the objectives of this study was to verify that products generated in the process of degradation of PCP do not show more toxicity than the initial compound. Using Daphnia magna as a representative species of an aquatic organism (Ecotoxicity assay) and Lactuca sativa and Panicum millaceum as terrestrial organism (Phytoxicity assay).

According to several reports, many catalytic processes can generate products that are more toxic than the initial substances. ^38,39,40 It was expected that the lower formation of halogenated intermediates would produce less toxic solutions.⁴¹ Our results indicate that the lowest formation of organohalide compounds is performed at pH 9, therefore this system was selected for the following ecotoxicity tests.

3.5 Ecotoxicity assay

The microcrustacean Daphnia magna was used according to Chilean Standards. The first assay consisted in evaluating the initial solution of PCP (30 mg/L) at different pH values, because PCP degradation process was dependent on the pH. The results are presented in Table 3. It was determined the IC₅₀ for PCP with average values of 0.8 mg/L and 0.6 mg/L, at 24 h and 48 h, respectively. An ecotoxicity positive control, IC₅₀ for K₂Cr₂O₇ (1.2 ± 0.07 mg/L) was included in the assay. These results are reasonably consistent with that previously reported for PCP (IC_50-24h = 0.62 mg/L) using the same microcrustaceans.⁴²

To study the evolution of toxicity during the ozonation reaction, samples from uncatalyzed and catalytic ozonation were taken at different reaction times and diluted to reach approximately twice the IC_50-48h. At time zero, when the degradation process had not started, the survival percentage was zero. The percentage of survival increased rapidly when the microcrustacean were incubated with both ozonation samples obtained from 15 minutes of reaction and beyond (data not shown). The decrease in toxicity must be related to the decrease in concentration of PCP and the formation of less toxic intermediates formed during the ozonation process. No significant differences were found in the survival profile between the MnO₂ catalyzed ozonation regarding the uncatalyzed process (data not shown).

3.6 Phytotoxicity assay

Phytotoxicity tests were performed using seeds of Lactuca sativa (dicotyledonous) and Panicum millaceum (monodicotyledonous). In two tests, the growth of root and stem of both species were measured after 6 days of incubation. The seeds were treated with samples obtained during the degradation of PCP, and distilled water was used as a control. In the case of lettuce assays, it was not observed any significant differences in presence and absence of catalyst.

The Fig. 7, shows the growth percentage of root and stem of Panicum millaceum exposed to PCP solutions sampled at different time intervals during ozonation and catalytic ozonation at pH9. An initial solution of 30 mg/L PCP (time 0 min) inhibited the root growth ~65% and stem growth ~50%.

Figure 7 Effect of samples taken at different time intervals during the PCP treatment for ozonation in the presence of (MnO₂/TiO₂) to pH 9 and in the absence of catalyst, expressed as percentage growth (a) and stem growth (b) of germinated seeds of Panicum millaceum, compared with a control (water). (p < 0.05). 

It is observed that stem and root growth is favored when the MnO₂/TiO₂ catalyst is used in the degradation process of PCP.

Fig. 7a shows that the root initially grows around 30% observing an increase in growth as the degradation of PCP occurs. In addition, it is observed that when using the catalyst, the growth of the root is very noticeable. This behavior could be explained by the generation of intermediaries in different proportion, Fig. 7b where stem growth is observed, also shows that the use of the catalyst helps in the reduction of toxicity, reflected in a greater growth of the stem, observing its greater effect after 10 min of degradation, reaching up to 90% stem growth.