Polyaromatic Hydrocarbons (PAHs) are toxic organic contaminants arising from fossil fuel combustion and industrial processing , and the need to remove these aromatic compounds is driven by their carcinogenic nature, environmental persistence and chronic toxicity . In soil, PAHs accumulate in the vadose zone due to their hydrophobic nature and remediation has been achieved via incineration, chemical oxidation, composting, land farming, and phytoremediation [3,4]. Controlled microbial degradations present an alternative cost-effective treatment for PAH contaminated soils, limited only by the poor aqueous solubility of PAHs.
Polycyclic Aromatic Hydrocarbons (PAHs) are a group of organic compounds which consist of at least two benzene rings in angular, linear, or cluster arrangement. In most cases, they are colorless, white, or yellow solids. PAHs are characterized by high melting and boiling points, low vapor pressure, and low solubility in aqueous media. Moreover, it is generally accepted that some PAHs are cancerogenic and mutagenic [1,2]. Even though there are many described substances occurring in the environment, most regulations aim at only some of them. For instance, U.S. Environmental Protection Agency (USEPA) listed 16 PAHs according to their ecological impact and effects on human health . PAHs are widely spread substances in the environment, resulting from natural and anthropogenic processes such as fires, petroleum spills, vehicles, and primarily by the incomplete combustion of the fuels. Many PAHs are also related to urban road surfaces due to vehicles and abrasion of asphalt. As a result, mentioned substances can migrate to air, water, and soil. They also have a tendency to deposit near to their source in the environment, due to low volatilization potential and high molecular weight . In general, aromatics are considered as substances that are hard to remove biologically, and thus, other techniques need to be developed . One which can be used in water treatment is acoustic cavitation technology, which is defined as the formation of acoustic bubbles due to ultrasound, and their subsequent collapse in given conditions. During the collapse of the bubbles, pollutants can be eliminated by the mechanical effects, oxidation, and occurrence of high pressure and temperature conditions [6,7]. This study aims to determine the PAH yields in a petrochemical industry wastewater via sonication The impact of ultrasound treatment on the removal of PAHs were investigated at different operatonal conditions and chemical additions.
Material and Methods
PAH Analytical Procedures
Analytical procedures were adapted from those described by Vandermeer and Daugulis (2007). PAH concentrations in methanol were quantified via fluorescence spectroscopy using a QuantaMaster QM-2000-6 fluorescence spectrometer All samples were diluted in methanol to within the linear range of detection (0-0.1 mg/L) as described by Rehmann et al. (2008) . Silicone oil concentrations were determined by diluting oil samples in methanol to within detection range and analyzing via fluorescent spectrophotometry. Polymeric concentrations of PAHs were obtained by desorbing a small quantity of polymers (circa 0.1 g) in 5 mL of methanol as detailed by Rehmann et al. (2008) . Extracts were then analyzed via fluorescence spectrophotometry, and equilibrium polymer concentrations were determined using partitioning coefficient curves, detailed subsequently.
Measurements of all pollutants in petrochemical industry
All the pollutants in petrochemical industry were measured according to Standard Methods (2012) .
Sonication of PCI ww
Raw wastewaters taken from the influent of the aeration unit of a PCI ww treatment plant in Izmir were analyzed. The characterization of PCI ww was shown in Table 1 for minimum, medium and maximum values. All measurements were carried out three times and the results given as the means of triplicate samplings with Standard Deviation (SD) values.
In order to detect the linear sensitivities between PAH removals and operational parameters ANOVA tet statistics were performed with Excel 2016.
Measurements of OH concentrations
The OH concentrations during sonication were measured by the equation generated by David, (2009) .
Results and Discussion
Effect of Increasing Sonication Temperature on the Removals of CODdis, TOC and PAHs in PCI ww
At the beginning of the studies the raw PCI ww samples were sonicated at 30oC and 60oC at a pH=7.0 at increasing sonication times from 5 min up to 60 min to determine the lowest sonication time for maximum CODdis removal efficiencies. The lowest sonication time was determined as 60 min for the maximum CODdis removals (Table 2).
44.05%, 61.22% and 89.94% CODdis removals were obtained after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 30oC. An increase of 9.78% in CODdis yield was obtained after 150 min at 30oC, compared to the control (E=80.16% CODdis at pH=7.0 and at 25oC). 46.50%, 67.37% and 92.48% CODdis removals were obtained after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 60oC. The contribution of 60oC temperature on CODdis removals were 5.07% and 12.32% after 120 and 150 min, respectively, compared to the control (E=62.30% and E=80.16% CODdis after 120 and 150 min at pH=7.0 at 25oC). The maximum CODdis removal was 92.48% after 150 min at pH=7.0 and at 60oC. A significant linear correlation between CODdis yields and increasing sonication temperature was observed (R2=0.93, F=15.43, p=0.01). Increasing temperatures (from 25oC to 30oC and to 60oC) increased the CODdis removal of PCI ww after sonication process since sonodegradation reaction rates in cavitation process increased with increasing temperature during sonication at increasing sonication times. As a result, increasing temperature increased the CODdis removal efficiency in PCI ww after sonication experiments.
47.70%, 63.38% and 90.89% TOC removals were measured after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 30oC. An increase of 12.52% in TOC removal was found after 150 min at 30oC, compared to the control (E=78.37% TOC at pH=7.0 and at 25oC). A significant linear correlation between TOC yields and increasing sonication temperature was observed (R2=0.95, F=17.78, p=0.01). 49.50%, 70% and 94.23% TOC removals were obtained after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 60oC. The contribution of 60oC temperature to the TOC removals were 7.26% and 15.86% after 120 and 150 min, respectively, compared to the control (E=62.74% and 78.37% TOC after 120 and 150 min at pH=7.0 and at 25oC). The maximum TOC removal efficiency was 94.23% after 150 min at pH=7.0 and at 60oC. A significant linear correlation between TOC yields and increasing sonication temperature was observed (R2=0.94, F=17.11, p=0.01). Increasing temperatures (30oC and 60oC) increased the TOC removals in PCI ww after sonication process. Sonodegradation reaction in cavitation process was rapidly performed with increasing temperatures at long sonication times such as 150 min. The treatment by sonication converts CODdis and TOC to much smaller sonodegraded compounds. Low sonication temperature (25oC) did not provide high degradation yields for CODdis and TOC.
Effect of Increasing Temperature on the Removal of PAHs in PCI ww at Increasing Sonication Times
Raw PCI ww samples were sonicated in a sonicator at 30oC and 60oC during 60 min, 120 and 150 min at pH=7.0. Similar total PAH removal yields were found at 25oC (E=54.92% total PAHs at pH=7.0) and 60oC (E=54.21% total PAHs at pH=7.0) after 60 min. In other words, increasing the temperature from 25oC to 30oC and 60oC did not contribute to the PAHs removal after 60 min. The total PAHs removal decreased slightly at a temperature of 30oC with the same sonication time. Similarly, the total PAHs removals at 30oC remained at the same level as 25oC after 120 min. Increasing the temperature from 25oC to 60oC increased the total PAHs removal efficiency after 120 and 150 min. In general, as the sonication time increased from 60 min to 150 min, the total PAHs removal increased. The maximum total PAHs removal efficiency was 96.90% after 150 min at pH=7.0 and at 60oC. A significant linear correlation between total PAHs yields and increasing sonication temperature was observed (R2=0.83, F=10.41, p=0.01).
Removal efficiencies in seventeen PAHs were measured in the influent and in the effluent of the sonication experiments after 60 min, 120 and 150 min at 30oC (Table 3). As seen in Table 4, all the removal yields of individual PAHs increased as the sonication time increased from 60 min to 150 min at a temperature of 60oC. The yields for all individual PAHs were above 91% except for BbF (86.21%) after 150 min of sonication time at a temperature of 60oC (Table 4). This showed that sonication at high temperature increased the yields in all PAHs species.
The results of this study showed that the PAHs removal was not dependent on the ring numbers of benzene for the individual PAHs species. Therefore, it can be concluded that a correlation between the removal of the PAHs and water solubility, Henry’s law constants and vapor pressure, was not observed at 60oC and the difference is not significant (R2=0.54, F=3.34, P=0.001). It was found that the PAHs degradation is a function of long sonication time (150 min) and high temperature (60oC). A high correlation was found between PAHs yields, time and temperature (R2=0.97). This correlation is also significant (F=17.78 p=0.001). The two experimental conditions (640 W sonication power and 35 kHz sonication frequency) employed in this study influenced the important physical parameters related to cavitation bubbles such as the extent of radical production from the bubble, the thickness of the liquid shell surrounding the bubble, the concentration of the PAHs in the interfacial region and extent of radical scavenging in the medium (Chakinala et al., 2008a; 2008b). For this reason, most probably, a significant difference was not observed between the lower molecular weight PAHs (e.g. those with two, three or four aromatic rings) and the higher molecular weight, more hydrophobic PAHs for their individual removals (R2=0.82, F=13.67, p=0.001) at 60oC. On the other hand, low-frequency ultrasound is expected to induce destructive effects for hydrophobic solutes, since they can easily diffuse near cavitation bubbles and undergo pyrolytic destruction inside the collapsing bubble or hydroxylation and thermal decomposition at its interfacial sheath [10-12].
Given that all PAHs with high molecular used in this study are relatively non-volatile, their ability to migrate towards the bubble and rapidly decompose at the interface is likely to be dictated by their hydrophobicity. It appears that the more hydrophobic PAHs are all readily susceptible to sonochemical degradation and high removal yields (86-98%) is achieved within 150 min of irradiation with the conditions under consideration (Table 3).
Among the PAHs studied, only in the case of PY increasing the temperature did not influence its removal (Table 3). The yield of PY decreased slightly as increasing the temperature from 120oC to 150oC while the removals of PHE, BghiP and the rest of the PAHs increased. The slight decrease in degradation rate observed for PY may be due to the increased solution temperature. For PY, an increased solution temperature might imply a slightly higher adsorption on the air-water interface and an increased diffusivity. These factors act to affect the slight accumulation of PY on the interface in different ways. As the temperature increased, the increased diffusivity may contribute to more available PY at the subsurface for adsorption. Thus, a slight increase in PY removal efficiency was observed from 25oC to 60oC (Table 3). The decrease in removal efficiency at 150oC may be due to less favorable adsorption resulting in reduced accumulation on the interface. Although, the effects of increasing temperature on the sonolytic removal efficiencies were also examined for all PAHs, in this section only PHE and BghiP are discussed.
The removal yields of PHE and BghiP increased with increasing temperature. For partitioning into the bubble, the increased solution temperature will allow PHE and BghiP molecules to more easily enter the cavitation bubble (i.e., increase diffusivity). At higher temperatures this effect will be enhanced and this may be the cause of the increase in removal rates for PHE and BghiP at 150oC.
The results of this study showed that although a strict correlation between the remaining percentage of the aforementioned PAHs and physicochemical properties was observed after 30 min, 60 and 120 min (R2=0.89, P=4.89, p=0.001), a significant correlation was not observed between the remaining percentages of PAHs and their properties after 150 min (R2=0.45, p=16.56, P=0.01) and over 90% removal rates of the all PAHs was achieved. Furthermore, it becomes evident that a larger hydrophobicity resulted in smaller reaction kinetic constants of the PAHs. Low initial PAHs concentrations led to low reaction rates and also to smaller residual concentrations. The coefficient of the correlation between the residual concentration and the total initial concentration of the single PAHs was strong and highly significant (R2=0.85, p < 0.001).
Several investigators have reported contradictory findings regarding the temperature effect. In certain reaction systems for instance, the net effect of an increase in To and consequently Tmax, is an increase in degradation rates. This occurs up to the point at which the cushioning effect of the vapor begins to dominate the system and further increases in liquid temperature result in reduced reaction rates. The fact that removal decreases with increasing liquid temperature is believed to be associated with the effect of temperature on both the bubble formation energy threshold and the intensity of bubble implosion. The maximum temperature (Tmax) obtained during the bubble collapse is given as follows Eq. (1):
where, To is the liquid bulk temperature, Po is the vapor pressure of the solution, P is the liquid pressure during the collapse and γ is the specific heat ratio (i.e. the ratio of constant pressure to constant volume heat capacities). Increased temperatures are likely to facilitate bubble formation due to an increase of the equilibrium vapor pressure; nevertheless, this beneficial effect is compensated by the fact that bubbles contain more vapor which cushions bubble implosion and consequently reduces Tmax. In addition to this, increased temperatures are likely to favor degassing of the liquid phase, thus reducing the number of gas nuclei available for bubble formation .
In order to detect the effect of increasing sonication time on the yields of less hydrophobic PAHs with low benzene rings (PHE, ANT, CHR, BbF and PY) and more hydrophobic PAHs with high benzene yields (DahA and BghiP) the raw PCI ww samples were sonicated at a temperature of 60oC at increasing sonication times (from 60 to 120 min and 150 min). The increase in temperature to 60oC will increase the kinetic reaction to a point at which the cushioning effect of the vapor in the bubble begins to dominate the system. Since the PAHs are relatively non-volatile, the degradation reaction took place in the gas-liquid film between the cavitating bubble and the bulk liquid mixture. As the reaction temperature increased, the rate of diffusion of PAHs from the bulk liquid phase to the reaction zone was accelerated. An increase in temperature up to 60oC improved the intensity of the cavitation, thus increasing the amount of free radicals produced within the bubble. It was suggested that these free radicals were required for the degradation reaction to occur and that they diffuse from the vapor cavity to the gas-liquid film where reaction ensues. As the rates of the counter diffusing reactants became comparable, a further increase in temperature (up to 80–90oC) had little or no effect on the reaction (i.e. the percent change in PAHs concentration reached a plateau as a function of temperature).
The results of the study showed that as the sonication time was increased the yields of BghiP, CHR, ANT and BbF increased while the destruction yields of DahA, PHE and PY decreased after 150 min. The effect of sonication time on the BghiP, CHR, ANT and BbF removals was significant for 150 min at 60oC (R2=0.98, F=14.56, p < 0.01). No significant correlation was found between the DahA, PHE and PY yields and 150 min at 60oC (R2=0.58, F=6.39, p < 0.01). The treatment by sonication converts PAHs with multiple benzene rings to much smaller compounds. In such cases it is obvious that higher sonication times are needed for complete mineralization. Short sonication times (60 min) did not provide high degradation yields for refractory PAHs since they were not exposed for a long enough time to ultrasonic irradiation.
Therefore, a decrease in the percentage of remaining PAHs was expected at longer sonication times due to sufficient radical reactions through cavitation. It was found that the yields in PAHs with high benzene rings (DahA and BghiP; E=90–92%) were as high as the PAHs with lower benzene rings (BbF, CHR, PHE, PY and ANT; E=92–95%). In order to explain this, it is important to mention some of the physical/chemical properties of these compounds. The solubilities of DahA and BghiP are approximately 15 times lower than for CHR and BbF, 100 times lower than for ANT and 10000 times lower than for PHE. Moreover, the Henry’s law constant of DahA and BghiP are also lower than ANT, PY, BbF and CHR. Given the lower solubilities and the low Henry’s law constant of DahA and BghiP and following the studies on non-volatile, hydrophobic compounds [14,15], we would expect lower degradation yields for DahA and BbF compared to ANT and CHR. There are two possible explanations for the enhancement in the yields for DahA and BghiP: (1) They have high removals as the ANT, PY, BbF and CHR, via the OHl pathway (hydroxylation) and/or (2) DahA and BghiP accumulate at the interface of the liquid gas phase to a greater degree than the other PAHs by a subsequent entrapment of the pollutant vapor in the cavitation bubble (pyrolysis). Reported second order reaction rate constants for the DahA and BghiP are lower than those of the other PAHs. Therefore, suggestion (1) can be ignored. The thickness of the liquid shell surrounding the bubble, in which temperature rises, is higher for hydrophobic organics like PAHs. In this shell, in principle, the thermal penetration depth in the bulk medium varies directly with the bubble size and the thickness of liquid layer around the bubble that gets heated up is larger for saturated medium. Finally, the extent of pyrolysis in the liquid shell depends on the thickness of this shell and concentration of the pollutant molecules in it.
If the pollutant is hydrophobic in nature, characterized by low solubility in water, it tends to partition between the bulk medium and bubble interface . The bubble–bulk interface also has a hydrophobic character, and hence, the concentration of the hydrophobic pollutant molecules in this region is much higher than the bulk. Therefore, hydrophobic PAHs concentrations were high in the interfacial region between bubble and bulk. The PAHs transfer process from the PCI ww to the cavitational bubbles and the removal of PAHs are jointly controlled by the hydrophobicity of PAHs. Increasing hydrophobicity by low Henry’s law constants induces destructive effects for hydrophobic PAHs, since they can easily diffuse near the cavitation bubbles and undergo pyrolytic destruction inside the collapsing bubble . Given that PAHs with high molecular weights in PCI ww have the ability to migrate towards the bubble, rapid decomposition at the interface is likely to be dictated by their hydrophobicity. It appears that the more hydrophobic DahA is readily susceptible to sonochemical degradation and nearly complete removal (99%) is achieved within 150 min of irradiation at 60oC. On the other hand, the hydrophobicity of an organic compound can be described fairly well by its octanol–water partition coefficient and water solubility. The higher octanol–water partition coefficient of hydrophobic PAHs results in higher PAHs removal although there are a few exceptional cases. In the exceptional cases, such as the yields of CHR and BbF, the vapor pressure and/or the reactivity of PAHs with intermediates (i.e. free radicals, atoms and active molecules) generated in situ in bulk liquid, play a simultaneous role, at least to a certain extent. Although, the BbF is more hydrophobic (having higher octanol–water partition coefficient) than that of CHR its removal efficiency is lower than that of CHR. This could be explained as follows: the hydrophobicity of BbF is higher with low Henry’s law constant (log POW=5.98 at 25oC) compared to CHR with high Henry’s law constant (log POW=5.71 at 25oC), but its vapor pressure (VP, 5.00x10-7 mm Hg at 25oC) is low compared to CHR (VP, 6.23x10-9 mm Hg at 25oC). Hence, the yield of BbF is lower compared to CHR after 150 min. A significant linear relationship was found between the hydrophobic PAHs yields and the Henry’s law constants and the solubilities of these PAHs (R2=0.96, F=14.67, p=0.001) while the relationship between vapor pressure and the PAHs removals was not significant (R2=0.65, F=7.95, p=0.001). Although, the BbF, DahA, CHR and ANT removals increased at increasing sonication times among the PAHs studied it was found that PHE, PY and BghiP concentrations decreased as the sonication time increased from 60 to 120 min while the concentration of these PAHs increased after 150 min. The reason of this could be explained by the ultimate destruction of these PAHs after 120 min. This sonication time could be accepted as the optimum time for the maximum degradation of PHE, PY and BghiP to the inter-metabolites.
Produced Metabolites from PHE, PY and BghiP PAHs in PCI ww
With the increase of sonication time to 150 min PHE was removed with an efficiency of 96.34% since this PAH degraded to the by-products NAP, p-hydroxy-benzoic acid and FLN (Table 5). PY was degraded to di–hydroxy pyrene and to benzoic acid while BghiP degraded to pyrene di-hydrodiol and to benzoic acid (Table 4).
Free radical and pyrolysis reactions produce different products, with relative abundances depending on the nature of the solute and its concentration. For example, t was found that FLN and benzoic acid are the sonication metabolites of PHE and methyl radicals (CH3l) formed from the pyrolysis of solvent-hexane [18,19]. Some researcher also found that sonolysis of simple hydrocarbons creates the same kind of products associated with very high-temperature pyrolysis. CH3l and ethyl (CH3CH2l) radicals are expected to be formed when hexane is decomposed sonochemically as a solvent [20,21]. CH3l has also been shown to form during the pyrolysis of acetone molecules. These alkyl radicals then react with PHE to form different types of methyl- and ethyl-phenanthrene by-products. In our study, although CH3l and CH3CH2l were not measured the metabolites found from the sonication of PHE (FL, NAP and benzoic acid) agree with the results found by older and more recent research as reported by some researchers [22,23]. The mechanism of pyrolysis of PHE had two pathways: (1) Loss of one carbon in PHE and yielding CH4 and FLN and (2) fragmentation resulting in a four carbon fragment and NAP. Therefore, FLN is an indication of a pyrolysis by-product formed from the PHE due to high-temperature reactions in or near a cavitation bubble as reported by Adewuyi (2001). In our study, CH4, H2 and CO2 gases were identified in the headspace of the sonicator reactor. The GC spectra of these gases are illustrated in (Figure 1).
Figure 1: CH4(g), H2(g) and CO2(g) spectra measured in the headspace of the sonicator by GC (sonication power=640 W, sonication frequency=35 kHz).
The increasing of PAHs concentrations after 150 min could be attributed to the re-formation of the PHE, PY and BghiP from the by-products. We suspected that the increase of PHE with longer sonication time may be due to the re-formation of PHE from the by-products mentioned above and from the FLN. The FLN formed during the sonication of PHE may be attacked by CH3l to regenerate PHE [26,27]. Cyclization reactions of PHE with methyl- or ethyl-naphthalene may also contribute to the re-formation of PHE. A radical mechanism proposed by It was shown that PHE formation from pyrolysis of 9,9–dimethyl–fluorene at 800oC by a free radical ring expansion process . It was also reported NAP and benzene formation during PHE pyrolysis at < 900oC . Furthermore, PHE pyrolysis at 700 and 850oC and reported that NAP is one of the pyrolysis products of PHE . Therefore, the NAP by-product detected in this study may be direct pyrolysis products of PHE. Similarly, PY yields increased after 120 min since the PY degraded to di–hydroxy pyrene and to benzoic acid (data not shown). Then we suspected that PY reproduced from the hydroxy-pyrene since the PY yields decreased. Similarly, the yield of BghiP increased after 120 min with sonodegradation to its by-products namely, benzoic acid and pyrene di-hydrodiol at 63oC (Table 5). However, the yields of the BghiP decreased after 150 min. This could be explained by the re-formation of BghiP from pyrene di-hydrodiol and benzoic acid (Table 5).
In this study, the presence of CH4, H2 and CO2 gases indicated not only the destruction of the PAHs but also confirmed the mechanism ‘‘pyrolysis” with degassing of the medium throughout sonication. The results given above are consistent with our results. When PHE is sonicated in an organic solvent, it is expected that a certain number of PHE molecules will migrate into the gaseous cavitation bubbles. Then PHE molecules are available to migrate towards the cavitation bubble interfaces or volatilize into the cavitation bubbles to react under pyrolysis thus leading to a lower percentage remaining (Adewuyi, 2001). In addition, at higher concentrations of PHE, the solute is more likely to compete for reaction with CH3l, which could also contribute to the loss of PHE .
It was observed that the PAHs with multiple benzene rings were also degradable with high yields, even though some studies demonstrated that sonication is not effective for PAHs with a large number of benzene rings (Psillakis et al., 2004). The PAHs yields obtained in our study are high in comparison to the removal performances of PAHs in the studies given below. In a study 77% PAHs removal efficiency was observed for the sonochemical degradation of 50 μg/l of initial PAHs mixture concentration (NAP, ACL and PHE) in water after 120 min, at 40oC, at 150 W and at 24 kHz . It was found 31-34% and 44-50% PAHs removals in mesophilic (35oC) and thermophilic (55oC) conditions for NAP and PY at 20 kHz and at 70 W, after 110 min, before anaerobic digestion . The yields obtained in the aforementioned studies are low in comparison to the removal performances of PAHs found in this study. It was observed that the PAHs with high benzene rings were also degradable with high yields, even though some studies demonstrated that sonication is not effective for PAHs with high benzene rings (Laughrey et al., 2001).
Effect of Dissolved oxygene (DO) Concentrations on the Removal of PAHs in PCI ww at Increasing Sonication Times and Temperatures
The raw PCI ww samples were oxygenated with increasing DO concentrations (2 mg/l, 4 mg/l, 6 and 10 mg/l) with pure O2 before the sonication experiments. 90.22%, 92.27%, 93.77% and 94.32% total PAHs removal efficiencies were measured in 2 mg/l, 4 mg/l, 6 and 10 mg/l DO concentrations, respectively, after 150 min at pH=7.0 and at 30oC (data not shown). Only 4, 6 and 10 mg/l DO concentrations increased the total PAHs removal efficiencies from 90.11% to 92.27-94.32% in comparison to the non-oxygenated samples (control, E=90.11% total PAHs at pH=7.0) at a temperature of 30oC and a sonication time of 150 min. In other words, the effect of increasing DO concentrations on the total PAHs removals was found to be insignificant at a temperature of 30oC and a sonication time of 150 min. The total PAHs yields also increased significantly (from 45.34% to 64.66% and from 62.40% to 80.34%) after 60 and 120 min, respectively, compared to the control at all DO concentrations for the same temperature. 94.67%, 95%, 96.79% and 97.23% total PAHs removal efficiencies were obtained in 2 mg/l, 4 mg/l, 6 and 10 mg/l DO at pH=7.0 and at 60oC after 150 min. As shown, the total PAHs removal efficiencies increased as the sonication time increased. However, the PAHs yields did not show a significant increase at increasing DO concentrations compared to the control (DO=0 mg/l while E=96.90% total PAHs at pH=7.0 and at 60oC). The maximum total PAHs removal efficiency was 97.23% after 150 min in DO=10 mg/l at 60oC. Increasing the temperature also did not increase the total PAHs removal efficiencies at increasing DO (ANOVA, F=2.51, p=0.001).
The yield of 2 and 3-ring PAHs removals was almost 96.47% for FLN and 96.77% for PHE, respectively, in raw PCI ww samples containing 10 mg/l DO at 60oC after 150 min. The removal efficiencies of four ring PAHs (CHR, BkF, BaP and IcdP) were also higher (efficiency varied between 93.98 and 98.22%) in the samples oxygenated with 10 mg/l DO at 60oC in comparison to the PAHs containing few benzene rings after 150 min. Similarly, around 96.87-97.01% removal efficiencies were obtained for PAHs with five rings BghiP, DahA. The PAHs intermediates (1–methylnaphthalene, 9–hydroxyfluorene, 9,10–phenanthrenequione, benzoic acid, 1,2,3–thiadiazole–4–carboxylic acid) in PCI ww were measured with HPLC in DO=10 mg/l after 120 min at 25oC. The initial total PAHs concentration of 1378.77 ng/ml decreased to 413.63 ng/ml in DO=10 mg/l after 120 min at 25oC. From 1378.77 ng/ml initial PAHs 274.24 ng/ml 1–methylnaphthalene, 176.07 ng/ml 9–hydroxyfluorene, 55.29 ng/ml 9,10–phenanthrenequione, 129.88 ng/ml benzoic acid and 48.26 ng/ml, 1,2,3–thiadiazole–4–carboxylic acid were produced after 120 min at 25oC. After 120 min the remaining PAHs concentration was found to be high (695.03 ng/ml). The initial PAHs concentration of 1378.77 ng/ml converted to 683.74 ng/ml PAHs intermetabolites in question as aforementioned. The low removal efficiency of the total PAHs (70%) could be attributed to the studied low temperature (25oC) although, DO increased the PAH yields at 60oC. The PAHs intermediates namely, 1–methylnaphthalene, 9–hydroxyfluorene, 9,10–phenanthrenequione, benzoic acid, 1,2,3–thiadiazole–4–carboxylic acid were sonodegraded with yields of 80.11%, 87.23%, 95.99%, 90.58% and 96.50%, respectively, in DO=10 mg/l after 150 min at 25oC.
O2 mediates the rate of mineralization of PAHs. The higher DO content of PCI ww resulted in faster rates of PAHs degradation through sonication, supporting the hypothesis that increased oxygenation was largely responsible for the enhanced PAHs degradation. Although, O2 exposure is an important factor in PAHs degradation it was reported that saturating the solutions cause decreases in O2Hl production resulting in low PAHs yields. In the presence of O2, the reactive radicals (Ol , OHl, OOHl) will be produced by a series of reactions and thus contribute to the removal of the PAHs [34,35]. Although, O2 and air have similar ratios of specific heats and thermal conductivity, the highest formation rate of H2O2, which was induced from the recombination of reactive radicals (OHl and O2Hl) was observed under O2 (Kojima et al., 2005). In oxygenated solutions, the O2Hl formed by; Eq. (2).
will decay with generation of H2O2. However, the production of the O2Hl increases the oxidation process due to further formation of H2O2 by its recombination reaction  Eq. (3):
The O2 dissolved in water contributes to the effective formation of many oxidants such as OHl and Ol by pyrolysis of O2. At the same time, these oxidants may further react to produce O2Hl [37,38]. On the other hand, the maximum temperature produced in collapsing cavitation bubble was higher by sonication in Ar than O2, because of high specific heat ratio of Ar (Kojima et al., 2005) and O2. (Suslik, 1988). With the combination of O2 and Ar, it is considered that many oxidants such as O3 and Ol were produced by stronger pyrolytic effect arising from the presence of O2 in addition to OHl typically formed by thermal decomposition of H2O. Moreover, it is also considered that production of OHl was accelerated by the reaction of O2Hl and H2O with oxidants O3 and Ol . Hence, it is assumed that more reactive species in Ar/O2 were formed compared with those in O2/N2 and in Ar. It was reported that the decomposition of dichlorinated groups in the Ar/O2 mixture was the highest due to the coupled effects of pyrolysis and radical reaction .
As aforementioned aqueous phase sonolysis is likely to result in the formation of H2O2 which may be formed through the recombination of OHl at the gas–liquid interface and/or in the solution bulk. Moreover, if the solution is saturated with O2, (H2O) and more OHl are formed in the bubble, the recombination of the former at the interface and/or in the solution bulk results in the formation of additional H2O2 . It was found that PAHs mainly decomposed at the gas–liquid interface through both OHl oxidation and pyrolytic reactions (for instance, CH4 and CO2 were identified as the primary pyrolysis by-products) rather than in the solution bulk . This was attributed to the fact that the PAHs, in this study, although less water soluble and non-volatile molecules, tended to accumulate at the gas–liquid interface. Furthermore, it was also found that the concentration of H2O2 formed during PAHs sonication experiments was substantially lower than that formed during water sonication in the absence of PAHs.
In order to verify the generation of H2O2, sonication tests in the absence of PAHs were carried out. As shown in Figure 2 the H2O2 concentration accumulated in deionized water (pH=6.93) increased with an increase of O2 dosage from 2 up to 6 mg/l. The yield of H2O2 reached a value as high as 84 mg/l under sonication after 150 min in deionized H2O whereas the H2O2 production was only 9 mg/l in wastewaters containing PAHs. This H2O2 level could only remove some less hydrophobic PAHs with low yields in comparison with the total amounts of PAHs removed with a mean yield of 95% after 150 min at 60oC. Since H2O2 production is an indicator of the presence of OHl through sonication a low H2O2 concentration shows that OHl did not contribute to the sonolysis of PAHs (Figure 2).
Figure 2: Effect of DO (2.00 mg/l, 4.00 and 6.00 mg/l) on H2S production in deionized water and PCI ww containing PAHs after 150 min at 60oC (sonication power=640 W, sonication frequency=35 kHz).
Consequently, it reveals that the hydrophobic PAHs in wastewater were principally sono-degraded mainly by way of pyrolytic reaction, deduced from the cavitation, while OHl made a minor contribution to the removal of PAHs. Throughout the studies some gaseous by-products were observed in the headspace of the sonication reactor. 38.21-45.35% CO2 and 11.20-15.43% CH4 were measured after 10 min in all the hydrophobic PAHs studied in the headspace of the sonicator reactor (Figure 3). The theoretical total carbon amount of initial total PAHs for CH4(g) and CO2(g) shows that the carbon balance reaches about 12.26-18.37% and 31.11-49.60%, respectively. It can be assumed that these gases are primarily formed as soon as more hydrophobic PAHs penetrate the cavitation bubbles in order to be pyrolyzed. The presences of these products indicate that the destruction of PAHs was performed via pyrolytic mechanism.
Figure 3: Theoretical carbon percentage based on CH4 and CO2 produced throughout sonication of PAHs after 150 min at 60oC (sonication power=640 W, sonication frequency=35 kHz).
Effect of Aeration on the Removal of PAHs in PCI ww at Increasing Sonication Time and Temperature
PCI ww were aerated for 1 h with an air pump before the sonication experiments. 48.95%, 75.39% and 94.48% total PAHs yields were measured under 1 h aeration after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 30oC (data not shown) . The contribution of aeration were 3.61%, 12.99% and 4.37% to the total PAHs yields after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 30oC, compared to the control (E=45.34%, 62.40% and 90.11% total PAHs after 60 min, 120 and 150 min, respectively, at pH=7.0 and at 30oC). 54.21%, 80.16% and 97.62% total PAHs removals was obtained under 1 h aeration at 60oC after 60 min, 120 and 150 min, respectively, while the control has a yield of 96.90% after 150 min. Aeration did not increase the total PAHs yields after 60 and 120 min compared to the control at 60oC. The maximum total PAHs yield was 97.62% after 150 min under 1 h aeration at 60oC. The contribution of aeration to the total PAHs removal was not significant at 60oC (R2=0.45, F=4.56, p=0.01).
The aeration contributed only to the removals of less hydrophobic PAHs. The yields in BbF, BkF and BaP increased from 62.15-67.21% to 93.94-97.99% with 1 h aeration after 150 min at 60oC. Air in the aqueous solution was reported to play a very important role in the generation of highly oxidative OHl enhancing the decomposition of less hydrophobic PAHs . In the presence of air, reactive radicals such as O●, OH● and O2H● will be produced by a series of reactions and may participate in the decomposition reaction of the less hydrophobic PAHs. These PAHs are degraded with hydroxylation reactions since the OHl ions concentrations increased from 23x10-21 up to 8x10-7 ng/ml after 120 min under 1 h aeration. However, aeration did not contribute to the yields of more hydrophobic PAHs. The removals of IcdP, DahA and BghiP remained around 76.18-79.80% after 150 min at 60oC. This reveals that the more hydrophobic PAHs in PCI ww were principally destroyed by way of pyrolytic reaction, deduced from the cavitation inside the bubble and/or at the interfacial region [44,45]. It was reported that the contribution of aeration was not so effective on the sonication of PAHs with high benzene rings . In this study, it was found that, under pyrolytic conditions, the H2O2 formation from the reactive radicals (OH● and O2H●) decreased trough sonication of more hydrophobic IcdP, DahA and BghiP. The studies including the H2O2 generation in deionized water showed that H2O2 accumulated in deionized water (pH=2.0) and increased up to 148 mg/l with 1 h aeration after 120 min at 30oC (Table 5).
The yield of H2O2 decreased to a value as low as 5 mg/l from 178 mg/l under sonication of less hydrophobic PAHs after 150 min, whereas it could only cause the destruction of some less hydrophobic PAHs (BbF, BkF and BaP) with a yield of 67-92% in comparison with the more hydrophobic PAHs removed (IcdP, DahA and BghiP) with a yield of (47-56%) (Table 6). The H2O2 concentration did not change for more hydrophobic PAHs before and after 120 and 150 min and remained between 4-5 mg/l. These differences could be explained by the differences in their physicochemical properties and their destruction ways. For example, the hydrophobicity of a more hydrophobic PAH “BghiP” is higher with low Henry’s law constant (log POW=5.98 at 25oC) compared to a less hydrophobic PAH “BbF” with high Henry’s law constant (log POW=5.71 at 25oC) and low vapor pressure (VP, 5.00x10-7 mm Hg at 25oC) and octanol water coefficients properties. The PAHs containing high benzene rings were removed with low efficiencies compared to the low benzene ring PAHs (BkF, BaP) since their solubilities, Henry’s law constant (5.84x10-7 and 4.57x10-7 atm m3 /mol at 25oC) and vapor pressures (9.70x10-10 and 5.49x10-9 mm Hg at 25oC) are low. The low removal efficiencies in more hydrophobic PAHs (IcdP, DahA) could probably be attributed to their non-hydroxylated sonication mechanism since the OHl ions produced through aeration are not favor for their sono-degradation [47,48]. A non-hydroxylated pathway was observed with aeration in the removal of some more hydrophobic PAHs (IcdP, DahA) at low frequency . In the presence of more hydrophobic PAHs (IcdP, DahA) more aeration inside the bubble dissolves into the medium during oscillations with a consequent rise in the intensity of collapse and radical production. Higher concentration at the bubble interface raises the partial pressure of the more hydrophobic PAHs (IcdP, DahA) which leads to entrapment of these PAHs into the cavitation bubble, resulting in pyrolytic sono-decomposition during the transient collapse of the bubble . The less hydrophobic (BkF, BaP) PAHs could not be destroyed under aerated sonication process since it can not be removed in hydroxylated mediums containing OHl ions as high as 12.00x10-8 ng/ml . The more hydrophobic BghiP could not be destroyed in hydroxylated mediums containing OHl ions as high as 23.00x10-21 and 8.00x10-7 ng/ml (Table 5). The destruction way of this PAH was mainly pyrolysis .
The effect of air on the removal of short chain, less hydrophobic, PAHs could be explained as follows: the hydroperoxyl radicals (O2H●) formed by Eq. (4);
will decay with generation of H2O2 as follows (53,54) in Eqs. (5), (6) and (7).
In the presence of air, reactive radicals such as O●, OH● and O2H● will be produced by a series of reactions and may participate in the decomposition reaction of the PAHs. In aerated solutions, the O2H● formed by Eq. (4) given in the section “effect of DO on the sonication of PAHs” and will decay with the generation of H2O2. Air in aqueous solution of less hydrophobic PAHs was reported to play a very important role in the generation of highly oxidative OHl, enhancing its decomposition . The effect of air on the removal of short chain, less hydrophobic, PAHs could be explained as follows: In aerated solutions, O2H● formed by Eq. (5);
The production of the O2H● increases the oxidation process due to further formation of H2O2 by its recombination reaction [56,57]. This was attributed to increase OHl entering to the bulk solution. It has been speculated that the rate of OHl formation in the gas phase is higher in an oxygenated atmosphere. Since in this study it was studied at low frequency (35 kHz) longer collapse times for the bubbles OHl have more time to recombine before being ejected from inside the bubble into the bulk liquid than at higher frequencies . This implies that OHl reacts with less hydrophobic PAHs primarily in the bulk solution and that the number of OHl capable of reaching the bulk at 35 kHz. It is clear that the destruction yields of less hydrophobic PAHs increases significantly up to 99% with an increase of aeration from 1 h up to 2 h . This observation may be partially ascribed to the generation of OHl, resulted from dissociation of molecular O2 (in the air) in the bubble, which is likely to recombine to form H2O2 at the gas–liquid interface of the bubbles as given in Eqs. (5), (6) and (7).
As a conlusion the more hydrophobic PAHs in PCI ww were principally destructed by way of pyrolytic reaction, deduced from the cavitation. It was reported that hydropholic compounds like IcdP, DahA and BghiP were degraded directly via pyrolytic reactions occurring inside the bubble and/or at the interfacial region or indirectly via radical reactions occurring at the interface and/or in the solution bulk of aerated solutions [60,61] It was reported that the contribution of aeration was not so effective on the sonication of PAHs with high benzene rings . However, the destruction of less hyrophobic PAHs also involve the participation of OHl and possibly H atoms which are formed from the water dissociation within the bubble and migrate towards the interface and the solution bulk in oxygenated environments.
Effect of N2(g) on the PAHs Removal Efficiencies in PCI ww Versus Sonication Times and Temperatures
15 and 30 min N2(g) (3.00 and 6.00 mg/l N2) was sparged in PCI ww before sonication experiments. Maximum 92.46%, 94.16% and 88.33%, 96.27% total PAHs removal efficiencies were found under 15 and 30 min of N2(g) (3 and 6 mg/l N2) sparging at 30oC and 60oC, respectively, after 150 min at pH=7.08 (Table 6). An increase of 6.08-13.18% and 18.51-18.71% in total PAHs yields were obtained under 15 and 30 min N2(g) (3 and 6 mg/l N2) sparging after 60 and 120 min while no significant increase in PAHs removals were observed after 150 min at pH=7.0, compared to the control [without N2(g) E=90.11% for total PAHs after 150 min at pH=7.0 and at 30oC]. The contribution of 15 and 30 min N2(g) (3 and 6 mg/l N2) sparging on the PAHs removals are significant at low sonication times (60 and 120 min) and temperature (30oC) (R2=0.84, F=13.09, p=0.001). N2(g) sparging did not significantly affect the PAHs removals compared to the control (E=96.90% total PAHs at pH=7.0) at 60oC at all sonication times (R2=0.32, F=1.34, p=0.001).
15 and 30 min N2(g) (3 and 6 mg/l N2) sparging did not contribute to the PAHs removals at 60 and 120 min at 60oC with the exception of 15 and 30 min N2(g) (3.00 and 6.00 mg/l N2) sparging after 120 min at 30oC. This could be attributed to the benzene ring-opening reactions of PAHs and other intermediates as well as the degradation of by-products throughout N2(g) sparging at low temperature.
The individual PAHs removals increased from 58.48-59.91% up to 71.79-96.91% after sonication time as long as 150 min. A significant correlation between PAHs removals and long sonication time was observed when 30 min N2(g) (6.00 mg/l N2) is sparged at 60oC (R2=0.86, F=17.98, p=0.001). In this study, it was found that N2(g) sparging increased the yields of less hydrophobic PAHs. 89.04% FL, 92.39% BaA and 96.91% BbF maximum PAHs yields were obtained after 150 min at 60oC for less hydrophobic PAHs with four benzene rings when 30 min N2(g) (6 mg/l N2) was sparged to the PCI ww. The yields obtained for more hydrophobic PAHs namely BaP, IcdP, DahA and BghiP were significantly lower (68.24-78.49%) than that of less hydrophobic PAHs after 150 min at 60oC. The PAHs containing high benzene rings were removed with low efficiencies compared to the low benzene ring PAHs since their solubilities, Henry’s law constant and vapor pressures are low.
In the presence of more hydrophobic PAHs more N2(g) inside the bubble dissolves into the medium during oscillations with a consequent rise in the intensity of collapse and radical production. Higher concentration at the bubble interface raises the partial pressure of the more hydrophobic PAHs which leads to entrapment of these PAHs into the cavitation bubble, resulting in pyrolytic sono-decomposition during the transient collapse of the bubble [63,64].
A non-hydroxylated pathway was observed in a work) with N2(g) sparging in the removal of some more hydrophobic PAHs at low frequency . The low removal efficiencies in more hydrophobic PAHs could probably be attributed to their non-hydroxylated sonication mechanism since the OHl ions produced through N2(g) sparging are not favor for their sono-degradation [65,66].
H2O2 which is preferentially formed by the recombination of OHl issue from the sonolysis of H2O can be used as a good indicator of the OHl production. The H2O2 measurement during acoustic cavitation, in absence and in the presence of PCI ww, is a suitable method to estimate the radical production rate for specific sonochemical conditions. The initial rate of H2O2 formation associated to the PAHs treatment by sonication in PCI ww decreases with increasing sonication time at 60oC. In the absence of PCI ww (in deionized water) the H2O2 was accumulated and its concentration was measured as 187 mg/l, whereas this level was only 9 mg/l in PCI ww after 150 min in samples containing N2 (Table 6). The OHl ion concentrations also increased from 10x10-62 to 43x10-7 mg/l after 150 min in PCI ww containing less hydrophobic FL, BaA and BbF. This showed that hydroxylation is the main mechanism for the removal of less hydrophobic PAHs. In other words, OHl is the major process for complete degradation of less hydrophobic PAHs.
In PCI ww the most sonogenerated OHl reacted with 89-97% less hydrophobic PAHs removals and radical recombination to produce H2O2 (Gogate et al., 2004). The OHl ion concentrations remained constant around 10x10-62 mg/l after 150 min in PCI ww containing more hydrophobic BaP, IcdP, DahA and BghiP. No increase in OHl ion concentrations was observed throughout sonication of more hydrophobic PAHs. This showed that hydroxylation is not the main mechanism for the removal of more hydrophobic PAHs. This indicates that the main process for the destruction of more hydrophobic PAHs is pyrolysis.
Since the sonooxidation of more hydrophobic BaP, IcdP, DahA and BghiP comprised 0.11%, 0.10%, 0.09% and 0.08% of the total sonodegradation process, OHl is not the major process for complete degradation of these PAHs (Table 7).
In other words, in this study, the contribution of OHl is minor for the ultimate sonodegradation of more hydrophobic PAHs. The formation of by-products (hydroxylated compounds namely phenanthrenediols) for possible OHl oxidation was not observed in HPLC. Similar results were obtained in some studies [66,67]. Since the sonooxidation of less hydrophobic FL, BaA and BbF comprised 0.87% and 0.77% and 79.00% of the total sonodegradation process, OHl is the major process for complete degradation of these PAHs. The contribution of the pyrolysis to the destruction of these PAHs is not significant.
Different suggestions were reported on the effect of N2(g) sparging on PAHs removals: Dissolved N2 present in aqueous solution might scavenge the free radical attacks to PAHs . Some studies showed that the PAHs are degraded in a N2(g) sparged system with high concentrations of OHl scavenger through sonication . Sparging of N2(g) could change the temperature within the cavitation site or other properties of the cavitation process. Gasses with lower thermal conductivity values such as N2 (18.70 mW/mol.K) increase the temperature inside the cavitation bubble upon collapse because they will allow less heat to the surrounding . The ratio of the specific heat at constant pressure and constant volume (Cp/Cv) plays a role in determining the maximum size of the cavitation bubble [69,70]. Since the Cp/Cv ratio of N2(g) is high (1.543) the bubble diameter increase during cavitation process resulting in high OHl ions (69,70). In this study it was found that N2(g) increase the OHl ions through sonication of less hydrophobic PAHs by increasing the temperature and bubble diameter.
The benefit of N2(g) sparging in enhancing the sonochemical activity is reported [71,72]. Sparging of N2(g) changed the temperature within the cavitation site and the variations of chemical properties of the system resulting in high PAHs removals (Laughrey et al., 2001). The mechanism affecting the yield of PAHs removals through sonication was explained as follows: N2(g) sparging changed the temperature within the cavitation site and the variations of chemical properties of the system resulting in high PAHs removals at 30oC. Gases with lower thermal conductivity (18.40 mW/mol.K) values such as N2(g) increased the temperature up to 49oC inside the cavitation bubble upon collapse because they allowed less heat to the surroundings such as 37oC [70,71]. The ultrasonic degradation of PAHs in aqueous solution is strongly affected by the average specific heat ratio of the gas dissolved in the solution, since the degradation is considered to proceed in local hot-spots formed by adiabatic collapsing bubbles . The ratio of the specific heat at constant pressure and constant volume (Cp/Cv) plays a role in determining the maximum size of the cavitation bubble . The specific heat (Cp) of N2(g) is 1.0397 J /g.K . The (Cp/Cv) ratios of N2(g) is 1.543. It is unlikely that any significant change in the cavitation process occured due to sparging with N2(g). The N2(g) or its sonolysis product may scavenge a fraction of the reactive radicals (Mendez-Arriaga et al., 2008). The greater the ratio, the greater the radius of the bubble before collapse. Significant changes in the cavitation process, due to changes in the chemical properties of PAHs occurred due to sparging with N2(g). The PAHs degradation in N2(g) sparged systems suggests that a non-oxygen dependent, non-hydroxyl radical pathway also exists more hydrophobic PAHs. This pathway is most likely pyrolysis occurring in the gas phase .
It was found that N2(g) bubbles produce the highest number of OHl in the sonication of less hydrophobic PAHs. Since the intensity of collapse of N2(g) bubbles indicated by the temperature peaks attained at the collapse of these bubbles . Although, some researchers mentioned that the trend of N2(g) in production of OHl is low through hydroxylation reaction [50,51] in a medium containing more hydrophobic PAHs the dissolved N2(g) slowly diffuses into the cavitation bubble during oscillations, as result of which the equilibrium of the bubble grows. This gas cushions the transient collapse of the bubble, the temperature and pressure peaks attained in the bubble reduce, resulting in the reduction of OHl production. This affects negatively the degradation of PAHs occurring through hydroxylation. In the presence of more hydrophobic PAHs more N2(g) inside the bubble dissolves into the medium during oscillations with consequent rise in the intensity of collapse and radical production. Higher concentration at the bubble interface raises the partial pressure of the pollutant that leads to higher evaporation and subsequent entrapment of the pollutant into the cavitation bubble, which under goes pyrolytic decomposition at the extreme conditions reached during the transient collapse of the bubble [73-75]. Sivasankar & Moholkar (2009) mentioned that the presence of N2(g) has a negative effect on the formation of H2O2 and on the degradation of PAHs during sonication under air. The HNO2 formed may be the scavenger of OHl, which eventually leads to the suppression in H2O2 production.
In this study, high total PAHs removals were found with N2(g) sparging in comparison with the other literature data. For example, in Laughrey et al. (2001) 80% total PAHs removals were found at 20oC under 30 min of N2(g) (6.00 mg/l N2) sparging at 20 kHz, and at 70 W, after 120 min. The PAHs yields reached higher than that found by Laughrey et al. (2001) under similar operational and sonicational conditions. 87% total PAHs removal was found for initial 0.90 µg/l total PAHs concentration at 400 W and at 24 kHz under 25 min N2(g) (5 mg/l N2) sparging at 60oC after 150 min. Different suggestions were reported on the effect of N2(g) sparging on PAH removals: Dissolved N2 present in aqueous solution might scavenge the free radicals, inhibiting the free radical attack to PAHs . These results agree with the data obtained for less hydrophobic PAHs. On the other hand, a non-hydroxylated pathway was observed with N2(g) sparging resulting in a non-complete removal of NAP, ACL and PHE at low frequency. The low removal efficiencies in more hydrophobic PAHs could be probably attributed to their non-hydroxylated sonication mechanism .
PAH removal kinetic
Some of the recent studies showed that there is a highly significant relationship between the average removal percentages and the hydrophobicity of PAHs, indicated as the octanol water partition coefficient is shown [55,56]. It becomes evident that a larger hydrophobicity resulted in smaller removal of the PAHs. However, in this study, the 4-ring, 5 and 6- ring PAHs, in contrast to the relationship mentioned above, exhibiting higher removals than expected from their log Kow at 60oC after 150 min. No significant correlation was observed between PAHs yields, water solubility, vapor pressure, C number in benzene rings, Henry’s law constant through sonication assisted 20 mg/l Fe+2 at 60oC after 150 min (R2=0.43, F=3.56, p=0.001). On the other hand, it was found that the coefficient of the correlation between the PAHs yields and the residual concentrations was strong and significant (R2=0.83, F=13.16, p < 0.001) at 60oC after 150 min (Figure 4). Similar PAHs removal in PAHs with low and high benzene rings could be attributed to the remaining PAHs percentages varying between 0.04% and 10% after 150 min since the hydrophobic PAHs can easily diffuse near cavitation bubbles under pyrolytic destruction.
Figure 4: The remaining (untreated) PAHs percentages versus sonication time in pseudo first order reaction kinetic (sonication power=640 W, sonication frequency=35 kHz).
A possible explanation of the postive effect of Fe+2 on the sonication of PAHs could be the reaction of Fe+2 with H2O2 to form OHl at a 34 kHz and a power of 450 W after 125 min at 60oC [70,71]. As this reaction proceeds, the concentration of Fe+2 declines, and consequently the rate of H2O2 consumption and OHl formation decline. The loss of Fe+2 is eventually balanced by the formation of Fe+2 through reduction of Fe+3 by reaction with H2O2 or O2Hl, and a steady state Fe+2 concentration is reached. At this point (> 60 s), pseudo first order loss of H2O2 is observed. This explanation is also supported by evidence that the OHl formation rate is significantly higher in the first 60 s. Psillakis et al. (2004) studied the sono-removal of 150 μg/l total initial concentration of PAHs mixture (NAP, ACT, PHE) in an aqueous solution. 92.20% of NAP, 96.25% of ACT and 89.80% of PHE removal efficiencies were found with Fe+2=14 mg/l concentration in a sonicator with a power of 150 W, a frequency of 80 kHz, a sonication temperature of 20oC, after 150 min. In our study the removal efficiencies for the aforementioned PAHs were found to be higher (E, NAP=99.00%, E, ACT=98.25% and E, PHE=98.11%) at 35oC for the same sonication time. During aqueous ultrasonic irradiation, OHl formed during the thermolytic reactions of H2O recombine to form H2O2 that tends to accumulate in the solution and does not usually play an important role in oxidizing organic species. However, the reaction between H2O2 and Fe+2 is known to produce OHl and is commonly referred to as the Fenton process .
Effect of HCO3-1 Concentrations on the Removal of PAHs in PCI ww at Increasing Sonication Times and Temperatures
Increasing HCO3-1 (0.50 g/l, 1 and 5 g/l) concentrations were added to the PCI ww before sonication process. 80.43%, 92.14% and 81.66% total PAHs yields were found in 0.50 g/l, 1 and 5 g/l HCO3-1, respectively, after 150 min at pH=7.0 and at 30oC. Total PAHs yields were sligthly increased in 1 g/l HCO3-1 (2%) compared to the control (without HCO3-1 while E=90.11% total PAHs) at pH=7.0 and at 30oC after 150 min (Figure 5). A significant linear correlation between total PAHs yields and increasing sonication time was not observed (R2=0.32, F=0.30, p=0.01). 78.09%, 81.69% and 82.62% total PAHs removals were found in 0.50 g/l, 1 and 5 g/l HCO3-1, after 150 min at pH=7.0 and at 60oC. Total PAHs yields did not change after 150 min compared to the control (E=96.90% total PAHs) at pH=7.0 and at 60oC. A significant linear correlation between total PAHs yields and increasing HCO3-1 concentrations was not observed (R2=0.34, F=0.40, p=0.01).
In order to identify the contribution of OHl in the sonolytic degradation of PAHs, the role of OHl was examined in the absence and presence of HCO3-1. Therefore, one refractory non-volatile, more hydrophobic PAHs namely DahA (4.57 ng/ml) was taken into consideration among the other seventeen PAHs present in PCI ww. Figure 5 presents the percent reduction of DahA during these experiments. As shown in this figure, the degradation curves can be mainly divided into three regions (initiation, acceleration and stabilization steps), but the most degradation of DahA was observed in two regions (initiation and acceleration steps). The degradation curves given in Figure 5 can be estimated with the sigmoidal model based on the best fits by trial and error using Eq. (8):
where, y is the degradation percent of PAHs and yo is the intercept, xo is the time at the inflection point (the time dividing between the first two stages), and a and b are dependent variables. This sigmoidal model Eq. (7) was used to divide the curves in Figure 5 mainly into two stages, the initiation and the acceleration step . By fitting the data into Eq. (7), R2 values (the conformity to the sigmoidal model) of each degradation curve in the runs with HCO3-1 and without HCO3-1 at 0 and 60 min were obtained, and found to be R12=0.98 and R22=0.97, respectively. The R2 values obtained also provided the estimation of reaction time in each step, and this estimated reaction time in each step is shown.
Figure 5: Percent reduction of DahA with HCO3-1 trough sonication times (sonication power=640 W, sonication frequency=35 kHz).
As shown in Figure 8, the total percent degradation values of DahA during the initiation steps were not significantly different (ANOVA, F=13.98, p=0.001) between these two cases. However, significant differences (ANOVA, F=2.67, p=0.001) were observed between the percent degradation values at the acceleration steps. Similar statistical results were obtained between the trials with HCO3-1 and without HCO3-1 after 60 min (ANOVA, F=1.98, p=0.001 and F=2.80, p=0.001) at the initiation steps. HCO3-1 is a well known OHl scavenger as shown in Eq. (9):
Since HCO3-1 is an ion, it will scavenge free radicals predominantly in the bulk water phase. Water-soluble compounds that are non volatile will be signicantly affected by the HCO3 concentration [60,61]. Since DahA is non-volatile, a refractory compound, the HCO3 can complete with DahA for available free radicals in the interfacial and bulk region and decrease the sonical destruction efficiency of the PAHs in question. Figure 8 also shows the contribution of HCO3-1 in the role of OHl scavenger during sonolysis. The effect of radical scavenging by HCO3-1 mainly occurred during the acceleration step, and suggesting that the most OHl in the sonication is generated at the acceleration step. From the results in Figure 8 during the sonication it can be induced that the reaction between OHl and HCO3-1 mainly occurs at the acceleration step rather than at the initial step. As aforementioned OHl are mainly produced at the acceleration step, which react with HCO3-1 to produce OHl ion, resulting in the increase in pH of the solution. However, at the initiation step, the production of OHl is insufficient for the reaction with HCO3-1, as shown in Eq. (9). Instead, the presence of 10 mg/l HCO3-1 may act as a buffer in the solution, resulting in almost no change of pH. The sonication of DahA at the initiation step may proceed by a thermal reaction, while the degradation of DahA is dominated by OHl reaction at the acceleration step. As shown in Figure 8, thermal degradation and chemical oxidation contributed approximately 25% and 34% to the degradation of 1,4–dioxane according to the result of the sonication without HCO3-1 .
The sonic degradation of DahA in the PCI ww was found to be varied between zero and pseudo first order with respect to PAHs concentration in the initiation and acceleration steps with HCO3-1 and without HCO3-1 at a frequency of 35 kHz and at 60oC in Eqs. (10) and (11):
Linear regression analysis was performed in order to determine the conformity to zero order or the pseudo-first order model in each step, based on Eq. (11) . R12 and R22 stand for the linear regression coefficients for zero order and pseudo-first order reaction kinetic model in each step, respectively. However, the kinetic constants found for DahA removal in the presence of HCO3-1 after 60 min exhibited a zero order kinetic rate model for both initiation (R12=0.97) and acceleration (R22=0.91) steps. Under these conditions, a maximum DahA removal efficiency was achieved.
The results of this part of this study showed that the degradation trend of the DahA at the acceleration step was also not significantly different between two conditions with and without HCO3-1 at the beginning of the sonication (0 min). However, the rapid decrease of DahA degradation was observed at the acceleration step in the presence of HCO3-1 after 60 min. The initiation step of the DahA degradation followed the zero order kinetic rate models, while the acceleration step followed the pseudo-first order. In the presence of HCO3-1 as a radical scavenger, the degradations of DahA were suppressed, indicating that OHl is an important factor in the sonication, especially at the acceleration step. Since DahA is hydrophobic, HCO3-1 can compete with DahA for available free radicals in the interfacial and bulk region and decrease its decomposition efficiency. They reported that water soluble compounds that are less volatile will be significantly influenced by the HCO3-1 concentration . Some reserarchers found that the degradation rate of Acid Blue 40 was increased in the presence of CO3 and HCO3-1, with a maximum around 610 mg/l total concentration at pH=9.0 and a saturation trend at pH=11.0 . They mentioned that in the presence of HCO3-1 not all the sonochemically formed OHl can induce the substrate degradation since with an elevated OH at the air–water interface, radical–radical recombination to yield H2O2 and H2O would be much more important than the reaction with the substrates. Furthermore, thus would also limit the diffusion of reactive radical species to the solution bulk in the sonicator with a frequency of 29 kHz and a power of 450 W.
Effect of S2O8-2 Concentrations on the Removal of PAHs in PCI ww in the Presence of iso-Butanol (C4H9OH)
To verify the PAHs sonodegradation pathways, the effects of the radical scavenger, iso-butanol, on the removals of PAHs in a sonicator containing S2O8-2 at pH=7.0 were examined. The addition of 200, 400 and 600 mg/l iso-Butanol to the sonicator containing Na2S2O8 reduced the the PAHs yields from 98% to 47% (Figure 8). As the iso-butanol concentrations increased the PAHs yields decreased. Iso-butanol molecules pass in the cavitation bubbles, they are able to scavenge OHl in the bubble and reduce significantly the degradation rate of the PAHs. Alcohols are commonly used to quench the OHl. The inhibition was explained by OHl competitive reactions with PAHs and iso-butanol. In other words, the sono-degradation of PAHs was found to deccelerate by iso-butanol where the competition for OHl increases as the quantity of iso-butanol relative to that of all individual more and less hydrophobic PAHs are increased. The inhibition in PAHs removals by adding iso-butanol indicates that the PAHs destruction pathway involves slightly OHl in the bulk liquid and/or the interface region of the cavitation bubbles for more (DahA, BghiP, BaA, CHR and BbF) and less (BkF, BaP, IcdP) hydrophobic PAHs. Since the PAHs utilized here are less soluble in H2O, and thus hydrophobic, their partitioning into the gas phase are possible and the solutes were pyrolyzed in the cavitation bubbles. Therefore, pyrolysis could be a major reaction pathway in a sonicator containing Na2S2O8 through sonication of hydrophobic PAHs . Figure 6 shows the effect of 2, 4, 6 and 10 mg/l S2O8-2 as an oxidant on the degradation of seven PAHs, compared with sonication-only condition (without S2O8-2). The degradation efficiency of seven PAHs in the sonication with 6 mg/l S2O8-2 was enhanced by 36% after 150 min (Figure 6).
Figure 6: Effect of increasing S2O8-2 on PAHs removal efficiencies at 60oC after 150 min (sonication power=640 W, sonication frequency=35 kHz).
In other words, the maximum yields in all PAHs (E=97-99%) was found at 6.00 mg/l S2O8-2. As the S2O8-2 increased from 6 to 10 mg/l the PAHs removals decreased from 98-99% to 72% and 75%. The kinetics at the initiation step in the sonication with S2O8-2 were found to be of the pseudo-first order rate model (k=0.008 1/day). We assumed that the OHl is also produced by the reactions between sonication of PAHs and S2O8-2, as shown in Eqs. (12), (13) and (14):
As shown in Figure 6, the degradation percentage of seven PAHs in the absence of S2O8-2 only was 78.12% after 150 min. This indicates that BghiP can be directly degraded only with sonication as shown in Eq. (14) . The degradation efficiencies of PAHs were not statistically different between those with S2O8-2 and those without S2O8-2 (only sonication) (R2=0.87, F=6.37, p=0.001). Persulfate ions trap the electrons generated through sonication. Thus, they prevent the recombination of S2O8-2 with positive holes, resulting in the simultaneous generation of sulfate free radicals (SO4l) (Eq. (15). Then the SO4l react with H2O molecules to form OHl (Eq. (16) . The mechanism of thermal decomposition is believed to involve the SO4l, which can abstract H2 from H2O to yield OHl  in Eqs. (16) and (17).
The effects of temperature were evaluated in reactors with S2O8-2 and without S2O8-2 at an ambient 21oC, 30, and 60oC. The experimental results indicated that the all the PAHs removals were highest in the reactor containing 9.27 mg/l S2O8-2 at 60oC, after 150 min. It can be concluded that the PAHs yields in a sonicator with S2O8-2 were higher than those in a sonicator without S2O8-2 (Figure 8). It was stated that the first order kinetics constants of the sonochemical decomposition of persulfate both increased with temperature .
When the sonolysis of H2O occurs, it leads to the formation of the non-specific oxidative species OHl. The Ultrasonic degradation of less hydrophobic organic compounds in the aqueous phase is mainly the result of an oxidation process by means of OHl which escapes from the bubble to the bulk solution. Hydrophobic compounds present in water are expected to penetrate the surrounding shell and/or the hot heart of the cavitation bubble in order to be pyrolyzed, burnt and/or ionized into the plasma core . Some studies suggested that PAHs (for example, PHE) are degraded by a free radical mechanism with OHl, since hydroxylated compounds like phenanthrenediols have not been detected [80-101]. On the other hand, some recent research has shown that plasma allows the ionization of organics like PAHs .
A comparison of calculated oxidation rates of OHl) for the experimental sonodegradation rates is at a frequency of 35 kHz for seven PAHs. Since the sonooxidation of hydrophobic DahA and BghiP comprised 0.04% and 0.03% of the total sonodegradation process, OHl is not the major process for complete degradation of these PAHs. In other words, in this study, the contribution of OHl is minor for the ultimate sonodegradation of more hydrophobic PAHs. This indicates the occupation of the heart and/or the surrounding shell of the bubble by hydrophobic PAH molecules, and limitation of the sonolysis of H2O and the formation of OHl. This demonstrates that the main process for destruction of hydrophobic PAHs is pyrolysis.
The formation of by-products (hydroxylated compounds namely phenanthrenediols) for possible OHl oxidation was not observed in HPLC. Similar results were obtained (76). Since the sonooxidation of less hydrophobic PHE, PY, CHR and ANT comprised 0.78%, 0.74% and 0.72%, 0.79% of the total sonodegradation process, OHl is the major process for complete degradation of these PAHs. The contribution of the pyrolysis to the destruction of these PAHs is not significant.
In order to determine the pyrolitic mechanism of PAH degradation a synthetic saturated solution of a mixture of seven PAHs (1.56 mg/l) was prepared to detect the gaseous by-products in the headspace of the sonication reactor (Figure 7). The theoretical total carbon amount of initial total PAHs for CH4 and CO2, shows that the carbon balance reaches about 11.13-15.28% and 38.40-45.26%, respectively (Figure 7).
Figure 7: Theoretical carbon percentage based on CH4(g) and CO2(g) produced throughout sonication of PAHs (sonication power=640 W, sonication frequency=35 kHz).
As mentioned previously the addition of 200 mg/l 1-butanol to the PAH mixture leads to a partial inhibition of the PAHs destruction and this is in agreement with the studies [68,69]. About 15.21-35.46% inhibition was observed for less hydrophobic PAHs with the highest vapour pressure (PHE, ANT, CHR, PY and BbF) while all the other strong hydrophobic PAHs (DahA and BghiP) were not or very slightly affected by the OHl scavenger. The oxidation of PAHs with OHl appears to be of relatively minor importance, because of the low inhibition of the degradation by means of 1-butanol and since the steady-state OHl concentration in the interfacial region of the cavitation bubble, where PAHs accumulate, is lower and not sufficient for a complete degradation of PAHs. Because of their properties, PAHs are then expected to be mainly localized in the heart and/or in the surrounding shell of the bubble, inhibiting the production of OHl and hence the oxidation pathway.
Throughout the studies some gaseous by-products were observed in the headspace of the sonication reactor. 38.20-45.32% CO2 and 11.10-15.03% CH4 were measured after 10 min of sonication time in all the PAHs studied in the headspace of the sonicator reactor (Figure 8). The theoretical total carbon amount of initial total PAHs for CH4(g) and CO2(g), shows that the carbon balance reaches about 12.33-18.12% and 31.40-49.09%, respectively. It can be assumed that these gases are primarily formed as soon as PAHs penetrate the cavitation bubbles in order to be pyrolyzed.
Figure 8: Theoretical carbon percentage based on CH4(g) and CO2(g) produced throughout sonication of PAHs (sonication power=640 W, sonication frequency=35 kHz).
The results of this study showed that sonodegradation is a very useful process in the removal of toxic and refractory compounds in petrochemical industry wastewaters. Low frequency (35 kHz) sonication proved to be a viable tool for the effective degradation of refractory compounds in petrochemical wastewaters. The removals increased after 120 and 150 min at 30oC and at 60oC. The sonication process could prove to be less land-intensive, less expensive and require less maintenance than traditional biological treatment processes and other AOPs processes. Sonication technology can provide a cost-effective alternative for destroying and detoxifying refractory compounds in petrochemical wastewaters. Sonicationn process can be easily applied all kind of wastewater type with only sonication and with the addition of some additives. Sonication removal efficiencies can be increased with the addition of some chemicals (solids or gas bubbles to act as nuclei). The optimization of sonication time can be made before the sonication experiments for effectively acoustic cavitation reaction. The optimization of sonication volume can be fixed before the sonication process for effectively cavitation reaction.
As a result of this study, sonication process is recommended for the treatment of petrochemical wastewaters. Sonication process can be applied as a pre-treatment or post-treatment in combination with other water purification processes.