- Email: [email protected]
Ultraviolet light emitting diodes and hydrogen peroxide in the photodegradation of aqueous phenol
Journal of Hazardous Materials 161 (2009) 1530–1534
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Ultraviolet light emitting diodes and hydrogen peroxide in the photodegradation of aqueous phenol Sari H. Vilhunen ∗ , Mika E.T. Sillanpa¨ a¨ Laboratory of Applied Environmental Chemistry, Department of Environmental Sciences, University of Kuopio, Patteristonkatu 1, FI-50100 Mikkeli, Finland
a r t i c l e
i n f o
Article history: Received 4 February 2008 Accepted 5 May 2008 Available online 9 May 2008 Keywords: Advanced oxidation processes Hydrogen peroxide Light emitting diode Phenol Ultraviolet
a b s t r a c t The novel system of ultraviolet (UV) light emitting diodes (LED) and hydrogen peroxide (H2 O2 ) was studied for the degradation of phenol as a model organic pollutant in water. The effect of different viewing angles (15 and 120◦ ), wavelengths (255, 265 and 280 nm) and phenol and H2 O2 concentrations were investigated in four photolytic batch reactors. Phenol degradation was observed to be most efficient with UV LEDs emitting at wavelength 280 nm, presumably due to the highest optical power. However, quantum yield for 280 nm reactor was only 0.23 compared to 0.33 of 255 nm reactor. Quantum yields for the rest of the reactors were 0.24 (265 nm, 120◦ ) and 0.22 (265 nm, 15◦ ). UV LEDs in combination with hydrogen peroxide are promising in wastewater treatment in degrading organic compounds, though development of both LEDs and reactor design is needed. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Organic micropollutants occur at trace amounts in municipal wastewater effluents and are released into the aquatic environment, although, industrial sector produces most of the harmful organic chemical load. Some xenobiotic chemicals, such as alkyl phenols and bisphenol A (BPA), in the municipal wastewaters are receiving more interest due to possible endocrine disrupting effects [1,2]. Phenol is a toxic compound and abundant in a variety of different industrial wastewaters. Significant amount of phenol is also found in groundwater [3]. Several biological, physical, electrochemical and photochemical methods are used to treat different wastewaters contaminated with organic compounds [4–7]. Usability of the method is dependent on the content of wastewater. For example, biological processes are not feasible, if the substances to be treated are highly durable or hazardous to microbial population. Ultraviolet (UV) light alone might not be effective in degrading refractory organic compounds and combination with oxidizing agent, e.g. hydrogen peroxide (H2 O2 ), is needed to produce highly efficient free radicals. UV/H2 O2 , UV/O3 and UV/TiO2 are few of the most common photochemical methods known as advanced oxidation processes (AOP). In UV/H2 O2 system, hydroxyl radicals are formed when water containing H2 O2 is exposed to UV light in the range of 200–280 nm, which is also effective in disinfection.
∗ Corresponding author. Tel.: +358 15 355 6236; fax: +358 15 355 6513. E-mail address: Sari.Vilhunen@uku.fi (S.H. Vilhunen). 0304-3894/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2008.05.010
The photolysis of H2 O2 can be described by the following reaction [8]: H2 O2 + UV( ≈ 200–280 nm) → HO• + HO•
(1)
Hydroxyl radical can attack organic molecules by radical addition (Eq. (2)), hydrogen abstraction (Eq. (3)) or electron transfer (Eq. (4)) [8]. Attack by hydrogen abstraction causes removal of hydrogen atom and results a chain reaction where the developed radical organic compound reacts with oxygen producing a peroxyl radical and so on. Two radicals can also combine to form a stable product by radical combination (Eq. (5)). In the following reactions, R is used to describe organic compound. R + HO• → ROH
(2)
R + HO• → R • + H2 O
(3)
R n + HO• → R n−1 + OH−
(4)
HO• + HO• → H2 O2
(5)
Due to the high energy demand of conventional UV lamps, like mercury vapor lamp, other sources of UV light are receiving more interest. Light emitting diode (LED) is a semiconductor p–n junction device, which emits light in a narrow spectrum, produced by a form of electroluminescence [9–11]. AlGaN and AlN are materials for deep ultraviolet light emitting diodes (UV LEDs). LEDs are free of toxicants (e.g. mercury) and consume less energy than traditional lamps, as LED transmits greater amount of energy into light and wastes little energy as heat. Compact size may also be a great
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1531
Nomenclature AOP FID GC LED SPE TOC Vis UV
Advanced oxidation process Flame ionization detector Gas chromatograph Light emitting diode Solid-phase extraction Total organic carbon Visible light Ultraviolet radiation
benefit. Moreover, UV LEDs are hard to break and emit only desirable wavelengths. The usage of UV LEDs in water purification is a new method because UV LEDs emitting wavelengths low enough have emerged just recently. However, preliminary studies utilizing LEDs in degrading organic compounds already exist and LEDs with emission at the wavelengths even as short as 210 nm have been developed [12,13]. The aim of this study was to determine the efficiency of currently marketed UV LEDs in degrading organic pollutants, such as phenol. Moreover, the effect of different emitted wavelengths and viewing angles was studied. Four self-designed batch reactors were used with 10 similar LEDs in each. Three reactors comprised LEDs with viewing angle of 120◦ and emitted wavelengths of 255, 265 or 280 nm, respectively. The LEDs in the fourth reactor emitted at the wavelength of 265 nm but the viewing angle was 15◦ . Optimization of the reaction conditions and pathway of the phenol degradation were not considered in this preliminary study. For the first time, utilizing UV LEDs in water purification was studied using phenol as a key organic pollutant. 2. Materials and methods 2.1. Standards and reagents Methanol (analysis grade, J. T. Baker), hydrogen peroxide ¨ (30%, synthesis grade, Riedel-de Haen) and inner standard 2¨ bromophenol (analytical standard, Riedel-de Haen) were used. Pro-analysis grade phenol and hydrochloric acid were from Merck and sodium chloride (purity 99.5%) from BDH. Inner standard solution for use was made by diluting 67 l of 2-bromophenol to 10 ml of methanol, resulting in a final concentration of 10 g/l. Phenol solutions used in the study were prepared in ultra-pure water.
Fig. 1. Schematic drawing of UV LED batch reactor (1) and other equipment needed: beaker (2), approximated sample height (3), magnetic stirrer (4), magnetic stirring bar (5) and clamp (6).
not supposed to have substantial effect on UV/H2 O2 process in the range of 4–9 [5,14]. pH measurements were carried out using WTW pH 340i pH-meter and electrode. One test was performed using lake ¨ arvi, ¨ water (from Pitkaj Mikkeli, Finland) as a background. 2.3. Photon flux and quantum yield calculations
2.2. Experimental design Four batch reactors using different UV LEDs were prepared with 10 similar LEDs in each. TO-39 LEDs were manufactured by Seoul Optodevice CO., LTD., Korea. Three distinct LED types, with viewing angle of 120◦ (low flat), differed from each other by emitted wavelength, i.e. 255, 265 and 280 nm. The fourth LED type emitted wavelength of 265 nm but the viewing angle was 15◦ caused by the distinct window (flat). LEDs were connected electrically and encapsulated in a black plastic cover. In each batch reactor (Fig. 1), the UV LED system (diameter 7 cm, height 20 cm) was attached with a clamp 1 cm above the sample in a beaker, unlike in conventional UV batch reactors in which the light source (usually mercury vapor lamp, 254 nm) is immersed in the solution [14,15]. Magnetic stirrer provided proper mixing. Sample volume was either 50 or 100 ml. All tests were conducted at room temperature (24 ± 3 ◦ C). Phenol concentration varied from 25 to 200 mg/l and H2 O2 :phenol molar ratio from 5 to 150. Every test was performed twice to bring reliability to the results. Initial pH was always between 6 and 7 because it was
The efficiency of LEDs was monitored throughout the experiments and it remained the same. Radiant flux was measured using integrating sphere (Gigahertz optik/Mitaten Oy) and results as well as other technical information are presented in Table 1. Results are calculated for 10 LEDs of each reactor. Measured radiant flux values and information received from manufacturer had some variation probably because of different ways to perform measurements. Amount of photons absorbed (q0n,p ) in the sample was calculated from radiation power flux (qp ) at wavelength and absorbance (A) of the sample (Eq. (6)) [16]. Radiation power flux was determined from measured radiant flux and typical emission curve of LED. Absorbances of the sample (phenol concentration 100 mg/l, H2 O2 :phenol molar ratio of 100), H2 O2 (3.6 g/l) and phenol (100 mg/l) are presented in Fig. 2 and values for the sample are 1.7 in 276 nm (280 nm reactor) and 2.3 in 269 nm (265 nm reactor) and 255 nm. Measurements were done by UV/vis spectrometer (Perkin Elmer Instruments, Lambda 45). Hydrogen peroxide has greater absorbance in lower wavelengths where absorbance of phe-
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1532
Table 1 Average of actual wavelengths, radiant fluxes and power inputs of different UV LED reactors Wavelength (nm) 255 265 280 265, 15◦
Actual wavelength (nm)
Radiant flux, from manufacturer (mW)
Radiant flux, measured (mW)
Power input (W)
255.45 268.61 276.07 ∼265
2.515 3.406 6.122 3.01–4.00
3.1 5.8 11.6 4.3
2.4 2.7 2.7 2.7
nol is less. Some approximations related to calculations had to be made like estimations of reflection and scattering. It is also important to take into account the radiation absorbed by H2 O2 and photoproducts of phenol degradation, so the determination was done at low conversion of phenol. Decomposition of H2 O2 was slow and concentration high and it was assumed to stay constant. Quantum yield (˚) of a photochemical reaction was calculated by dividing the amount of destroyed phenol (number of events) by the number of adsorbed photons (Eq. (7)) [16]. qp (abs, ) = q0n,p ()(1 − 10−A() )
(6)
() = [dx/dt]/[q0n,p ()(1 − 10−A() )]
(7)
concentration. Column used was HP-5 (Agilent, 5% phenyl methyl siloxane, capillary 30.0 m × 320 m × 0.25 m nominal). Injection mode was splitless and the volume 1 l. Injector and detector temperatures were 300 ◦ C. Carrier gas was helium with the flow of 1.2 ml/min. Oven temperature program was the following: 35◦ (hold 2 min) → 20 ◦ C/min to 260◦ → 15 ◦ C/min to 300◦ (hold 1 min). Phenol removal was reported as a percentage of initial phenol concentration (100 × (C/C0 )). Degradation products were not identified. Total organic carbon (TOC) analysis was performed using SHIMADZU TOC-VCPH total organic carbon analyzer.
3. Results and discussion
2.4. Sample preparation
3.1. Degradation of phenol by UV/H2 O2 method
Solid-phase extraction (SPE) columns (Chromabond C18 ec (octadecyl—modified silica, endcapped), Macherey-Nagel) had a volume of 3 ml and amount of C18 phase 500 mg. Columns were first conditioned with three column volumes of methanol and two column volumes of 0.01 M hydrochloric acid. For the sample treatment, 200 mg of NaCl was added for every 1 ml of sample. HCl (1 M) was used to adjust pH to 2. Sample (2.5 ml) was pipetted to SPE column and slowly, approximately 0.5 ml/min, aspirated through the column. Sample was washed twice with 500 l of 0.01 M HCl and dried under vacuum for 10 min. Elution was performed with 3 × 500 l of methanol. Recovery rate of SPE was evaluated with different concentrations of phenol and it was found to be 92%. Extraction method was based on Macherey-Nagel Application-No.: 301860. Vacuum chamber for extraction was from J. T. Baker (BAKER spe-12G). Prior to the phenol analysis, 50 l of 2-bromophenol solution (10 g/l in methanol) was added to the sample and mixed well.
A gas chromatograph (Agilent 6890 GC) equipped with a flame ionization detector (FID) was used for the determination of phenol
Degradation of phenol (100 mg/l) was insignificant when only UV light from LEDs or hydrogen peroxide (dark test, H2 O2 :phenol molar ratio 100) was used (Fig. 3). In both cases, the maximum degradation rate was only 6% after 6 h. Results are in agreement with the phenol degradation rates of literature when only UV radiation or H2 O2 is used [4,5,17]. Thus, UV light alone is not expected to efficiently degrade phenol. Phenol degradation was fastest with the 280 nm UV LED reactor, presumably because of the highest optical power (Table 1). The sample volume was 100 ml, initial phenol concentration 100 mg/l and H2 O2 :phenol molar ratio 100 in each case. In a six-hours test, the degradation rate ranged from 35 to 65% of the initial concentration depending on the wavelength in the order of 255 nm < 265 nm (viewing angle 15◦ ) < 265 nm < 280 nm (Fig. 3). A 265 nm reactor with viewing angle of 15◦ was only slightly more effective than a 255 nm reactor. The comparison of efficiencies of UV LED reactors with conventional laboratory scale UV systems cannot be thorough with the current preliminary results due to the different equipment structures. However, it seems that the rate of phenol degradation is considerably slow compared to other studies, although, the energy consumption with LEDs is less [5].
Fig. 2. Absorption spectrum of H2 O2 (3.6 g/l), phenol (100 mg/l) and mixture of both.
Fig. 3. Degradation of phenol (100 mg/l) using only H2 O2 or 280 nm reactor or 280, 265 and 255 nm UV reactor with H2 O2 (H2 O2 :phenol molar ratio 100).
2.5. Chemical analysis
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1533
and estimations of the penetration depth of the radiation must be done.
Fig. 4. Degradation of phenol with initial phenol concentration of 25, 100 and 200 mg/l (H2 O2 :phenol molar ratio 100).
3.1.1. H2 O2 :phenol molar ratio The effect of H2 O2 :phenol molar ratio on degradation of phenol was noticed to be mostly irrelevant with ratios 25–150 (265 nm LED, initial phenol concentration 100 mg/l, 6 h), which is in accord with previous studies with conventional UV lamps [14]. However, with molar ratio of 5, degradation was slower, only 25% degraded, compared to results with higher ratios mentioned previously (approximately 50% degraded). Alnaizy and Akgerman [5] showed that very high H2 O2 :phenol molar ratio (e.g. 500) starts to inhibit the phenol photolytic degradation. At high concentrations, H2 O2 can act as a free-radical eliminator and decrease the amount of hydroxyl radicals in the liquid. As a conclusion, the amount of hydroxyl radicals has to be sufficient but not too generous to provide most effective degradation. 3.1.2. Initial phenol concentration and sample volume Initial phenol concentration had a major effect on phenol degradation rate using 280 nm LEDs and H2 O2 :phenol molar ratio of 100. Three different concentrations studied were 25, 100 and 200 mg/l and the percentual phenol degradation was noticeably slower with higher concentration (Fig. 4). With 25 mg/l, nearly 75% of phenol was degraded in 3 h, while with 100 mg/l, the degradation rate was 65% in 6 h and only 40% with phenol concentration of 200 mg/l. Several studies confirm that initial phenol concentration has a remarkable effect on photolytic percentual degradation rate of parent compound, and the higher the concentration, the lower the degradation rate [5,17]. Changing the sample volume from 100 to 50 ml affected the degradation rate substantially, as expected (Fig. 4). Decomposition of phenol was much faster with smaller sample volume when sample layer is thinner and radiation penetrates it better. Almost 90% degradation was achieved during three-hour experiment (H2 O2 :phenol molar ratio 100, initial phenol concentration 25 mg/l). When optimizing the reaction conditions the sample volume and the thickness of the layer to be treated plays a major role
3.1.3. Effect of background Degradation of phenol in lake water was found to be somewhat slower than in ultra-pure water. TOC value of lake water was 8.2 mg/l. In lake water, 54% of phenol (100 mg/l, H2 O2 :phenol molar ratio 100) was degraded within 6 h, while in ultra-pure water the respective degradation was 65%. Differences are caused by various reacting and UV light absorbing compounds, besides added phenol, in lake water (e.g. humic substances and particles). The sample with lake water was colorless in the beginning of the experiment and turned to slightly red by the end of the test suggesting the formation of some degradation products. Color change in ultra-pure water was less visible (from colorless to slightly orange). TOC values were also monitored during the tests but no significant decrease was observed.
3.2. Reactor results Estimated quantum yields at initial time for four reactors were 0.33 (255 nm), 0.24 (265 nm), 0.23 (280 nm) and 0.22 (265 nm, v.a. 15◦ ). In light of these results, the 255 nm radiation seems to be the most efficient opposite to the longest wavelength used. When absorption of H2 O2 is greater (short wavelengths) more radicals are formed to attack phenol molecules. Also in lower wavelengths the absorption of phenol is weaker causing less interference to radical formation. The most inefficient LED type was the one emitting 265 nm radiation with viewing angle of 15◦ having smallest quantum yield and poor effect on phenol degradation. Thus, viewing angle has an impact to reactions in water and must be considered when reactor conditions are optimized.
3.3. Kinetics of phenol degradation Degradation products were not identified in this study but according to the literature they are expected to be organic acids or aromatic byproducts (catechol, hydroquinone, bentzoquinone and resorcinol) [14]. TOC values remained the same during the study, which reveals that degradation of phenol was not complete and the formed products were still organic compounds. In this study the phenol degradation reaction was indicated to be pseudo-first-order, which is accordant to some previous studies [14,17,18]. Some reaction rate constants for different treatments were approximated by integral method and are presented in Table 2 [19]. However, more tests and monitoring of the intermediate products are needed to determine the degradation pathways. Defined kinetic modeling of photolytic degradation of phenol and phenolic compounds can be found from various sources [5,14,18]. The degradation is a complex series of events and complicated kinetic schemes have been proposed [5,20].
Table 2 Reaction rate constants (min−1 ) for phenol degradation by UV/H2 O2 treatment Wavelength of radiation (nm)
Initial C (mg/l)
H2 O2 :phenol molar ratio
Sample volume (ml)
Reaction rate constant (min−1 )
280 280 280 280 265 265 255
25 100 200 25 100 100 100
100 100 100 100 100 5 100
100 100 100 50 100 100 100
0.00760 0.00305 0.00137 0.01206 0.00201 0.00164 0.00133
1534
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
4. Conclusions Degradation rate of phenol by UV LEDs combined with H2 O2 was dependent of the emitted wavelength, viewing angle and initial reactant concentration. Quantum yields of phenol degradation (100 mg/l) were 0.33, 0.24, 0.23 and 0.22 for 255, 265, 280 and 265 nm (v.a. 15◦ ) reactors revealing the most efficient wavelength to be 255 nm. Efficiency of short wavelength is caused by high H2 O2 and low phenol absorption enabling the production of greatest amount of important hydroxyl radicals. The novel method operates similarly with conventional UV techniques. The concentration of H2 O2 effected to phenol decomposition only in considerably high or low amounts and when increasing the initial phenol concentration the percentual degradation rate decreased. As the development of deep UV LEDs is in its first stages and no optimized technical design for reactor solutions has been developed, current preliminary results are encouraging and show potential for UV LEDs in water purification purposes. Taking the characteristics of the water to be treated into account the LED type can be selected so that no energy is consumed to radiation emitted in ineffective wavelengths. Due to the high efficiency and durability of LEDs, the system has great expectations. Acknowledgements This work was financially supported by the Finnish Environmental Science and Technology graduate school (EnSTe), EU and city of Mikkeli. The authors thank MSc. Anshy Plamthottathill and Dr. Sari Luostarinen for language revision and technical assistance in writing the manuscript. Dr. Jaroslaw Puton is thanked for the photochemical calculations. References ´ A.S. Pereira, F.R. Aquino Neto, Characterization of organic [1] M. Castillo, D. Barcelo, pollutants in industrial effluents by high-temperature gas chromatographymass spectrometry, Trends Anal. Chem. 18 (1999) 26–36. [2] D. Voutsa, P. Hartmann, C. Schaffner, W. Giger, Benzotriazoles, alkylphenols and bisphenol A in municipal wastewaters and in the Glatt river, Switzerland, Environ. Sci. Pollut. Res. 13 (2006) 333–341.
[3] Z. He, J. Liu, W. Cai, The important role of hydroxy ion in phenol removal using pulsed corona discharge, J. Electrostat. 63 (2005) 371– 386. [4] O. Legrini, E. Oliveros, A.M. Braun, Photochemical processes for water treatment, Chem. Rev. 93 (1993) 671–698. [5] R. Alnaizy, A. Akgerman, Advanced oxidation of phenolic compounds, Adv. End. Res. 4 (2000) 233–244. [6] L. Amor, M. Eiroa, C. Kennes, M.C. Veiga, Phenol biodegradation and its effect on the nitrification process, Water Res. 39 (2005) 2915–2920. [7] X.-Y. Li, Y.-H. Cui, Y.-J. Feng, Z.-M. Xie, J.-D. Gu, Reaction pathways and mechanisms of the electrochemical degradation of phenol on different electrodes, Water Res. 39 (2005) 1972–1981. [8] G. Tchobanoglous, F.L. Burton, F. Burton, H.D. Stensel, Wastewater Engineering: Treatment and Reuse, 4th ed., Inc. Metcalf and Eddy, McGraw-Hill Companies Inc., New York, USA, 2005. [9] A.H. Crawford, M.A. Banas, M.P. Ross, D.S. Ruby, J.S. Nelson, R. Boucher, A.A. Allerman (Eds), Final LDRD report: Ultraviolet water purification systems for rural environments and mobile applications, Sandia National Laboratories, Albuquerque, USA, 2005. [10] M.A. Khan, AlGaN multiple quantum well based deep UV LEDs and their applications, Phys. Stat. Sol. 203 (2006) 1764–1770. [11] X. Hu, J. Deng, J.P. Zhang, A. Lunev, Y. Bilenko, T. Katona, M.S. Shur, R. Gaska, M. Shatalov, A. Khan, Deep ultraviolet light-emitting diodes, Phys. Stat. Sol. 203 (2006) 1815–1818. [12] W. Wang, Y. Ku, Photocatalytic degradation of reactive red 22 in aqueous solution by UV-LED radiation, Water Res. 40 (2006) 2249–2258. [13] Y. Taniyasu, M. Kasu, T. Makimoto, An aluminum nitride light-emitting diode with a wavelength of 210 nm, Nature 441 (2006) 325–328. [14] H. Kusic, N. Koprivanac, A.L. Bozic, Minimization of organic pollutant content in aqueous solution by means of AOPs: UV- and ozone-based technologies, Chem. Env. J. 123 (2006) 127–137. [15] H. Kusic, N. Koprivanac, A.L. Bozic, I. Selanec, Photo-assisted fenton type processes for the degradation of phenol: a kinetic study, J. Haz. Mat. B136 (2006) 632–644. [16] H.J. Kuhn, S.E. Braslavsky, R. Schmidt, Chemical actinometry (IUPAC technical report), Pure Appl. Chem. 76 (2004) 2105–2146. [17] S.G. Poupopoulos, F. Arvanitakis, C.J. Philippopoulos, Photochemical treatment of phenol aqueous solutions using ultraviolet radiation and hydrogen peroxide, J. Haz. Mat. B129 (2006) 64–68. [18] C. Scheck, F.H. Frimmel, Degradation of phenol and salicylic acid by ultraviolet radiation/hydrogen peroxide/oxygen, Water Res. 29 (1995) 2346–2352. [19] H.S. Fogler, Elements of Chemical Reaction Engineering, Prentice-Hall, Inc., Upper Saddle River, New Jersey, 1999. • [20] A.K. De, B. Chaudhuri, S. Bhattacharjee, B.K. Dutta, Estimation of OH radical reaction rate constants for phenol and chlorinated phenols using UV/H2 O2 photo-oxidation, J. Haz. Mat. B64 (1999) 91–104.
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Ultraviolet light emitting diodes and hydrogen peroxide in the photodegradation of aqueous phenol Sari H. Vilhunen ∗ , Mika E.T. Sillanpa¨ a¨ Laboratory of Applied Environmental Chemistry, Department of Environmental Sciences, University of Kuopio, Patteristonkatu 1, FI-50100 Mikkeli, Finland
a r t i c l e
i n f o
Article history: Received 4 February 2008 Accepted 5 May 2008 Available online 9 May 2008 Keywords: Advanced oxidation processes Hydrogen peroxide Light emitting diode Phenol Ultraviolet
a b s t r a c t The novel system of ultraviolet (UV) light emitting diodes (LED) and hydrogen peroxide (H2 O2 ) was studied for the degradation of phenol as a model organic pollutant in water. The effect of different viewing angles (15 and 120◦ ), wavelengths (255, 265 and 280 nm) and phenol and H2 O2 concentrations were investigated in four photolytic batch reactors. Phenol degradation was observed to be most efficient with UV LEDs emitting at wavelength 280 nm, presumably due to the highest optical power. However, quantum yield for 280 nm reactor was only 0.23 compared to 0.33 of 255 nm reactor. Quantum yields for the rest of the reactors were 0.24 (265 nm, 120◦ ) and 0.22 (265 nm, 15◦ ). UV LEDs in combination with hydrogen peroxide are promising in wastewater treatment in degrading organic compounds, though development of both LEDs and reactor design is needed. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Organic micropollutants occur at trace amounts in municipal wastewater effluents and are released into the aquatic environment, although, industrial sector produces most of the harmful organic chemical load. Some xenobiotic chemicals, such as alkyl phenols and bisphenol A (BPA), in the municipal wastewaters are receiving more interest due to possible endocrine disrupting effects [1,2]. Phenol is a toxic compound and abundant in a variety of different industrial wastewaters. Significant amount of phenol is also found in groundwater [3]. Several biological, physical, electrochemical and photochemical methods are used to treat different wastewaters contaminated with organic compounds [4–7]. Usability of the method is dependent on the content of wastewater. For example, biological processes are not feasible, if the substances to be treated are highly durable or hazardous to microbial population. Ultraviolet (UV) light alone might not be effective in degrading refractory organic compounds and combination with oxidizing agent, e.g. hydrogen peroxide (H2 O2 ), is needed to produce highly efficient free radicals. UV/H2 O2 , UV/O3 and UV/TiO2 are few of the most common photochemical methods known as advanced oxidation processes (AOP). In UV/H2 O2 system, hydroxyl radicals are formed when water containing H2 O2 is exposed to UV light in the range of 200–280 nm, which is also effective in disinfection.
∗ Corresponding author. Tel.: +358 15 355 6236; fax: +358 15 355 6513. E-mail address: Sari.Vilhunen@uku.fi (S.H. Vilhunen). 0304-3894/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2008.05.010
The photolysis of H2 O2 can be described by the following reaction [8]: H2 O2 + UV( ≈ 200–280 nm) → HO• + HO•
(1)
Hydroxyl radical can attack organic molecules by radical addition (Eq. (2)), hydrogen abstraction (Eq. (3)) or electron transfer (Eq. (4)) [8]. Attack by hydrogen abstraction causes removal of hydrogen atom and results a chain reaction where the developed radical organic compound reacts with oxygen producing a peroxyl radical and so on. Two radicals can also combine to form a stable product by radical combination (Eq. (5)). In the following reactions, R is used to describe organic compound. R + HO• → ROH
(2)
R + HO• → R • + H2 O
(3)
R n + HO• → R n−1 + OH−
(4)
HO• + HO• → H2 O2
(5)
Due to the high energy demand of conventional UV lamps, like mercury vapor lamp, other sources of UV light are receiving more interest. Light emitting diode (LED) is a semiconductor p–n junction device, which emits light in a narrow spectrum, produced by a form of electroluminescence [9–11]. AlGaN and AlN are materials for deep ultraviolet light emitting diodes (UV LEDs). LEDs are free of toxicants (e.g. mercury) and consume less energy than traditional lamps, as LED transmits greater amount of energy into light and wastes little energy as heat. Compact size may also be a great
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1531
Nomenclature AOP FID GC LED SPE TOC Vis UV
Advanced oxidation process Flame ionization detector Gas chromatograph Light emitting diode Solid-phase extraction Total organic carbon Visible light Ultraviolet radiation
benefit. Moreover, UV LEDs are hard to break and emit only desirable wavelengths. The usage of UV LEDs in water purification is a new method because UV LEDs emitting wavelengths low enough have emerged just recently. However, preliminary studies utilizing LEDs in degrading organic compounds already exist and LEDs with emission at the wavelengths even as short as 210 nm have been developed [12,13]. The aim of this study was to determine the efficiency of currently marketed UV LEDs in degrading organic pollutants, such as phenol. Moreover, the effect of different emitted wavelengths and viewing angles was studied. Four self-designed batch reactors were used with 10 similar LEDs in each. Three reactors comprised LEDs with viewing angle of 120◦ and emitted wavelengths of 255, 265 or 280 nm, respectively. The LEDs in the fourth reactor emitted at the wavelength of 265 nm but the viewing angle was 15◦ . Optimization of the reaction conditions and pathway of the phenol degradation were not considered in this preliminary study. For the first time, utilizing UV LEDs in water purification was studied using phenol as a key organic pollutant. 2. Materials and methods 2.1. Standards and reagents Methanol (analysis grade, J. T. Baker), hydrogen peroxide ¨ (30%, synthesis grade, Riedel-de Haen) and inner standard 2¨ bromophenol (analytical standard, Riedel-de Haen) were used. Pro-analysis grade phenol and hydrochloric acid were from Merck and sodium chloride (purity 99.5%) from BDH. Inner standard solution for use was made by diluting 67 l of 2-bromophenol to 10 ml of methanol, resulting in a final concentration of 10 g/l. Phenol solutions used in the study were prepared in ultra-pure water.
Fig. 1. Schematic drawing of UV LED batch reactor (1) and other equipment needed: beaker (2), approximated sample height (3), magnetic stirrer (4), magnetic stirring bar (5) and clamp (6).
not supposed to have substantial effect on UV/H2 O2 process in the range of 4–9 [5,14]. pH measurements were carried out using WTW pH 340i pH-meter and electrode. One test was performed using lake ¨ arvi, ¨ water (from Pitkaj Mikkeli, Finland) as a background. 2.3. Photon flux and quantum yield calculations
2.2. Experimental design Four batch reactors using different UV LEDs were prepared with 10 similar LEDs in each. TO-39 LEDs were manufactured by Seoul Optodevice CO., LTD., Korea. Three distinct LED types, with viewing angle of 120◦ (low flat), differed from each other by emitted wavelength, i.e. 255, 265 and 280 nm. The fourth LED type emitted wavelength of 265 nm but the viewing angle was 15◦ caused by the distinct window (flat). LEDs were connected electrically and encapsulated in a black plastic cover. In each batch reactor (Fig. 1), the UV LED system (diameter 7 cm, height 20 cm) was attached with a clamp 1 cm above the sample in a beaker, unlike in conventional UV batch reactors in which the light source (usually mercury vapor lamp, 254 nm) is immersed in the solution [14,15]. Magnetic stirrer provided proper mixing. Sample volume was either 50 or 100 ml. All tests were conducted at room temperature (24 ± 3 ◦ C). Phenol concentration varied from 25 to 200 mg/l and H2 O2 :phenol molar ratio from 5 to 150. Every test was performed twice to bring reliability to the results. Initial pH was always between 6 and 7 because it was
The efficiency of LEDs was monitored throughout the experiments and it remained the same. Radiant flux was measured using integrating sphere (Gigahertz optik/Mitaten Oy) and results as well as other technical information are presented in Table 1. Results are calculated for 10 LEDs of each reactor. Measured radiant flux values and information received from manufacturer had some variation probably because of different ways to perform measurements. Amount of photons absorbed (q0n,p ) in the sample was calculated from radiation power flux (qp ) at wavelength and absorbance (A) of the sample (Eq. (6)) [16]. Radiation power flux was determined from measured radiant flux and typical emission curve of LED. Absorbances of the sample (phenol concentration 100 mg/l, H2 O2 :phenol molar ratio of 100), H2 O2 (3.6 g/l) and phenol (100 mg/l) are presented in Fig. 2 and values for the sample are 1.7 in 276 nm (280 nm reactor) and 2.3 in 269 nm (265 nm reactor) and 255 nm. Measurements were done by UV/vis spectrometer (Perkin Elmer Instruments, Lambda 45). Hydrogen peroxide has greater absorbance in lower wavelengths where absorbance of phe-
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1532
Table 1 Average of actual wavelengths, radiant fluxes and power inputs of different UV LED reactors Wavelength (nm) 255 265 280 265, 15◦
Actual wavelength (nm)
Radiant flux, from manufacturer (mW)
Radiant flux, measured (mW)
Power input (W)
255.45 268.61 276.07 ∼265
2.515 3.406 6.122 3.01–4.00
3.1 5.8 11.6 4.3
2.4 2.7 2.7 2.7
nol is less. Some approximations related to calculations had to be made like estimations of reflection and scattering. It is also important to take into account the radiation absorbed by H2 O2 and photoproducts of phenol degradation, so the determination was done at low conversion of phenol. Decomposition of H2 O2 was slow and concentration high and it was assumed to stay constant. Quantum yield (˚) of a photochemical reaction was calculated by dividing the amount of destroyed phenol (number of events) by the number of adsorbed photons (Eq. (7)) [16]. qp (abs, ) = q0n,p ()(1 − 10−A() )
(6)
() = [dx/dt]/[q0n,p ()(1 − 10−A() )]
(7)
concentration. Column used was HP-5 (Agilent, 5% phenyl methyl siloxane, capillary 30.0 m × 320 m × 0.25 m nominal). Injection mode was splitless and the volume 1 l. Injector and detector temperatures were 300 ◦ C. Carrier gas was helium with the flow of 1.2 ml/min. Oven temperature program was the following: 35◦ (hold 2 min) → 20 ◦ C/min to 260◦ → 15 ◦ C/min to 300◦ (hold 1 min). Phenol removal was reported as a percentage of initial phenol concentration (100 × (C/C0 )). Degradation products were not identified. Total organic carbon (TOC) analysis was performed using SHIMADZU TOC-VCPH total organic carbon analyzer.
3. Results and discussion
2.4. Sample preparation
3.1. Degradation of phenol by UV/H2 O2 method
Solid-phase extraction (SPE) columns (Chromabond C18 ec (octadecyl—modified silica, endcapped), Macherey-Nagel) had a volume of 3 ml and amount of C18 phase 500 mg. Columns were first conditioned with three column volumes of methanol and two column volumes of 0.01 M hydrochloric acid. For the sample treatment, 200 mg of NaCl was added for every 1 ml of sample. HCl (1 M) was used to adjust pH to 2. Sample (2.5 ml) was pipetted to SPE column and slowly, approximately 0.5 ml/min, aspirated through the column. Sample was washed twice with 500 l of 0.01 M HCl and dried under vacuum for 10 min. Elution was performed with 3 × 500 l of methanol. Recovery rate of SPE was evaluated with different concentrations of phenol and it was found to be 92%. Extraction method was based on Macherey-Nagel Application-No.: 301860. Vacuum chamber for extraction was from J. T. Baker (BAKER spe-12G). Prior to the phenol analysis, 50 l of 2-bromophenol solution (10 g/l in methanol) was added to the sample and mixed well.
A gas chromatograph (Agilent 6890 GC) equipped with a flame ionization detector (FID) was used for the determination of phenol
Degradation of phenol (100 mg/l) was insignificant when only UV light from LEDs or hydrogen peroxide (dark test, H2 O2 :phenol molar ratio 100) was used (Fig. 3). In both cases, the maximum degradation rate was only 6% after 6 h. Results are in agreement with the phenol degradation rates of literature when only UV radiation or H2 O2 is used [4,5,17]. Thus, UV light alone is not expected to efficiently degrade phenol. Phenol degradation was fastest with the 280 nm UV LED reactor, presumably because of the highest optical power (Table 1). The sample volume was 100 ml, initial phenol concentration 100 mg/l and H2 O2 :phenol molar ratio 100 in each case. In a six-hours test, the degradation rate ranged from 35 to 65% of the initial concentration depending on the wavelength in the order of 255 nm < 265 nm (viewing angle 15◦ ) < 265 nm < 280 nm (Fig. 3). A 265 nm reactor with viewing angle of 15◦ was only slightly more effective than a 255 nm reactor. The comparison of efficiencies of UV LED reactors with conventional laboratory scale UV systems cannot be thorough with the current preliminary results due to the different equipment structures. However, it seems that the rate of phenol degradation is considerably slow compared to other studies, although, the energy consumption with LEDs is less [5].
Fig. 2. Absorption spectrum of H2 O2 (3.6 g/l), phenol (100 mg/l) and mixture of both.
Fig. 3. Degradation of phenol (100 mg/l) using only H2 O2 or 280 nm reactor or 280, 265 and 255 nm UV reactor with H2 O2 (H2 O2 :phenol molar ratio 100).
2.5. Chemical analysis
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
1533
and estimations of the penetration depth of the radiation must be done.
Fig. 4. Degradation of phenol with initial phenol concentration of 25, 100 and 200 mg/l (H2 O2 :phenol molar ratio 100).
3.1.1. H2 O2 :phenol molar ratio The effect of H2 O2 :phenol molar ratio on degradation of phenol was noticed to be mostly irrelevant with ratios 25–150 (265 nm LED, initial phenol concentration 100 mg/l, 6 h), which is in accord with previous studies with conventional UV lamps [14]. However, with molar ratio of 5, degradation was slower, only 25% degraded, compared to results with higher ratios mentioned previously (approximately 50% degraded). Alnaizy and Akgerman [5] showed that very high H2 O2 :phenol molar ratio (e.g. 500) starts to inhibit the phenol photolytic degradation. At high concentrations, H2 O2 can act as a free-radical eliminator and decrease the amount of hydroxyl radicals in the liquid. As a conclusion, the amount of hydroxyl radicals has to be sufficient but not too generous to provide most effective degradation. 3.1.2. Initial phenol concentration and sample volume Initial phenol concentration had a major effect on phenol degradation rate using 280 nm LEDs and H2 O2 :phenol molar ratio of 100. Three different concentrations studied were 25, 100 and 200 mg/l and the percentual phenol degradation was noticeably slower with higher concentration (Fig. 4). With 25 mg/l, nearly 75% of phenol was degraded in 3 h, while with 100 mg/l, the degradation rate was 65% in 6 h and only 40% with phenol concentration of 200 mg/l. Several studies confirm that initial phenol concentration has a remarkable effect on photolytic percentual degradation rate of parent compound, and the higher the concentration, the lower the degradation rate [5,17]. Changing the sample volume from 100 to 50 ml affected the degradation rate substantially, as expected (Fig. 4). Decomposition of phenol was much faster with smaller sample volume when sample layer is thinner and radiation penetrates it better. Almost 90% degradation was achieved during three-hour experiment (H2 O2 :phenol molar ratio 100, initial phenol concentration 25 mg/l). When optimizing the reaction conditions the sample volume and the thickness of the layer to be treated plays a major role
3.1.3. Effect of background Degradation of phenol in lake water was found to be somewhat slower than in ultra-pure water. TOC value of lake water was 8.2 mg/l. In lake water, 54% of phenol (100 mg/l, H2 O2 :phenol molar ratio 100) was degraded within 6 h, while in ultra-pure water the respective degradation was 65%. Differences are caused by various reacting and UV light absorbing compounds, besides added phenol, in lake water (e.g. humic substances and particles). The sample with lake water was colorless in the beginning of the experiment and turned to slightly red by the end of the test suggesting the formation of some degradation products. Color change in ultra-pure water was less visible (from colorless to slightly orange). TOC values were also monitored during the tests but no significant decrease was observed.
3.2. Reactor results Estimated quantum yields at initial time for four reactors were 0.33 (255 nm), 0.24 (265 nm), 0.23 (280 nm) and 0.22 (265 nm, v.a. 15◦ ). In light of these results, the 255 nm radiation seems to be the most efficient opposite to the longest wavelength used. When absorption of H2 O2 is greater (short wavelengths) more radicals are formed to attack phenol molecules. Also in lower wavelengths the absorption of phenol is weaker causing less interference to radical formation. The most inefficient LED type was the one emitting 265 nm radiation with viewing angle of 15◦ having smallest quantum yield and poor effect on phenol degradation. Thus, viewing angle has an impact to reactions in water and must be considered when reactor conditions are optimized.
3.3. Kinetics of phenol degradation Degradation products were not identified in this study but according to the literature they are expected to be organic acids or aromatic byproducts (catechol, hydroquinone, bentzoquinone and resorcinol) [14]. TOC values remained the same during the study, which reveals that degradation of phenol was not complete and the formed products were still organic compounds. In this study the phenol degradation reaction was indicated to be pseudo-first-order, which is accordant to some previous studies [14,17,18]. Some reaction rate constants for different treatments were approximated by integral method and are presented in Table 2 [19]. However, more tests and monitoring of the intermediate products are needed to determine the degradation pathways. Defined kinetic modeling of photolytic degradation of phenol and phenolic compounds can be found from various sources [5,14,18]. The degradation is a complex series of events and complicated kinetic schemes have been proposed [5,20].
Table 2 Reaction rate constants (min−1 ) for phenol degradation by UV/H2 O2 treatment Wavelength of radiation (nm)
Initial C (mg/l)
H2 O2 :phenol molar ratio
Sample volume (ml)
Reaction rate constant (min−1 )
280 280 280 280 265 265 255
25 100 200 25 100 100 100
100 100 100 100 100 5 100
100 100 100 50 100 100 100
0.00760 0.00305 0.00137 0.01206 0.00201 0.00164 0.00133
1534
¨ / Journal of Hazardous Materials 161 (2009) 1530–1534 S.H. Vilhunen, M.E.T. Sillanp¨ aa
4. Conclusions Degradation rate of phenol by UV LEDs combined with H2 O2 was dependent of the emitted wavelength, viewing angle and initial reactant concentration. Quantum yields of phenol degradation (100 mg/l) were 0.33, 0.24, 0.23 and 0.22 for 255, 265, 280 and 265 nm (v.a. 15◦ ) reactors revealing the most efficient wavelength to be 255 nm. Efficiency of short wavelength is caused by high H2 O2 and low phenol absorption enabling the production of greatest amount of important hydroxyl radicals. The novel method operates similarly with conventional UV techniques. The concentration of H2 O2 effected to phenol decomposition only in considerably high or low amounts and when increasing the initial phenol concentration the percentual degradation rate decreased. As the development of deep UV LEDs is in its first stages and no optimized technical design for reactor solutions has been developed, current preliminary results are encouraging and show potential for UV LEDs in water purification purposes. Taking the characteristics of the water to be treated into account the LED type can be selected so that no energy is consumed to radiation emitted in ineffective wavelengths. Due to the high efficiency and durability of LEDs, the system has great expectations. Acknowledgements This work was financially supported by the Finnish Environmental Science and Technology graduate school (EnSTe), EU and city of Mikkeli. The authors thank MSc. Anshy Plamthottathill and Dr. Sari Luostarinen for language revision and technical assistance in writing the manuscript. Dr. Jaroslaw Puton is thanked for the photochemical calculations. References ´ A.S. Pereira, F.R. Aquino Neto, Characterization of organic [1] M. Castillo, D. Barcelo, pollutants in industrial effluents by high-temperature gas chromatographymass spectrometry, Trends Anal. Chem. 18 (1999) 26–36. [2] D. Voutsa, P. Hartmann, C. Schaffner, W. Giger, Benzotriazoles, alkylphenols and bisphenol A in municipal wastewaters and in the Glatt river, Switzerland, Environ. Sci. Pollut. Res. 13 (2006) 333–341.
[3] Z. He, J. Liu, W. Cai, The important role of hydroxy ion in phenol removal using pulsed corona discharge, J. Electrostat. 63 (2005) 371– 386. [4] O. Legrini, E. Oliveros, A.M. Braun, Photochemical processes for water treatment, Chem. Rev. 93 (1993) 671–698. [5] R. Alnaizy, A. Akgerman, Advanced oxidation of phenolic compounds, Adv. End. Res. 4 (2000) 233–244. [6] L. Amor, M. Eiroa, C. Kennes, M.C. Veiga, Phenol biodegradation and its effect on the nitrification process, Water Res. 39 (2005) 2915–2920. [7] X.-Y. Li, Y.-H. Cui, Y.-J. Feng, Z.-M. Xie, J.-D. Gu, Reaction pathways and mechanisms of the electrochemical degradation of phenol on different electrodes, Water Res. 39 (2005) 1972–1981. [8] G. Tchobanoglous, F.L. Burton, F. Burton, H.D. Stensel, Wastewater Engineering: Treatment and Reuse, 4th ed., Inc. Metcalf and Eddy, McGraw-Hill Companies Inc., New York, USA, 2005. [9] A.H. Crawford, M.A. Banas, M.P. Ross, D.S. Ruby, J.S. Nelson, R. Boucher, A.A. Allerman (Eds), Final LDRD report: Ultraviolet water purification systems for rural environments and mobile applications, Sandia National Laboratories, Albuquerque, USA, 2005. [10] M.A. Khan, AlGaN multiple quantum well based deep UV LEDs and their applications, Phys. Stat. Sol. 203 (2006) 1764–1770. [11] X. Hu, J. Deng, J.P. Zhang, A. Lunev, Y. Bilenko, T. Katona, M.S. Shur, R. Gaska, M. Shatalov, A. Khan, Deep ultraviolet light-emitting diodes, Phys. Stat. Sol. 203 (2006) 1815–1818. [12] W. Wang, Y. Ku, Photocatalytic degradation of reactive red 22 in aqueous solution by UV-LED radiation, Water Res. 40 (2006) 2249–2258. [13] Y. Taniyasu, M. Kasu, T. Makimoto, An aluminum nitride light-emitting diode with a wavelength of 210 nm, Nature 441 (2006) 325–328. [14] H. Kusic, N. Koprivanac, A.L. Bozic, Minimization of organic pollutant content in aqueous solution by means of AOPs: UV- and ozone-based technologies, Chem. Env. J. 123 (2006) 127–137. [15] H. Kusic, N. Koprivanac, A.L. Bozic, I. Selanec, Photo-assisted fenton type processes for the degradation of phenol: a kinetic study, J. Haz. Mat. B136 (2006) 632–644. [16] H.J. Kuhn, S.E. Braslavsky, R. Schmidt, Chemical actinometry (IUPAC technical report), Pure Appl. Chem. 76 (2004) 2105–2146. [17] S.G. Poupopoulos, F. Arvanitakis, C.J. Philippopoulos, Photochemical treatment of phenol aqueous solutions using ultraviolet radiation and hydrogen peroxide, J. Haz. Mat. B129 (2006) 64–68. [18] C. Scheck, F.H. Frimmel, Degradation of phenol and salicylic acid by ultraviolet radiation/hydrogen peroxide/oxygen, Water Res. 29 (1995) 2346–2352. [19] H.S. Fogler, Elements of Chemical Reaction Engineering, Prentice-Hall, Inc., Upper Saddle River, New Jersey, 1999. • [20] A.K. De, B. Chaudhuri, S. Bhattacharjee, B.K. Dutta, Estimation of OH radical reaction rate constants for phenol and chlorinated phenols using UV/H2 O2 photo-oxidation, J. Haz. Mat. B64 (1999) 91–104.