H 2 O 2 Based Oxidation Processes for the Treatment of Real High Strength Aqueous Wastes

This work was aimed at studying the applicability of H2O2-based oxidation processes (namely H2O2/UV, photo-Fenton, and Fenton) for the treatment of six real aqueous wastes. These wastes derived from chemical, pharmaceutical, and detergent production, and were characterised by high COD (chemical oxygen demand) and, in four cases, surfactant concentrations: overall, about 100 tests were conducted. The H2O2/UV and photo-Fenton processes proved to be very effective in COD removal, the efficiency being greater than 70%. The optimal treatment conditions for the H2O2/UV process were: 120 min reaction, H2O2/CODinitial dosage ratio = 1/2; the radiation intensity (up to 2000 W·L−1) revealed to be a crucial factor, especially in the earlier stage of the process (about 40 min): this aspect can be exploited to reduce the costs related to energy consumption. For the photo-Fenton process the following conditions were chosen: Fe/H2O2 ratio = 1/30; specific power input = 125 W·L−1; H2O2/CODinitial = 1/2; reaction time = 240 min. Photolytic reactions and the presence of dissolved oxygen revealed to be crucial factors for COD removal. The Fenton process, while showing a moderate efficiency (25% COD removal) in the treatment of high loaded wastewaters, provided excellent results in the treatment of aqueous wastes with high content of surfactants. An average yield removal of 70% for non-ionic surfactants (TAS) and 95% for anionic surfactants (MBAS) was obtained, under the following optimal conditions: Fe/H2O2 = 1/4, H2O2/CODinitial ratio = 1, and contact time = 30 min.


Introduction
Industrial aqueous wastes are significantly heterogeneous, even within the same factory, their characteristics changing by time, depending on the ongoing production activities.This poses a challenge for their treatment, there being three alternatives: (a) advanced biological processes, which may be able to treat high organic loads with a typical operational stability [1][2][3] and a very low sludge production [1,4]; (b) advanced chemical processes [5][6][7][8][9]; (c) or a combination of both [4,10,11].In the past, chemical oxidation has been used for reducing the concentrations of residual organics, removing ammonia, controlling odors, and for disinfection purposes.Nowadays, chemical oxidation processes are recommended for improving the biological treatability of refractory organic compounds and reducing the inhibitory effects of specific substances towards the microbial growth [12,13].
The present research was focused on the application of some advanced oxidation processes (AOPs) for the treatment of high strength industrial aqueous wastes.These processes combine several oxidants and/or physical treatments, such as ultraviolet light and ultrasonic irradiation with or without the use of chemical catalysts.AOPs are based on the formation of hydroxyl radicals ( • OH), which promote radical chain reactions leading to the destruction of aromatic compounds, adsorbable organic halogen (AOX), detergents, pesticides, azo dyes, and phenols [5,6,12,[14][15][16][17][18][19][20].
UV radiation is often used in combination with O 3 , H 2 O 2 , Fenton's (H 2 O 2 /Fe 2+ ) reagent and TiO 2 catalyst to accelerate the radical formation and, thus, cause an indirect photolysis [21].UV-based processes are negatively affected by high turbidity and intensive colour of wastewater [14].
In the UV/H 2 O 2 combined process, the UV radiation activates H 2 O 2 , finally leading to the formation of the • OH radical formation [22,23].The effectiveness of the UV/H 2 O 2 process depends on various conditions that affect its ability to degrade organic molecules.These conditions include the type and the concentration of the organic contaminants or dissolved inorganics (such as carbonates and metallic cations), light transmittance of the solutions, pH, temperature, and hydrogen peroxide dosage.An excessive concentration of H 2 O 2 would act as a radical scavenger reducing the rate of oxidation, while a too low H 2 O 2 dosage brings to an insufficiently hydroxyl radicals formation, thus decreasing the oxidation rate.The UV/H 2 O 2 process is sensitive to the scavenging effects of carbonate ions for pH values in the range 8-9 [7,18].
The Fenton process has been the most widely used AOP [24] for wastewater treatment due to its simplicity in terms of equipment and management operation.However, it presents some disadvantages such as the production of chemical sludge, the high acid consumption for decreasing the pH (especially in case of high alkalinity wastewaters), the high concentrations of chloride and sulphate ions (depending on the kind of ferrous salt used) in the treated waste and the significant operating costs, due to sludge disposal and hydrogen peroxide consumption.
Fenton and H 2 O 2 /UV processes involve a significant ferrous salt dosage and a high energy consumption, respectively.
The photo-Fenton process (i.e., the Fenton process with additional exposure to UV radiation) overcomes these drawbacks.The reactions involved in this process are the following [7]: The two oxidation-reduction reactions occur repeatedly until complete mineralization of the organic pollutants to CO 2 and H 2 O is achieved [7].The main advantage of the photo-Fenton process compared to the Fenton one lies in the important reduction of reagent consumption and sludge production.
Notwithstanding the huge mole of literature findings, which underline the combined influence of several parameters on process efficiency (temperature, pH, reagent dosage, inorganic salts concentration, etc.) and suggest possible reaction mechanisms [7,11,25,26], the applicability of AOPs to real wastewater is still an open issue.Therefore, any hypothesis has to be fully validated by means of experimental tests that must be carried out under conditions which must be as close as possible to the real ones.In particular, the composition of the wastewater represents a crucial factor: hence, the real wastes to be treated should be used at this scope, instead of synthetic solutions (which use is more appropriate for theoretical investigations).
The present work was aimed at testing three AOPs (namely H 2 O 2 /UV, Fenton and photo-Fenton) for the pre-treatment, upstream a biological process, of six high strength aqueous wastes, four of them also being characterised by a high content of anionic and non-ionic surfactants.Case by case, the preferable treatment process and the optimal operating conditions were defined.Results are thought to be of general interest for practitioners facing the problem of treating such kinds of real wastewaters.
During the second experimental period (phase II), mainly focused on surfactant removal, four aqueous wastes deriving from the detergents production were treated.In this case, together with very high COD concentrations, a significant presence of anionic (MBAS) and non-ionic (TAS) surfactants (up to 13,000 and 17,000 mg•L −1 , respectively) was measured.In order to reduce the acid dosage during the Fenton treatment, A.W.#3 was mixed with an acidic aqueous waste (COD = 30,000 mg•L −1 , pH < 1.5), thus obtaining two mixtures: mix#1 (83% A.W.#3 + 17% acidic waste) and mix#2 (17% A.W.#3 + 83% acidic waste).For some of the studied wastewaters, the biodegradability (as appears from the BOD 5 /COD ratio) was relatively high, thus suggesting the biological treatability also without the need of a chemical pre-treatment.Nevertheless, BOD measurements are obtained after the dilution of the samples.Indeed, in real applications, this is not the case: the so high level of contamination, poses serious problems to the biomass and microfauna of the activated sludge plant, in terms of metabolic inhibition (namely the nitrification process) and sludge settle-ability [27][28][29].The results of OUR (Oxygen Uptake Rate) tests carried out on the aqueous wastes (the values obtained vary from 3.7 to 4.2 mgO 2 g VSS with respect to the exogenous value of 4.5 ± 0.7 mgO 2 g VSS −1 •h −1 ) clearly demonstrated this.
These troubles may be even emphasized by surfactants.Therefore, a chemical pre-treatment might be a proper choice.

Pilot Scale Plants
Photo-Fenton and H 2 O 2 /UV tests were carried out by means of three different plants (A, B, and C, respectively) with the aim of studying the influence of the UV lamp type (having different energy consumption and emission spectrum) and the reactor shape/geometry, both on the atmospheric oxygen transfer and on process performance.H 2 O 2 /UV tests were conducting using all the three plants; photo-Fenton tests were performed on plants A and B.
The main characteristics of the pilot scale plants (Figure 1) are reported below.


Plant A consists of a 2 L glass reactor; the cover is welded and a discharge valve is placed on the bottom.An external jacket connected to a cryostat is used for cooling the system.On the central cone a stirring device is applied.In the lateral cone two medium-high-pressure UV lamps are placed: each lamp has a power of 125 W (emission spectrum: 280-400 nm). Plant B consists of AISI 316L stainless steel photo-reactor (8 L volume), containing one UV lamp.During the experimental work, two different kinds of lamp were used: the first is a low-pressure lamp with a power of 36 W (emission spectrum: 254 nm); the second is similar but with a power of 120 W. The following advantages may be ascribed to the use of low-pressure lamps: low surface temperature (40-50 °C), high power conversion efficiency (35%-40% of electric energy is converted into useful UV energy) and long duration (8000-10,000 h). Plant C consists of: an AISI 316L stainless steel photo-reactor (10 L volume), a high-pressure UV lamp (power: 10-30 kW; emission spectrum: 200-700 nm), a feeding pump (with flowrate adjustable from 2 to 10 L•min −1 ), a pump for H2O2 dosage (flowrate adjustable up to 8 mL•min −1 ), • Plant A consists of a 2 L glass reactor; the cover is welded and a discharge valve is placed on the bottom.An external jacket connected to a cryostat is used for cooling the system.On the central cone a stirring device is applied.In the lateral cone two medium-high-pressure UV lamps are placed: each lamp has a power of 125 W (emission spectrum: 280-400 nm).

•
Plant B consists of AISI 316L stainless steel photo-reactor (8 L volume), containing one UV lamp.During the experimental work, two different kinds of lamp were used: the first is a low-pressure lamp with a power of 36 W (emission spectrum: 254 nm); the second is similar but with a power of 120 W. The following advantages may be ascribed to the use of low-pressure lamps: low surface temperature (40-50 • C), high power conversion efficiency (35%-40% of electric energy is converted into useful UV energy) and long duration (8000-10,000 h).

•
Plant C consists of: an AISI 316L stainless steel photo-reactor (10 L volume), a high-pressure UV lamp (power: 10-30 kW; emission spectrum: 200-700 nm), a feeding pump (with flowrate adjustable from 2 to 10 L•min −1 ), a pump for H 2 O 2 dosage (flowrate adjustable up to 8 mL•min −1 ), probes for the measurement of flowrate, electrical conductivity, pH, redox potential, and temperature.The UV lamp used in this plant simulates solar radiation.
As regards the Fenton process, each test was carried out with the use of 1 L flask, rapidly mixed (by means of a magnetic stirrer).

Experimental Tests
During phase I, the high COD aqueous wastes (A.W.#1 and A.W.#2) were submitted to H 2 O 2 /UV, photo-Fenton and Fenton processes.During phase II, the Fenton process was tested for the treatment of A.W.#3 to #6, which were characterised by high concentrations of surfactants (see Table 1).
In Tables 2 and 3 (which concerns the phase I and phase II, respectively) the operating conditions (oxidant dosage, plant used, reaction time) of all tests carried out during the experimental work are reported.The dosage of reagents is reported in terms of absolute concentrations and weight ratio between reagents (H 2 O 2 /COD initial and Fe 2+ /H 2 O 2 ).The aqueous wastes were diluted in order to limit the amount of reagents and simplify the experimental procedures at the laboratory scale: the concentrations of reagents shown in Tables 2 and 3 are those actually employed for treating the diluted wastewaters.
Since the AOPs were supposed to be used as a pre-treatment to a biological stage, the oxidant dosage was generally under the stoichiometric ratio, with respect to the initial COD of the sample.
As regards UV/H 2 O 2 and photo-Fenton tests, the specific power of lamp, expressed as power (W) per volume of reactor (L), is shown in Table 2.
During some tests, air was inflated into the reactors, in order to assess possible effects on the overall efficiency of the process (e.g., in terms of mixing and mass transfer improvement or oxygen supply).Actually, in real facilities, pressurized air pipelines are often present for other purposes (e.g., the biological treatment plant or other industrial needs), so that the possibility to exploit this opportunity may be an interesting option.
Reaction time was varied from 30 to 240 min, based on the author's experience on full scale facilities.The tests exhibiting good performances were repeated in order to confirm the results obtained; about 100 oxidation tests were performed overall.

H 2 O 2 /UV Process
Figure 2 shows the effect of the hydrogen peroxide dosage on COD removal yields for A.W.#1 and A.W.#2.As highlighted above, the reagent dosages were kept below the stoichiometric ratio (with respect to the initial COD concentration), since chemical oxidation was considered to be applied as a pre-treatment upstream a biological process.
The chemical analyses were carried out three times on the same sample; the average values are reported in the results session.

H2O2/UV Process
Figure 2 shows the effect of the hydrogen peroxide dosage on COD removal yields for A.W.#1 and A.W.#2.As highlighted above, the reagent dosages were kept below the stoichiometric ratio (with respect to the initial COD concentration), since chemical oxidation was considered to be applied as a pre-treatment upstream a biological process.For both tested aqueous wastes, the increase of hydrogen peroxide dosage enhanced the COD removal efficiency.Nevertheless, the efficiency was only slightly improved for H2O2/CODinitial dosage ratios greater than 1/2, which can be regarded as an optimal value.Similarly, doubling the reaction time (from 120 up to 240 min) did not lead to an appreciable improvement of process performance.
The dosing mode of hydrogen peroxide did not affect the process efficiency in terms of COD removal (data not shown).
For all tests (with the exception of the one carried out on A.W.#2 with plant B and H2O2/CODinitial = 3/4), the ratio (H2O2 consumed/COD removed) was far below the stoichiometric value 2.125 gH2O2•gCOD −1 and it was proportional to the dosage of hydrogen peroxide dosage (Figure 2).This may be explained considering the reciprocal role of hydrogen peroxide and UV radiation: likely, the role of the oxidation mechanisms that involve H2O2, with respect to photolysis, depend on the hydrogen peroxide dosage.
The effect of the UV lamp type and power (expressed as W•L −1 ) on COD removal efficiency (tests performed on A.W.#2 with a contact time of 120 min) are shown in Figure 3.For both tested aqueous wastes, the increase of hydrogen peroxide dosage enhanced the COD removal efficiency.Nevertheless, the efficiency was only slightly improved for H 2 O 2 /COD initial dosage ratios greater than 1/2, which can be regarded as an optimal value.Similarly, doubling the reaction time (from 120 up to 240 min) did not lead to an appreciable improvement of process performance.
The dosing mode of hydrogen peroxide did not affect the process efficiency in terms of COD removal (data not shown).
For all tests (with the exception of the one carried out on A.W.#2 with plant B and H 2 O 2 /COD initial = 3/4), the ratio (H 2 O 2 consumed/COD removed) was far below the stoichiometric value 2.125 gH 2 O 2 •gCOD −1 and it was proportional to the dosage of hydrogen peroxide dosage (Figure 2).This may be explained considering the reciprocal role of hydrogen peroxide and UV radiation: likely, the role of the oxidation mechanisms that involve H 2 O 2 , with respect to photolysis, depend on the hydrogen peroxide dosage.
The effect of the UV lamp type and power (expressed as W•L −1 ) on COD removal efficiency (tests performed on A.W.#2 with a contact time of 120 min) are shown in Figure 3.
Under the same dosage of hydrogen peroxide dosage conditions, the use of a high-pressure UV lamp (plant C) led to COD removal yields higher compared to those achieved with a low-pressure UV lamp (plant B).
In all case, the COD removal yields did not significantly increase after 80 min reaction (or 40 min in case of high pressure UV lamp).Therefore, the advantage of adopting a more powerful UV lamp was clear especially during the earliest stages of each test; this aspect is very important for practical applications: in case of a batch process, the power input could be progressively decreased over time, with an important energy saving.Under the same dosage of hydrogen peroxide dosage conditions, the use of a high-pressure UV lamp (plant C) led to COD removal yields higher compared to those achieved with a low-pressure UV lamp (plant B).
In all case, the COD removal yields did not significantly increase after 80 min reaction (or 40 min in case of high pressure UV lamp).Therefore, the advantage of adopting a more powerful UV lamp was clear especially during the earliest stages of each test; this aspect is very important for practical applications: in case of a batch process, the power input could be progressively decreased over time, with an important energy saving.

Photo-Fenton Process
Photo-Fenton tests were carried out on plants A and B, equipped with medium-high and low pressure UV lamps, respectively.
The COD removal efficiency was not significantly influenced by the Fe 2+ dosage (Figure 4 shows the results obtained with plant A).In effect, under the same hydrogen peroxide and iron dosage conditions, the results were more significantly influenced by the UV radiation intensity (data not shown).This could be partially related to the production of O3 due to irradiation of dissolved oxygen.Furthermore, the UV radiation "scavenging", due to the reduction of Fe 3+ to Fe 2+ , did not affect the overall process efficiency.This can be explained considering the following aspects: the UV radiation was oversupplied (the specific energy consumption being rather high; 830 kWh•kgCODremoved −1 ); the radiation consumed was compensated by the production of • OH (due to the reaction between H2O2 and Fe 2+ ).Based on these outcomes, additional photo-Fenton tests with the use of plant C (equipped with a high-pressure lamp with an emitted light wavelength in the range 200-700 nm) were not carried out.

Photo-Fenton Process
Photo-Fenton tests were carried out on plants A and B, equipped with medium-high and low pressure UV lamps, respectively.
The COD removal efficiency was not significantly influenced by the Fe 2+ dosage (Figure 4 shows the results obtained with plant A).In effect, under the same hydrogen peroxide and iron dosage conditions, the results were more significantly influenced by the UV radiation intensity (data not shown).This could be partially related to the production of O 3 due to irradiation of dissolved oxygen.Furthermore, the UV radiation "scavenging", due to the reduction of Fe 3+ to Fe 2+ , did not affect the overall process efficiency.This can be explained considering the following aspects: the UV radiation was oversupplied (the specific energy consumption being rather high; 830 kWh•kgCOD removed −1 ); the radiation consumed was compensated by the production of • OH (due to the reaction between H 2 O 2 and Fe 2+ ).Based on these outcomes, additional photo-Fenton tests with the use of plant C (equipped with a high-pressure lamp with an emitted light wavelength in the range 200-700 nm) were not carried out.
Sustainability 2017, 9, 244 9 of 14 Table 4 shows the COD removal efficiency achieved in two tests with different duration: 120 min (photo-Fenton test 1-P-F 1 ) and 240 min (photo-Fenton test 2-P-F 2 ).These tests were carried out on A.W.#1 using plant A (medium-high-pressure UV lamp, specific power = 125 W•L −1 ); H 2 O 2 /COD initial and Fe 2+ /H 2 O 2 ratios were kept at 1/2 and 1/30, respectively.It can be observed that after 120 min reaction, an additional COD removal was achieved.Although after 80 min the residual hydrogen peroxide was equal to zero (Table 4), the oxidation was still in progress.It can be argued that the possible presence of dissolved oxygen (activated by UV radiation) and photolysis effect were crucial factors for COD removal.The presence of dissolved oxygen, due to the strong mixing conditions, was in effect also confirmed by [31][32][33].

Fenton Process
The Fenton tests were carried out both on A.W.#1 and A.W.#2 with the same hydrogen peroxide dosage (H 2 O 2 /COD initial = 1/2) and a contact time of 120 min.The results are reported in Table 5.
A higher dosage of iron increased the COD removal yields, especially for A.W.#2 (Table 5); however, the removal efficiencies obtained during Fenton tests were lower, compared to the other AOPs.This was probably due to the low amount of dosed iron (the ratio [Fe 2+ ] dosed /[H 2 O 2 ] dosed being lower than 1) that was not sufficient to completely consume all the added hydrogen peroxide (see Figure 5).Likely, the high presence of organic matter negatively interferes with the Fe 2+ regeneration process, thus leading to the accumulation of H 2 O 2 [33][34][35]: organic matter is an • OH scavenger, involving the interruption of the radical chain reactions that lead to Fe 2+ regeneration through this equation: where the radical HO 2 • can be produced by the following reaction: The dosing mode did not significantly influence the process efficiency; in fact, the impulse addition of reagents (initial dosage) led to an initial increase in the reaction rate, but the final results were similar (data not shown).
Table 5. Fenton process: effect of iron dosage on COD removal efficiency (H2O2/CODinitial = 1/2; contact time = 120 min).Likely, the high presence of organic matter negatively interferes with the Fe 2+ regeneration process, thus leading to the accumulation of H2O2 [33][34][35]: organic matter is an • OH scavenger, involving the interruption of the radical chain reactions that lead to Fe 2+ regeneration through this equation:

Fenton Test
where the radical HO2 • can be produced by the following reaction: • OH + H2O2 → HO2 • + H2O (4) The dosing mode did not significantly influence the process efficiency; in fact, the impulse addition of reagents (initial dosage) led to an initial increase in the reaction rate, but the final results were similar (data not shown).

Fenton Process
The effect of iron dosage on the removal efficiency of surfactants (MBAS and TAS) is reported in Figure 6.
The increase of iron dosage (from 1/4 to 1/2) involved a significant enhancement of total surfactant removal only for mix#1.Moreover, the anionic surfactants (MBAS) were more easily removed, with respect to the non-ionic surfactants (TAS).

Fenton Process
The effect of iron dosage on the removal efficiency of surfactants (MBAS and TAS) is reported in Figure 6.The influence of hydrogen peroxide dosage on the surfactant removal efficiency is reported in Figure 7.The increase of iron dosage (from 1/4 to 1/2) involved a significant enhancement of total surfactant removal only for mix#1.Moreover, the anionic surfactants (MBAS) were more easily removed, with respect to the non-ionic surfactants (TAS).
The influence of hydrogen peroxide dosage on the surfactant removal efficiency is reported in Figure 7.The influence of hydrogen peroxide dosage on the surfactant removal efficiency is reported in Figure 7.It can be seen that the increase of H2O2 dosage up to a value of the H2O2/CODinitial ratio equal to 1, led to an improvement of surfactants removal, especially for TAS.A further increase of the hydrogen peroxide dosage (from 1 to 1.5 H2O2/CODinitial) did not lead to any improvement in surfactant removal yields (both for anionic and non-ionic forms).
The influence of contact time on surfactants removal rate is shown in Table 6.
The removal of total surfactants did not significantly increased along with contact time, with the exception of A.W.#4: in this case, the removal yields, especially for TAS, passed from 40% to 60% and 80%, by increasing the reaction time from 30 min, to 60 and 120 min, respectively.As concerns A.W.#5 and A.W.#6, the effect of reaction time on surfactants removal was not significant, so that the optimal contact can be identified in 30 min.It can be seen that the increase of H 2 O 2 dosage up to a value of the H 2 O 2 /COD initial ratio equal to 1, led to an improvement of surfactants removal, especially for TAS.A further increase of the hydrogen peroxide dosage (from 1 to 1.5 H 2 O 2 /COD initial ) did not lead to any improvement in surfactant removal yields (both for anionic and non-ionic forms).
The influence of contact time on surfactants removal rate is shown in Table 6.
The removal of total surfactants did not significantly increased along with contact time, with the exception of A.W.#4: in this case, the removal yields, especially for TAS, passed from 40% to 60% and 80%, by increasing the reaction time from 30 min, to 60 and 120 min, respectively.As concerns A.W.#5 and A.W.#6, the effect of reaction time on surfactants removal was not significant, so that the optimal contact can be identified in 30 min.
Additionally, for mix#1, the slightly higher surfactant removal yield was due to the modification of iron dosage (from 1/5 to 1/2) rather than the increase of reaction time.

Figure 1 .
Figure 1.Pilot plants used for photo-Fenton (A and B) and H2O2/UV (A, B, and C) tests.

Figure 1 .
Figure 1.Pilot plants used for photo-Fenton (A and B) and H 2 O 2 /UV (A, B, and C) tests.

Figure 2 .
Figure 2. H2O2/UV process: effect of hydrogen peroxide dosage on COD removal efficiency.

Figure 2 .
Figure 2. H 2 O 2 /UV process: effect of hydrogen peroxide dosage on COD removal efficiency.

Figure 6 .
Figure 6.Fenton process: effect of iron dosage on the removal of surfactants.

Figure 6 .
Figure 6.Fenton process: effect of iron dosage on the removal of surfactants.

Figure 7 .
Figure 7. Fenton process: effect of hydrogen peroxide dosage on the removal of surfactants.

Figure 7 .
Figure 7. Fenton process: effect of hydrogen peroxide dosage on the removal of surfactants.

Table 1 .
Main characteristics of aqueous wastes.

Table 2 .
Operating conditions adopted during H 2 O 2 /UV, photo-Fenton, and Fenton tests (phase I).

Table 3 .
Operating conditions adopted during Fenton tests (phase II).

Table 6 .
Fenton process: effect of contact time on the removal of surfactants.