Uv irradiation versus combined uv/hydrogen peroxide and uv/ozone treatment for the removal of persistent organic pollutants from water

International Conference Ozone and UV, April 3rd 2006 UV Irradiation versus combined UV / Hydrogen Peroxide and
UV / Ozone Treatment for the Removal of Persistent Organic
Pollutants from Water

Muriel Sona, Christine Baus, Heinz-Jürgen Brauch
DVGW - Technologiezentrum Wasser, Karlsruher Str. 84, 76139 Karlsruhe, Germany
Keywords: UV, UV / H2O2, UV / ozone, quantum yield, EE/O, sulfamethoxazole, iopromide,
Introduction
Most often, conventional drinking water treatment technologies such as aeration, filtration or adsorption on activated carbon are not able to remove persistent organic pollutants completely. Therefore new technologies have to be evaluated with regard to their removal efficiency. Chemical oxidation is a promising option and with ozonation a wide range of substances can already be eliminated. However, even this treatment proves to be inefficient for a number of emerging contaminants. Therefore more advanced oxidation technologies are investigated, which are able to produce highly reactive OH radicals to accelerate the degradation of persistent substances. In this study the elimination potential of UV-irradiation and advanced oxidation processes, namely UV / hydrogen peroxide and UV These substances include pharmaceuticals (sulfamethoxazole and iopromide) and gasoline additives such as MTBE (methyl-tertiary-butyl-ether) and TAME (tertiary-amyl-methyl-ether). The antibiotic sulfamethoxazole is disposed aga . Its widespread use and low degradability in the aquatic environment leads to a rather frequent detection in surface water as well as groundwaters in Germany. Concentrations levels amount to up to 100 ng/L. [1] Iopromide belongs to the family of iodinated X-ray contrast media. They are used for displaying soft tissues and vessels during radiological examinations. After application they leave the human body to 95% non-metabolised within 24 h, mostly via urine excretion. Since no elimination occurs in sewage treatment plants, levels of up to 1µg/L can be detected in surface waters. [2-4] Fuel oxygenates such as MTBE and TAME are used to enhance the octane number of gasoline. Due to their persistency during subsoil passage they are often found in the raw water of water treatment plants using riverbank filtrate as raw water source. The acute toxicity of MTBE and TAME is rather low; experts debate, however, whether there are chronic toxic effects. However, the taste and odour threshold of MTBE lies at 10 to 30 µg/L and therefore it is not desirable in drinking water. Concentrations levels found in German surface waters average 0.2 to 0.3 µg/L, but peak concentrations of more than 10 µg/L were measured in the river Rhine. [5-9] Material and Methods
The substances under investigation were used as purchased from Merck, Darmstadt, Germany (MTBE, TAME, H2O2), Sigma Aldrich, Taufkirchen, Germany (sulfamethoxazole) and Promochem, Wesel, Germany (iopromide) in analytical grade. A stock solution of 1 g/L was prepared in demineralised water from each of the compounds. MTBE and TAME were analysed using purge&trap extraction with subsequent GC/MS detection [10]. Sulfamethoxazole was quantified using a HPLC MS-MS technique [2]. For the detection of iopromide a newly developed method was used combining inductive coupled plasma with mass spectrometry. [11] Hydrogen peroxide concentration was measured photometrically with titanium reagent using the German standard method DIN 38409. International Conference Ozone and UV, April 3rd 2006 Gaseous ozone was produced from pure oxygen, medical grade, by an ozone generator provided by Anseros GmbH, Tübingen, Germany. The oxygen/ozone mixture was introduced in a bubble column filled with demineralised water at 5 °C resulting in an ozone stock solution with concentrations of approx. 25 mg/L ozone. Ozone was analysed photometrically using the indigo method according to DIN 38408 - G3/3. Drinking water spiked with the substance under investigation was used in most of the experiments. Characteristic parameters are on display in Table 1. Laboratory-scale experiments were carried out in batch scale. A closed loop UV reactor equipped with a 40 W low pressure mercury lamp was used for the irradiation experiments. The reservoir, a glass vessel, contains 8 L of drinking water that was pumped through the UV reactor. The lamp has two main emission lines at λ= 184.9 nm and 253.7 nm. Initial concentration of the substances under investigation was 1 mg/l. After defined time intervals samples were taken and analysed for the compounds. For the advanced oxidation experiments, H2O2 or ozone was added after the annealing of the lamp. Concentrations were adjusted by admixing a defined amount of stock solutions of either H2O2 or ozone. Initial concentration of H2O2 and ozone were 4 and 3-6 mg/L, respectively. During the experiments the concentration decline of the oxidants was monitored additionally. Run-time of the UV-, UV/H2O2- and UV/ozone- experiments was 30, 30 and 15 min, respectively. Table 1: Characteristic parameters for the drinking water used in the experiments Parameter
Parameter
Results and Discussion
UV irradiation
UV irradiation experiments were carried out to show the degradability of the substances under investigation by UV photooxidation. In Figure 1 the results of the irradiation experiments in drinking water are on display. Sulfamethoxazole is removed to nearly 100% within the duration of the experiment. Iopromide is also noticeably reduced in concentration, although to a lesser extent. Though the fuel additives MTBE and TAME exhibit no absorbance in the UV spectrum, a slight concentration decline was detected (23-26 %). This can probably be attributed to the reaction of the substances with radicals which are formed during UV irradiation. A measure for the efficiency by which a substance is photochemically transformed, is the quantum yield Φ. It is defined as the amount of molecules of the substance reacting per time unit in ratio to the amount of photons per time unit which are absorbed by the photoreactive substance [12]. In Table 2 the calculated quantum yields for sulfamethoxazole and iopromide are on display. Sulfamethoxazole exhibits a high quantum yield, indicating that the photoreaction is highly efficient. On an average 1.2 to 1.6 absorbed photons yield a conversion of the molecule. Iopromide on the other side needs a much larger amount of absorbed photons in order to be converted. The calculated quantum yields are dependent on the chosen matrix. In demineralised water the values are decisively higher than in drinking water. This can be attributed to "dark" secondary reactions taking place in the reaction mixture apart from the initial photochemically induced direct conversion. These reactions often involve the formation of radicals which in turn attack International Conference Ozone and UV, April 3rd 2006 non specifically other water ingredients. In drinking water the amount of radical scavenging species such as carbonate and bicarbonate is higher, thus interfering any side reactions where radicals are involved. The assumption that secondary radicals are formed, is strengthened by the observation that during UV irradiation MTBE is eliminated faster in demineralised water than in drinking water (cp. reaction rate constant kUV in Table 2). Since MTBE does not absorb UV light, secondary reactions have to take place which are responsible for the MTBE conversion. iopromide
sulfamethoxazole
concentration decline c/c
time [min]
Concentration decline of sulfamethoxazole, iopromide and MTBE/TAME in drinking water during
UV-irradiation. The error bars in the graph show the relative measurement uncertainty (c0=1
mg/L)
Table 2: Calculated quantum yield and reaction rate constants for the substances under investigation during UV irradiation, combined UV / H2O2 and UV / ozone
Advanced oxidation processes: combined UV / H2O2 and UV / ozone
In order to enhance the elimination efficiency an oxidant was added during UV irradiation. This substance itself absorbs UV light and reacts with water forming highly reactive OH radicals. Commonly used oxidants include hydrogen peroxide (H2O2 ) or ozone. Introducing H2O2 at a concentration of 4 mg/L to the reaction mixture induced only a slight enhancement of the degradation (Fig. 2). During the experiments only a slight decline in the peroxide concentration was observed. This can be attributed to the reformation of H2O2 by the homolytic cleavage of water during UV irradiation at lower wave lengths (<200nm). Furthermore the absorbance efficiency of H2O2 is dependent on the concentration. The chosen concentration of 4 mg/L may be less efficient in the production of OH radicals. This also explains the low removal efficiency of the advanced oxidation process. Experiments of the UV/H2O2 process for the removal of MTBE with higher concentrations of H2O2 (up to 100 mg/L, data not shown) showed a much better performance. International Conference Ozone and UV, April 3rd 2006 However, in this study, H2O2 concentrations close to those commonly used in drinking water treatment were chosen. For the UV / ozone experiments ozone was added instead of hydrogen peroxide. The decline of sulfamethoxazole was not investigated since this substance is already quickly eliminated by pure UV irradiation and therefore the application of UV / ozone is not considered to lead to major improvements. In contrast to H2O2 the ozone was consumed very fast; within 5 min the ozone concentration reached nil. This is mirrored in the concentration decline of MTBE and iopromide (cp. Fig. 3). A very steep concentration curve indicates the fast formation of OH radicals by the photolytic destruction of the applied ozone and the subsequent non-selective reaction of the radicals with MTBE and iopromide. Up to 60% of the MTBE was eliminated. The prolongation of the experiment proved to be inefficient, the concentration of MTBE decreased no further. In the case of iopromide the concentration decline continued - however, to a lesser extent. This final decline shows the influence of the pure UV irradiation. With a constant dosing of ozone during the experiment a complete conversion of the substances could be reached. concentration decline c/c
time [min]
Comparison of the concentration decline during UV-irradiation with the UV/H2O2-process (c0=1
mg/L, cH2O2=4 mg/L; experiments carried out in drinking water if not otherwise stated)
International Conference Ozone and UV, April 3rd 2006 entration decline c/c
time [min]
Comparison of UV-irradiation with UV/H2O2 and UV/O3-process
A comparison of the reaction rate constants calculated from the experiments is on display in
Table 2. The reaction rates are assumed to be pseudo first order and they are used as “overall
rate constants” for the evaluation of advanced oxidation processes. This proceeding is
established from experience and literature [13]. It can clearly be seen, that the combined
process UV / ozone includes the highest ration of the fast radical reactions with its reaction rate
constants being ten times higher than those of the other processes.

Bromate and Nitrite formation
The ozonation of bromide containing natural waters can lead to the formation of bromate [14,15]. Bromate is considered to be potentially carcinogenic, so it is limited in the German drinking water decree to 10 µg/L (applicable from 2008). During the pure UV irradiation and UV / H2O2 experiments no bromate formation of was observed as expected. However, in all UV / ozone experiments bromate was formed in concentrations between 25 and 32 µg/L thus exceeding the drinking water limit. Nitrite can be formed during UV-irradiation of nitrate containing natural waters [16]. In the
experiments nitrite formation was only observed in UV irradiation and UV / H2O2 experiments
partly exceeding the limit set by the German drinking water decree of 0.1 mg/L.
Concentrations found ranged between 0.06 to 0.21 mg/L. During the experiments with ozone,
no nitrite was formed since ozone was able to oxidise any nitrite accidentally being formed by
the irradiation.
Metabolite research
Apart from quantifying iopromide the analytical method for iopromide allows for a detection of iodide and iodate and also a detection of iodine containing metabolites. The main degradation products were iodide and iodate, which accounts for a destruction of the iodine-carbon bond by the UV irradiation. However, other iodine containing metabolites were observed as well. The quantification of iodine over the experiment duration (the ‘iodine sum’) showed an almost complete recovery of the initially added iodine in iopromide (cp Fig 4). The calculation of total organic iodine (TOI) and total inorganic iodine (TII, i.e. iodide and iodate) showed a proportional increase of the inorganic iodine during the decrease of the organic iodine fraction. That underlines that the main part of the iodine contained in iopromide converts into inorganic iodide or iodate. Remarkably, only a very low part of iodate is formed during UV irradiation and International Conference Ozone and UV, April 3rd 2006 in applying the UV / H2O2 process; during the UV / ozone process, however, a complete conversion of iodide to iodate is observed. UV irradiation
UV / ozone
Fig. 4: Metabolite formation during oxidation of iopromide (x-axis: time [min]; y-axis: calculated
Economic evaluation of the processes
The figure-of-merit ‘electrical energy per order of magnitude’ (EE/O) is an evaluation parameter independent of the process procedure. It represents the electrical energy (kWh) required to diminish the concentration of a pollutant one order of magnitude in 1 m³ (1000 L). In order to compare the different advanced oxidation processes the EE/O was calculated. The results show that the UV / ozone process is the most efficient process for the removal of the organic substances under investigation showing the lowest EE/O values (Table 3). With the lowest input of electrical energy the highest elimination rates could be achieved. For the gasoline additives MTBE and TAME UV irradiation and the combination of UV with hydrogen peroxide at the concentrations applied is not a suitable removal method as can be seen by the high EE/O. However, sulfamethoxazole is easily removed with UV irradiation and UV /H2O2 represented by the low values of the EE/O parameter. Iopromide shows a fairly good EE/O for the UV and UV / H2O2 process, but both, iopromide and MTBE, are eliminated much more efficiently by the UV / ozone process. However, higher investment and maintenance costs have to be taken into account, since ozone has to be produced on-site and the equipment is much more sophisticated. International Conference Ozone and UV, April 3rd 2006 Table 3 Comparison of the parameter EE/O for all applied processes Conclusion
It is possible to use the advanced oxidation processes UV / H2O2 and UV / ozone for the degradation of persistent organic compounds. The order of elimination efficiency was identical in all applications and derived as follows: Sulfamethoxazole > Iopromide > TAME > MTBE
The most efficient technology for the removal of these substances proved to be the UV / ozone process. However, the applicability and manageability of this process requires extended resources and knowledge. The addition of H2O2 in low concentrations showed only slight enhancement of the degradation profile. More investigations into the best dosage of H2O2 are under way. The gasoline additives are only removed noticeably by the application of advanced oxidation processes such as UV / H2O2 or UV / ozone. Further investigations have to be conducted concerning the behaviour of alternative ethers (TAME, DIPE, GTBE or ETBE). Metabolite formation could be proven for the X-ray contrast medium iopromide. Generally speaking, advanced oxidation processes mostly cause by-products which might be more harmful than the original substance. For example it is well known, that the decomposition of MTBE causes intermediate products such as the carcinogenic substance tert-butyl-alcohol (TBA). Therefore further studies have to be made into the formation of reaction by-products during the oxidation experiments. References
[1] Sacher F, Brauch H-J: Abschlussbericht vom Forschungsvorhaben des Ministeriums für Umwelt und Verkehr des Landes Baden-Württemberg: Vorkommen von Pharmaka und Hormonen in Grund-, Oberflächenwässern und Böden in Baden-Württemberg. 2002. [2] Sacher F, Lange FT, Brauch H-J, Blankenhorn I. Pharmaceuticals in groundwaters - analytical methods and results of a monitoring program in Baden-Württemberg, Germany. Journal of Chromatography A 2001; 938:199-210. [3] Putschew A, Wischnack S, Jekel M. Occurrence of triiodinated X-ray contrast agents in the aquatic environment. The Science of the Total Environment 2000; 255:129-134. [4] Wischnack S, Oleksy-Frenzel J, Jekel M. Abbauverhalten und Vorkommen organischer Jodverbindungen im Raum Berlin. GdCH Fachgruppentagung Wasserchemie 1998;96-99. [5] Schmidt TC. Analysis of methyl tert-butyl ether (MTBE) and tert-butyl alcohol (TBA) in ground and surface water. Trends in Analytical Chemistry 2003; 22(10):776-784. [6] Transfer pathways of MTBE into groundwater: the role of diffuse vs. point sources. First European Conference on MTBE, September 8-9: Dresden University of Technology, 2003. [7] Achten C, Kolb A, Püttmann W, Seel P, Gihr R. Methyl tert-butyl ether (MTBE) in river and wastewater in Germany. Environmental Science and Technology 2002; 36:3652-3661. [8] Baus C, Hung H-W, Sacher F, Fleig M, Brauch H-J. MTBE in drinking water production - occurrence and efficiency of treatment technologies. Acta Hydrochimica et Hydrobiologica 2005; 33(2):118-132. International Conference Ozone and UV, April 3rd 2006 [9] Baus C, Sacher F, Brauch H-J. Efficiency of ozonation and AOP for methyl tert-butylether (MTBE) removal in waterworks. Ozone: Science and Engineering 2005; 27(1):27-35. [10] Klinger J, Stieler C, Sacher F, Brauch H-J. MTBE (methyl tertiary-butyl ether) in groundwaters: monitoring results from Germany. Journal of Environmental Monitoring 2002; 4:276-279. [11] Sacher F, Raue B, Brauch H-J: Analysis of iodinated X-ray contrast agents in water samples by ion chromatography and inductively-coupled plasma mass spectrometry, Journal of Chromatography A, 1085 (2005), 117-123 [12] Oppenländer T: Photochemical purification of water and air, Wiley-VCH-Verlag, Weinheim (2003) [13] Graham J L, Striebich R, Patterson C L, Krishnan E, Haught R C: MTBE oxidation by-products from the treatment of surface waters by ozonation and UV-ozonation, Chemosphere, 54 (2004), 1011-1016 [14] Sacher F., Schmidt W., Böhme U., Brauch H.-J.: Bromat - Ein Problem für die Trinkwasserversorgung in Deutschland? Zeitschrift des Gas- und Wasserfaches, 138 (4) (1997) [15] von Gunten U., Hoigné J.: Bromate Formation during Ozonation of bromide-containing Waters: Interaction of Ozone and Hydroxyl Radical Reactions, Environmental Science & Technology 28 (7) (1994), S. 1234-1242 [16] Gonzalez M C, Braun A M: VUV photolysis of aqueous solutions of nitrate and nitrite, research ob Chemical Intermediates, 21 (8/9), (1995), 837-859

Source: http://berlin-water-week.com/fileadmin/user_upload/pdf/veranstaltungen/IOA_Ozone_and_UV/14_IOAEA3G_Baus.pdf