Simultaneous Removal of MTBE and Benzene from Contaminated Groundwater Using Ultraviolet-Based Ozone and Hydrogen Peroxide

Efficiency of ultraviolet-ozone (UV/O 3 ) and ultraviolet-hydrogen peroxide (UV/H 2 O 2 ) processes was investigated for simultaneous removal of methyl tertiary butyl ether (MTBE) and benzene from contaminated ground water.The photoreactor employed housed 15-watt low pressure (LP) and 150-watt medium pressure (MP) mercury UV lamps. Oxidation of contaminants was studied at two different levels of ozone and hydrogen peroxide. Brackish groundwater samples were spiked with MTBE and benzene up to a concentration of 500μg L. Removal potential was evaluated at different parameters such as UV type and intensity and peroxide and ozone dosages, as well as contact time. Results indicated that no removal of the contaminants was attained when treated with hydrogen peroxide or ozone alone. However, about 50% and 30% removal of MTBE were achieved in 30 minutes when irradiated with MP-UV and LP-UV lamps, respectively. On the other hand, UV/H 2 O 2 process was found to be superior in removal of MTBE (90% in 10min.) and benzene (95% in 5min.) compared to UV/O 3 process. Furthermore, removal of benzene was comparatively easier than MTBE in both approaches. It is hence concluded that higher UV intensities and elevated doses of H 2 O 2 accelerate simultaneous removal of MTBE and benzene from water.


Introduction
In oil producing countries, contamination of water resources is generally attributed to fuel additives and different fuel products.These contaminants reach water bodies through different routes such as the process of transportation, mixing, refining, and filling operations.Fuel products such as benzene, toluene ethyl benzene, and xylenes (BTEX) and fuel additives such as methyl tert-butyl ether (MTBE) are among the organic pollutants commonly detected in water resources.MTBE is added to fuel to enhance the octane number and reduce air emissions of carbon monoxide since the issuance of US Clean Air Act (USCAA) in 1970s [1].Major causes of MTBE pollution are leakages in underground storage tanks and pipelines and accidental spills during refining processes.Many studies have reported their presence and impacts in atmosphere [2] as well as in groundwater [3].Toxicity analysis conducted on test animals has shown that direct exposure to MTBE, even for shorter periods of time, can cause nausea, vomiting, and sleepiness [3].
On the other hand, benzene has received a lot of attention due to its carcinogenic nature [4].It has been reported that, in drinking water, more than 0.005 mg/L of benzene can seriously increase the risk of cancer [5].Benzene concentration in gasoline is being regulated at 1% by volume that renders it harmful carcinogen for living organism.Studies conducted on animal models show that benzene toxicity can increase the chances of cancer mutagenesis at p53 gene which results into uncontrolled growth of cells [6].
Despite the fact that both MTBE and benzene may exist together, they have different physical and chemical characteristics which affect their transportation and fate in the environment.For example, solubility of MTBE in water bodies is 20 times higher (43,000-54,300 mg L −1 ) than benzene (1780 mg L −1 ) and other compounds in BTEX [7].This enhances the movement of MTBE in underground formations and renders it difficult to trace.Furthermore, lower  oc and  ow values of MTBE indicate its lesser ability for adsorption to solid particles than benzene [8].

International Journal of Photoenergy
A number of techniques have been employed for the removal of MTBE and benzene from contaminated water.Bioremediation has not been very successful due to low biomass production and slow growth kinetics of bacteria [9].Furthermore, air stripping has been reported by some researchers as a noneconomical technique for MTBE removal from water [9] and adsorption to granular activated carbon (GAC) or synthetic resins only leads to transfer of MTBE from one phase to another instead of remediation.However, superior results in removal of MTBE have been reported by employing advanced oxidation processes (AOPs) [10,11].Commonly used AOPs include the use of ozone-hydrogen peroxide oxidation (O 3 /H 2 O 2 ), UV-hydrogen peroxide oxidation UV/H 2 O 2 , ultrasonic ozonation (O 3 ), Fenton and photo-Fenton reactions, and photocatalysis with titanium dioxide (TiO 2 ) [9,12,13].These processes generate highly reactive hydroxyl radicals (OH • ) that in turn react with the contaminants and degrade them.Under optimum conditions, these hydroxyl radicals can completely mineralize the organic pollutants to their basic constituents, carbon dioxide and water [9].Systems coupled with O 3 /H 2 O 2 possess great potential to degrade high amount of MTBE from water than O 3 and H 2 O 2 .However, the only disadvantage found in the system is the capability for bromate formation through adjustment of H 2 O 2 /O 3 ratio [9].Benzene has also been removed by using modern techniques such as UV/H 2 O 2 [14] and photocatalytic process of UV/TiO 2 [15].Likewise, other studies have also reported the photocatalytic degradation of gaseous benzene over UV-irradiated TiO 2 nanoballs [16,17].A system possesses higher efficiency of degradation than O 3 or UV alone but requires high cost and energy sometimes [9].
However, there is lack of convincing data for simultaneous removal of both benzene and MTBE using AOPs, despite the fact that both contaminants have quite often been found together in industrial wastes [18].Therefore, the main objective of this study was to evaluate the efficacy of AOPs in simultaneous removal of benzene and MTBE from contaminated water.

Solutions.
A control solution was prepared from deionized water.Highly pure (>99.5%)MTBE and benzene were obtained from Sigma-Aldrich (UK).Stock standard solutions of 100 ppm of MTBE and benzene were prepared in deionized water and spiked to initial concentration of about 500 ppb.
A test solution was prepared from groundwater samples that were collected from a well located in Dhahran, Saudi Arabia.The well has a total depth of 180 m and taps Umm er Radhuma (UER) aquifer.The well, being pumped at an average rate of about 1000 gallons min −1 , is mainly used for domestic cleansing and toilet flushing purposes.Chemical analysis of groundwater was performed to determine levels of dissolved salts (Table 1).Both MTBE and benzene were not detected in the raw water samples and hence were spiked at concentrations of 500 ppb.
Hydrogen peroxide (H 2 O 2 ) of approximately 35% purity, purchased from local market, was used to prepare the stock Ozone oxidation was facilitated by ozone generator model C-Lasky C-L1010 (Air Tree Ozone Technology Co., Taiwan).The unit can produce O 3 at the rate of 10 g h −1 if pure oxygen is used while at 2 g h −1 if air is used at flow rate of 6 L min −1 (25 ∘ C).Stock solution of O 3 in water (30 ppm) was prepared on daily basis.The amount of ozone dissolved in the aqueous phase was determined by indigo method by Hoigne and Bader [19].Ozone was spiked into previously MTBEbenzene spiked waters at concentrations of 1 and 5 ppm.

UV Photoreactor.
Spiked water samples (500 mL) with MTBE and benzene were transferred to the bench-scale UV photoreactor (Figure 1) (NORMAG Co., Hessenstraße, Hofheim, Germany).Two types of ultraviolet lamps were housed in the photoreactor: (i) 15 W low pressure (LP) Hg lamp which emits light at 254 nm with intensity of 6.5 × 10 −3 W/cm 2 and (ii) 150 W medium pressure (MP) Hg lamp which emits light at a wider spectrum ranging from 200 to 400 nm with intensity of 5.3 × 10 −2 W cm −2 .Due to volatile nature of MTBE and benzene, the temperature of solutions was kept constant at room temperature during treatment.In all treatments, water samples were collected after runtimes of 0, 5, 10, 20, and 30 min and analyzed for residual amount of MTBE, benzene, and by-products.During the rigorous bench-scale testing of the process, operational factors such as concentration of H 2 O 2 and O 3 and contact time were optimized for the removal of MTBE and benzene.

Chemical Analyses of Samples.
Effluent samples from the UV photoreactor were collected periodically and analyzed for its concentration levels of MTBE and benzene using an ISQ single quadrupole GC/MS system (Thermo Scientific) equipped with TriPlus for headspace injection system.A 60 m long, 0.32 mm i.d.Rtx-502.2 (Restek Corp., USA) capillary column was used for the separation purpose.The carrier gas was helium flowing at the rate of 1.7 mL min −1 .The column temperature was programmed to rise from 50 to 220 ∘ C at the rate of 20 ∘ C min −1 .The mass spectrometer was operated in the selected ion mode (SIM).Calibration curves were prepared for MTBE and benzene and some by-products.

Calculation of EEO Values for UV/H
2 O 2 and UV/O 3 .Electrical energy per order (EEO) of contaminant provides a convenient way to assess the effectiveness of removal of various organic compounds for a given water sample using UV irradiation.It is estimated in terms of kWh/1000 gal/order of removal, which is determined at the optimum where  is the power input (kW),  is the irradiation time (min),  is the volume (L) of the water in the reactor, and   and   are the initial and final pollutant concentrations.
For one order of removal, concentration should be reduced to   /  = 10 and log (  /  ) would be equal to 1 [21].1, the chemical analyses of groundwater revealed its brackish nature with a total of dissolved solids of over 3600 ppm.Moreover, levels of chloride, bromide, and sulfate ions are on the higher side at 1220, 4.2, and 546 ppm, respectively.High levels of these anions were expected to compete with the target pollutants (MTBE and benzene) for the hydroxyl radicals [22][23][24][25].In addition, the effect of water quality on the UV exposure and irradiation efficiency is low due to the low turbidity level of 0.1 NTU.The scavenging activities of the anions especially chlorides, carbonates, bicarbonate, and other species has also been reported by other researchers [23,25].Classical Fenton's system also supports this behavior of anionic species in terms of nonproductive reactions [23,26,27].Figures 2 and 3 show that both H 2 O 2 and O 3 were not able to influence the removal of MTBE or benzene from contaminated groundwater under the conditions used in the study.This can be attributed to the strong bonding within molecular structure of MTBE and benzene [28,29].The second reason could be the relatively low oxidative potential of H 2 O 2 and O 3 [30].However, significant removal of MTBE was observed when the spiked water was treated by UV radiation.Approximately, 50% of MTBE was removed within 30 minutes of exposure to 150 W MP UV lamp and about 30% in samples exposed to 15 W LP UV lamp.These findings support the results of previously conducted studies that reported partial degradation of MTBE by UV photolysis [31].

Results and Discussion
Similarly, removal of benzene was substantial in samples exposed to UV photolysis (Figure 3).Approximately 75% of benzene was removed by 15 W LP-UV lamp within 30 minutes and 95% benzene was removed by 150 W MP-UV irradiation.This accelerated removal of benzene was apparently due to the fact that benzene is slightly less stable against UV radiation than MTBE [32].

Treatment of MTBE-Benzene Spiked Waters by UV/H 2 O 2
Process.Results of remediation of MTBE-benzene spiked waters by UV/H 2 O 2 process are presented in Figures 4 and  5. Removal of the MTBE by UV/H 2 O 2 was significantly enhanced as compared to removal by H 2 O 2 or UV alone (Figure 4).All four possible combinations of UV types (15 W LP and 150 W MP) and H 2 O 2 concentrations (20 and 100 ppm) were tested, namely, (i) LP/20 ppm, (ii) LP/100 ppm, (iii) MP/20 ppm, (iv) and MP/100 ppm.The removal efficacy was found to be influenced by the concentration of H 2 O 2 as well as by the intensity of UV radiation.With highest removal efficiency at MP/100 ppm and LP/100 ppm followed by MP/20 ppm and LP/20 ppm, the concentration of H 2 O 2 seems to be the dominating factor as irradiating with both LP-UV and MP-UV lamps removed >95% of MTBE at 100 ppm H 2 O 2 within 20 minutes of exposure (Figure 4).Results of this study are consistent with previous studies that have also reported additive effect of coupled UV/H 2 O 2 techniques in removal of MTBE from contaminated waters [33].
Similarly, superior efficacy for benzene removal was observed in samples treated with UV/H 2 O 2 process compared to the samples treated separately with UV and H 2 O 2 (Figure 5).Benzene was completely removed within 10 min by all four possible combinations of UV intensities (LP and MP)  and H 2 O 2 concentrations (20 and 100 ppm).More than 95% of benzene was removed within 5 minutes of treatment time in MP/100 ppm and LP/100 ppm.Hence, use of UV/H 2 O 2 is one of the most efficient methods for remediation of benzene contaminated water.It would be interesting to compare benzene removal by this treatment with alternative remediation methods, such as air stripping and activated carbon adsorption.These results are in tendon with earlier studies in which the successful degradation of benzene using UV/H 2 O 2 coupled reactions has been reported [14].

Treatment of MTBE-Benzene Spiked Waters by UV/O 3
Process.This study was designed to evaluate combined effect of UV irradiation and ozone oxidation (UV/O 3 ) on remediation of MTBE-benzene spiked waters.Again, all four possible combinations of UV intensities (LP and MP) and O 3 concentrations (1 and 5 ppm) were tested, namely, (i) LP/1 ppm, (ii) LP/5 ppm, (iii) MP/1 ppm, (iv) and MP/5 ppm.
Results of the experiment are shown in Figures 5 and 6.Remediated process seemed to be influenced mainly by O 3 concentration as maximum removal of MTBE was recorded in the samples treated with MP/5 ppm O 3 as 80%  followed by LP/5 ppm O 3 as 70% in 30 min, while only a little amount of MTBE (40%) was removed at 1 ppm of O 3 concentration in both LP and MP treated samples (Figure 6).Benzene removal by UV/O 3 process was chiefly influenced by the type of UV radiation (Figure 7) as more than 95% of benzene was removed after 30 minutes in samples treated with MP/5 ppm and MP/1 ppm compared to 80% removal in samples treated with LP/1 ppm and LP/5 ppm.Hence, removal of benzene was superior at higher UV intensities.
Results of the remediation assays clearly demonstrate superiority of using MP over LP radiation and H 2 O 2 over O 3 in an AOP.In all studied treatments benzene was removed relatively quickly compared to MTBE.Hence, MTBE may become a slight limiting factor when both contaminants are treated simultaneously by AOPs.
The MTBE degradation by UV/H 2 O 2 process is demonstrated by the formation of by-products (i.e., tertiary butyl formate (TBF) and tertiary butyl alcohol (TBA)) as presented in Figure 8 which shows the formation of TBF and TBA when MTBE is treated with LPUV lamp with 50 ppm H 2 O 2 .

Conclusion
Both MTBE and benzene can be removed simultaneously from contaminated water using advanced oxidation processes employing different combinations of UV, H Layout.Three separate sets of experiments were designed for each type of spiked water.(I) In the first set of experiments, MTBE-benzene spiked waters were treated with two levels of each of H 2 O 2 (20 and 100 ppm), O 3 (1 and 5 ppm), and UV radiation

H 2 O
2 or O 3 concentrations.Higher values of EEO of a contaminant indicate that the contaminant is more difficult to remove [20].The EEO (kWh/m 3 /order) was calculated from the following equation: EEO =  ×  × 1000  × 60 × log (  /  ) ,

Figure 8 :
Figure 8: By-products formed due to degradation of MTBE.

Table 1 :
Characteristics of groundwater samples.
solution at a concentration of 3750 ppm.Actual concentration of H 2 O 2 solution was determined by titration method using standardizing solution (0.1 N) of potassium permanganate (KMNO 4 ).

Table 2 :
The EEO values for MTBE and benzene by LP-UV and MP-UV at 100 ppm of H 2 O 2 .The EEO for Benzene Removal by UV/H 2 O 2 .The optimum conditions for removal of MTBE using the 15 W LP-UV lamp were peroxide (H 2 O 2 ) of 100 ppm (Figure4).Table2summarizes the calculations of the EEO values for MTBE and benzene when treating contaminated groundwater and reducing the concentration of MTBE and benzene by 10 orders of magnitude using both LP 15 W and MP 150 W combined with 100 ppm H 2 O 2 .From the data in Table2, it is estimated that 10 and 2 minutes are needed to irradiate the water spiked with MTBE and benzene, respectively, when using LP-UV+ 100 ppm H 2 O 2 .The values of EEO for MTBE and benzene are 21 kW⋅hr/1000 gal and 6.3 kW⋅hr/1000 gal, respectively, while the EEO values for the 2 contaminants calculated for the 150 W MP-UV lamp were 15.7 and 8.41, respectively.Hence, benzene is relatively easier to remove from water.These results are slightly higher than the EEO levels reported in previously published work (Calgon AOT Handbook, 1996) which stated EEO values of 10 kWh/1000 gal for MTBE and 2-5 kWh/1000 gal for benzene.The main reason for the high values of EEO is attributed to the quality of the groundwater treated which consists of high levels of radical scavengers as discussed above.The EEO values for MTBE removal by UV/O 3 process were not calculated since 10 orders of removal of MTBE were not attained within the 30 minutes of reaction time used in the study.
2 O 2 , and O 3 .Remediation assays clearly demonstrate superiority of International Journal of Photoenergy using MP over LP radiation and H 2 O 2 over O 3 .Superiority of using UV/H 2 O 2 over UV/O 3 process and also its efficiency in removing benzene as compared to MTBE, from contaminated water, were established.It was also confirmed by lower EEO values for UV/H 2 O 2 as compared to that of UV/O 3 .The efficacy of these processes clearly depends upon the concentrations of H 2 O 2 and O 3 and intensity of UV radiation.Furthermore, benzene was comparatively easier to degrade than MTBE.Hence, MTBE may become a slight limiting factor when both contaminants are targeted simultaneously.Detailed studies are required to unearth exact mechanism behind variable behaviors of benzene and MTBE towards AOPs.