VUV irradiation emitted from microwave discharge electrodeless lamps

Available online at www.sciencedirect.com

Chemosphere 71 (2008) 1774–1780 www.elsevier.com/locate/chemosphere

Technical Note

Photolysis of low concentration H2S under UV/VUV irradiation emitted from microwave discharge electrodeless lamps Xia Lan-Yan a, Gu Ding-Hong a, Tan Jing b, Dong Wen-Bo a,*, Hou Hui-Qi a,* b

a Institute of Environmental Science, Fudan University, Shanghai 200433, China Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117576, Singapore

Received 18 August 2007; received in revised form 14 January 2008; accepted 14 January 2008 Available online 10 March 2008

Abstract The photolysis of simulating low concentration of hydrogen sulfide malodorous gas was studied under UV irradiation emitted by selfmade microwave discharge electrodeless lamps (i.e. microwave UV electrodeless mercury lamp (185/253.7 nm) and iodine lamp (178.3/ 180.1/183/184.4/187.6/206.2 nm)). Experiments results showed that the removal efficiency (gH2 S ) of hydrogen sulfide was decreased with increasing initial H2S concentration and increased slightly with gas residence time; H2S removal efficiency was decreased dramatically with enlarged pipe diameter. Under the experimental conditions with pipe diameter of 36 mm, gas flow rate of 0.42 standard l s1, gH2 S was 52% with initial H2S concentration of 19.5 mg m3 by microwave mercury lamp, the absolute removal amount (ARA) was 4.30 lg s1, and energy yield (EY) was 77.3 mg kW h1; gH2 S was 56% with initial H2S concentration of 18.9 mg m3 by microwave iodine lamp, the ARA was 4.48 lg s1, and the EY was 80.5 mg kW h1. The main photolysis product was confirmed to be SO2 4 with IC. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Microwave discharge electrodeless lamp; Hydrogen sulfide; Photolysis; Mercury; Iodine

1. Introduction There is a growing concern on the malodorous pollution at sewage and industrial wastewater treatment plants because it is not only a nuisance in the ambient environment but also poses adverse health effects to humans. Sulfur compounds such as hydrogen sulfide, methanethiol (CH3SH), dimethyl sulfide (DMS), dimethyl disulfide, are mainly responsible for the nauseous odors emitted from these plants. They are characterized by high toxicity and extremely low odor threshold values, e.g. 7.6  104 mg m3 for H2S (De Zwart and Keunen, 1992). Therefore, the effective removal of these sulfur compounds from air is highly desired. Photo-induced processing has been widely applied in industries including automobile manufacturing, electron*

Corresponding authors. Tel.: +86 021 65642230; fax: +86 021 65642293 (W.-B. Dong). E-mail address: [email protected] (W.-B. Dong). 0045-6535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2008.01.050

ics, textiles, chemicals, medical treatment and wastes control. The principle of sulfur compounds removal by photolysis may find strong support from some kinetic studies on UV photolysis of H2S, CH3SH and DMS. Conventional photodissociation processing is performed on narrow wavelength irradiated by lasers (193–252 nm), deuterium lamps (190–400 nm) or black light lamps (330– 400 nm) (Rogers et al., 1996; Wilson et al., 1996; Martinez-Haya et al., 1999; Koda et al., 2001). However, their application on waste removal is restricted by high cost or technical limitations. Taking lasers for example, they are generally expensive and provide only small area beams. Some popular photolysis technologies need the assistance of catalysts, and the problems of the poisoning or deactivation of photocatalyst caused by accumulation of organic products and/or sulfates on the surface of catalyst have not been well solved. Moreover, conventional UV light sources, such as mercury vapor lamps and high-pressure xenon lamps, are incapable of producing intense near UV or deep UV lights. Taking a low pressure mercury lamp

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

as an example, it typically only emits irradiation a few mW cm2 at 185 nm and around 3–4 times of this amount at 253.7 nm (Zhang and Boyd, 2000). Therefore, UV sources and techniques, with low cost, high intensity and energy efficiency, are extremely desirable for waste removal. In this paper, a new-type UV light source, microwave discharge electrodeless lamps (MDELs) are employed. Unlike the conventional lamps energized by the electric field between the electrodes, the principle of MDELs is that the gas or materials in the lamp absorb the microwave (MW) energy to form stable UV-emitting discharge plasma. Such unique discharge pattern endows the MDELs with lots of advantages, in terms of long lifetime of UV output, high UV radiant power (Al-Shamma’a et al., 2001), and adaptable lamp shapes. Moreover, more types of working gases or materials are available to produce a variety of atomic or excimer emissions in short-wavelength UV spectral regions, e.g. 206.2 nm from I* (Kerst et al., 1998), 222 nm from KrCl* (Hassal and Ballik, 1991), and 228 nm from Cd* (Limbeck, 2006). Most of these MDELs are still under laboratorial investigation and need further optimizations to improve the efficiency. Only the mercury MDELs have been commercialized now, as mercury is the most readily to be excited, and even a domestic microwave-oven may act as the MW power supply (Kla´n et al., 1999). Employing mercury MDELs to water sterilization (Iwaguch et al., 2002; Pandithas et al., 2003) and photodissociation of organic pollutants in aqueous solution (Horikoshi et al., 2002, 2004; Ai et al., 2005; Muller et al., 2005; Hong et al., 2006; Zhang et al., 2006a,b; Gao et al., 2007) have been studied extensively in the past few years. Yet to the best of our knowledge, little has been done on the photolysis of odor gas pollutants using MDELs as the UV light source. In this paper, a preliminary study on the photolysis of H2S with two self-developed MDELs was investigated. Filled with binary mixtures of Hg–Ar and I2–Kr, respectively, MDEL-Hg and MDEL-I2 were found to emit

Microwave power supply

intense atomic lines of mercury and iodine in both UV and vacuum-UV (VUV) region. A relatively low concentration level of H2S, in the range of 0–25 mg m3, was selected for this work, in consideration of the generally low concentrations of H2S monitored at wastewater treatment plants. The removal performance was examined under different conditions with different initial H2S concentration, gas residence time and cross-section area of gas path. The energy consumptions by this method were also evaluated. 2. Experimental section 2.1. Microwave electrodeless lamps MDELs consisted of a MW power supply and electrodeless lamps. The MW power supply, operated at a fixed frequency of 2.45 GHz, was developed with the assistance of Shanghai Yalian Microwave Tec. Co. As shown in Fig. 1, it is assembled with a upper box for generating high voltage and a lower box equipped with a magnetron and a resonant cavity. A high-quality electric cable was used to transmit the high voltage to the magnetron to establish a uniform MW filed in the resonant cavity. With 200 W of input electrical power supply, the output of MW was 80 W which was sufficient to ignite the quartz lamp. The rest power is converted to heat and/or consumed by other auxiliary equipments such as cable and ventilator. Further investigations on the conversion ratio of input electric power to MW energy are needed. The electrodeless lamps were made of quartz tubes with 20 mm external diameter and sealed up at both ends. Two MDELs were investigated in this work, one was 100 cm long MDEL-Hg filled with Hg (20 mg) and Ar (0.267 kPa), and another was MDEL-I2 with I2 (2 mg) and Kr (0.067 kPa) with the length of 80 cm. With one end inserted into the cavity, the remaining part of the lamp outside the cavity is fixed at the center of a glass tube to serve as the photo-reactor for H2S photolysis.

High voltage power

Outflow

Electric cable

Mass flow meter

H2S gas

1775

Electrodeless lamp

Sampling position II

Baffle plate Blower

Magnetron Resonant cavity

Sampling position I Cyclone device

Inflow

Fig. 1. The sketch of experimental photolysis reaction system.

1776

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

2.2. Experimental setup

3. Results and discussion

The experimental setup designed for the evaluation of MDEL removal efficiency is shown in Fig. 1. The system consisted of a continuous flow gas generation system, a MDEL reactor and a gas analysis system. H2S (99.9%, Nanjing Mucop Nanfen Special Gas Co.) was mixed with ambient air through a cyclone device before introducing into the photolysis area. The gas stream, driven by a terminal blower, passed through the reactor for 15 min to allow the system to reach the steady state. Power was then applied to inspire the lamp for another 10 min to make sure the new steady state had been achieved before the measurement of concentration of H2S in the gas stream. The initial H2S concentrations were ranged from 0 to 25 mg m3, and gas residence time was between 1 and 5 s. They were controlled by a mass flow meter and adjusting a baffle plate located at the end of the reaction area before the air blower, respectively. Two glass tubes with internal diameter of 36 mm (Tube-A hereafter) and 46 mm (Tube-B hereafter) were served as photolysis area. All the experiments were carried out at atmospheric pressure and room temperature (25 ± 2 °C) with ca. 40% relative humidity.

3.1. Characteristics of the MDELs With one end inserted into the resonant cavity, the MDEL was illuminated by high frequency of MW radiation. The filling gas in the lamp was excited to produce plasma which transferred the MW energy from the inner resonant cavity to outside to excite the whole lamp in 1– 2 s. The UV radiation emitted by MDELs was detected by an Acton VM-505 VUV monochromater. The emission spectrum and a picture of the MDELs are shown in Figs. 2 and 3 under 200 W of MW output power. The MDEL-Hg containing a mixture of mercury and argon showed mainly atomic Hg emission lines at 185 nm (6s6p(1P1)–6s6p(1S0)) and 253.7 nm (62P1–62S0) (Gross et al., 2000). The radiation at 185 nm accounted for 12% of the total UV output in this study compared with conventional mercury of 7%. The MDEL-I2 containing a mixture of iodine and krypton, emitted six atomic I emission lines at 178.3, 180.1, 183 (5p46s(4P5/2)–5p5(2P3/2)), 184.4 (5p46s(4P3/2)–5p5(2P1/2)), 187.6 (5p46s(4P1/2)–5p5(2P1/2)), 206.2 nm (5p46s(2P3/2)–5p5 (2P1/2)) and one I2 emission line at 342 nm (Gross et al., 2000; Spietz et al., 2001), with maximal emission at 206.3 nm.

2.3. Chemical analysis 3.2. Parameters of H2S photodecomposition The concentrations of H2S ([H2S]) before and after MDEL treatment were determined by a spectrophotometer (Cassella et al., 1999; Santos and Korn, 2006). H2S in the gas stream was absorbed by a solution of N,N-dimethylp-phenylenediamine in acidic medium in which FeCl3 was added afterward to form methylene blue. [H2S] was then determined by measuring light absorbance at 665 nm with spectrophotometer (Model 721, Shanghai Precision & Scientific Instrument Co., Ltd.). H2S quantification was performed using a 7-point linear calibration (r = 0.9997) with diluted standards. Two parallel samples were detected and standard deviation was calculated. The standard deviation of all data is between 0.007 and 0.07 mg m3.

253.7 nm

1.0

0.8

0.6

0.0 175

313 nm

0.2

365 nm

0.4 185 nm

Intensity (a.u.)

3.2.1. Effects of H2S inlet concentration Fig. 4a show the removal efficiency of H2S (gH2 S , gH2 S = ([H2S]before  [H2S]after)/[H2S]before, the subscripts denote whether the sampling position is before or after the photolysis area) as a function of inlet H2S concentration at fixed gas flow rate of 0.14 standard l s1. High efficiency of H2S decomposition by MDELs was observed (Fig. 4). gH2 S decreased from 92% to 62% for MDEL-Hg as the inlet [H2S] increases from 1.4 to 21.9 mg m3 in Tube-A at gas residence time of 4.5 s. gH2 S changed in the same manner for MDEL-I2, reaching the highest (88%) at the inlet [H2S] of 1.1 mg m3. The number and energy of

200

225

250

275

300

325

350

375

Wavelength (nm)

Fig. 2. The emission spectrum and a photo of the working microwave electrodeless mercury lamp.

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

206.2 nm

1.0

0.2

342 nm

187.6 nm

0.4

180.1nm 183 nm 184.4 nm

0.6 178.3 nm

Intensity (a.u.)

0.8

0.0 175

200

225

250

275

300

325

350

Wavelength (nm)

Fig. 3. The emission spectrum of microwave electrodeless iodine lamp.

a b c d

100 90

Tube-A, gas residence time 4.5 s Tube-A, gas residence time 1.5 s Tube-B, gas residence time 4.5 s Tube-B, gas residence time 1.5 s

Removal efficiency (%)

80 70 a

60

b' b

50 40 d'

30

c c'

d

a' b' c' d'

a'

Tube-A, gas residence time 3.6 s Tube-A, gas residence time 1.2 s Tube-B, gas residence time 3.6 s Tube-B, gas residence time 1.2 s

20 10 0 0

5

10

15

20

25

-3

Inlet [H2S] (mg m )

Fig. 4. gH2 S as a function of initial [H2S] at different gas residence time within Tube-A and Tube-B. a, b, c, d: gH2 S achieved with MDEL-Hg; a0 , b0 , c0 , d0 : gH2 S achieved with MDEL-I2.

photons and active radicals in the reaction area did not change at a fixed input MW power. Therefore, with constant gas flow rate, as inlet [H2S] increases, unit H2S obtains less energy, which results in a lower gH2 S . MDELs shows a high efficiency for H2S removal in the presence of low [H2S], and also to some extent are capable of dealing with the variation in feed gas concentration, which implies the potential application of MDEL on controlling malodorous H2S. 3.2.2. Effects of gas residence time (gas flow rate) Experiments were then performed in Tube-A to determine the effect of gas residence time on gH2 S with MDEL-Hg (Fig. 4a – Tube-A, b – Tube-A) and MDELI2 (Fig. 4a0 – Tube-A, b0 – Tube-A), respectively. Since for the 100 cm-length of MDEL-Hg and the 80 cm-length of MDEL-I2 only 90 cm and 72 cm were inserted into the reaction area, respectively, the photolysis volumes of the reactors for calculations were 633.0 and 506.4 cm3. The gas residence time thus varied from 1.5 to 4.5 s as

1777

the gas flow rate changed from 0.42 to 0.14 standard l s1 in MDEL-Hg, and varied from 1.2 to 3.6 s in MDEL-I2. Similar trends were observed for MDEL-Hg and MDEL-I2, which are present in Fig. 4. It is observed that with increasing of gas residence time, the removal efficiency increases slightly. It is probably because the longer gas residence time provides both longer collision time and higher collision possibility, which enhances the decomposition of H2S. However, gH2 S did not increase much as expected with longer gas residence time provided (tripled). It is possibly because the life time of radicals is nanoseconds, it is much shorter than the gas residence time (1.2–4.5 s) of H2S. Therefore, the reaction of H2S with radicals is not limited kinetically by gas residence time. With 1.2 s gas residence time and 20 mg m3 inlet [H2S], gH2 S could reach as high as 50% in Tube-A. 3.2.3. Effects of the radial UV radiation diffusion In industrial wastes treatment plants, larger reaction area is always desirable in order to treat more wastes per unit time at an acceptable removal efficiency. Therefore, we designed another reaction system, Tube-B, which has a larger diameter compared with Tube-A. Experiments were conducted in Tube-B to determine the effects of radial diffusion of UV radiation on H2S removal under the same reaction condition mentioned in Sections 3.2.1 and 3.2.2. As shown in Fig. 4, gH2 S decreased dramatically with increasing tube diameter. Within Tube-B, gH2 S decreased to about 40% as inlet [H2S] increased to 5 mg m3. Compared with N2 whose absorption band is lower than 150 nm, and CO2 whose absorption coefficient is with magnitude of 103 atm1 cm1 in the region of 185–206 nm, H2S has a relative strong absorption coefficient of ca. 0.1 mm Hg1 cm1 (Hideo, 1978). UV radiation decreases radially within the reaction tube because of the absorbing of photon by H2S gas. H2S molecule is prone to be decomposed closer to the lamp wall, and those H2S molecular far from the lamp wall are difficult to absorb photon and thus with less possibility to be decomposed. Although we employed a gas mixture to homogenize H2S/gas mixture, we could not guarantee the uniformity of radial UV radiation. That results in decreasing of gH2 S . It is no doubt that enhancing tube diameter would decrease the gH2 S . The absolute removal amount (ARA) and the energy yield (EY) were used in this study to evaluate the energy efficiency of MDELs in H2S removal. ARA equals to C 0  gH2 S  Q (C0 is inlet [H2S], mg m3; Q is gas flow rate, standard l s1) with the unit of lg s1. EY (mg kW h1), a well-accepted parameter to indicate the energy efficiency, is calculated by 3600ARA/P (P is overall input power, 200 W). The relative ARA and EY with respect to gas residence time of 4.5/1.5 s with MDEL-Hg, and 3.6/1.2 s with MDEL-I2, within both Tube-A and Tube-B, are listed in Tables 1 and 2, respectively. With higher inlet [H2S], gH2 S decreased, on the contrary, ARA and EY increased significantly. It is because with fixed input MW power, H2S would not be fully decomposed due to the limited amount

1778

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

Table 1 ARA and EY as function of initial H2S concentration with different gas residence time with MDEL-Hg

Table 2 ARA and EY as function of initial H2S concentration with different gas residence time with MDEL-I2

Gas residence time (s)

Gas residence time (s)

Tube-A 4.5

1.5

Tube-B 4.5

1.5

Inlet [H2S] (mg m3)

ARA (lg s1)

EY (mg kW h1)

1.4 3.3 5.3 6.7 10.4 14.1 16.5 21.9

0.19 0.40 0.64 0.75 1.08 1.27 1.56 1.93

3.3 7.2 11.5 13.5 19.5 22.8 28.0 34.7

1.6 4.1 6.3 9.9 11.2 14.1 19.5

0.61 1.46 2.14 3.04 2.90 3.26 4.30

10.9 26.3 38.6 54.7 52.2 58.6 77.3

0.9 1.3 3.6 4.3 5.0 11.7 12.4 18.05

0.22 0.25 0.39 0.44 0.49 0.57 0.47 0.50

3.9 4.5 7.0 7.9 8.8 10.2 8.4 9.0

0.6 1.8 4.4 6.7 11.1 17.3

0.30 0.56 0.96 1.22 1.36 1.42

5.4 10.0 17.2 22.0 24.4 25.5

of photons and radicals if the inlet [H2S] is higher than the threshold. However, more H2S molecules will enhance the utilization of photons and radicals, which thus results in higher ARA and EY. Tables 1 and 2 also show that gas residence time and tube diameter affect ARA and EY remarkably. At gas residence time of 4.5 s with MDEL-Hg, and 3.6 s with MDEL-I2, gas flow rate of Tube-A (0.14 standard l s1) was about half of that of Tube-B (0.27 standard l s1), but ARA and EY of Tube-A were about 3 times to Tube-B. On the other hand, prolonging gas residence time and/or reducing tube diameter would decrease gas flow rate and increase energy consumption. Therefore, in the practical application of MDELs, appropriate gas residence time and tube diameter should be considered together with initial waste concentration, required removal efficiency, target removal amount, apparatus capability and capital investment. MDELs placed in parallel modes would be another suggestion to benefit both technically and economically. Notably, we achieved an EY of ca. 80 mg kW h1 with inlet [H2S] of 20 mg m3 by MDELs, while Ma et al. (2006) achieved a maximal removal rate of 667 mg d1 with

Tube-A 3.6

1.2

Tube-B 3.6

1.2

Inlet [H2S] (mg m3)

ARA (lg s1)

EY (mg kW h1)

1.1 3.0 5.6 9.6 11.7 14.8 18.0 22.6

0.14 0.36 0.65 1.08 1.30 1.57 1.63 1.80

2.5 6.4 11.7 19.3 23.3 28.2 29.4 32.3

1.3 4.1 5.9 8.5 14.3 15.8 18.9

0.49 1.42 1.96 2.62 4.13 4.24 4.48

8.8 25.5 35.2 47.2 74.2 76.3 80.5

1.2 2.1 2.7 5.2 7.5 11.2 14.0 18.3

0.29 0.30 0.35 0.45 0.53 0.62 0.46 0.44

5.1 5.3 6.3 8.1 9.5 11.1 8.3 7.9

0.9 1.1 2.1 5.2 9.3 12.5 18.1

0.52 0.53 0.77 1.14 1.43 1.26 1.37

9.2 9.5 13.8 20.6 25.7 22.6 24.7

activated carbon bioreactor to eliminate H2S. Comparing with activate carbon bioreactor, MDEL is economically favorable in terms of energy efficiency and operating expense because no regular replacement and regeneration of activated carbon is required. Therefore, MDEL photolysis could be a new alternative technique for odor gas removal, in consideration of the fairly good H2S removal efficiency we achieved with the short gas residence time of 1.5 s and the low energy consumption. 3.3. Photolysis products analysis The mechanism of photodissociation of H2S has been previously studied (Brad et al., 1989; Canela et al., 1998; Liu et al., 1999). It is believed that the primary photolysis in UV region produces H + HS. HS can be generated by absorbing hm with energy higher than 385.92 kJ mol1 (Chang and Tseng, 1996), corresponding to a wavelength of ca. 309 nm, which means UV radiation with wavelength shorter than 309 nm could break HS–H if it is absorbed by H2S molecule. The absorption coefficient of H2S is with the

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

1779

Table 3 The concentration of SO2 4 analyzed by IC MDEL Air H2S–air H2S–air

Hg/Ar Hg/Ar I2/Kr

Inlet [H2S] (mg m3) 0 16.7 18.0

Sample (l)

H2 S þ hm= O= OH ! HS þ H !    ! Final products Canela et al. (1998) confirmed the photolysis products of H2S are mainly SO2 4 . We employed ion chromatograph (IC) to analysis photolysis products and their concentrations. Gas products were sampled into 10 mL distilled water after photolysis with MDEL-Hg and MDEL-I2 within Tube-A at a gas flow rate of 0.14 standard l s1. Meanwhile, products of photolysis of pure air were also sampled into 10 mL distilled water as a blank sample. The gas products in samples were then identified and quantified by the IC technique. The species of sulfate is supposed to be H2SO4 in outlet gas, since no other cations were detected in the sample by using IC. The results are present in Table 3. Under the above experimental condition, gH2 S obtained at inlet [H2S] of 16.7 mg m3 is ca. 65% with MDEL-Hg, at inlet [H2S] of 18.0 mg m3 is ca. 64% with MDEL-I2, according to the experimental results shown in Fig. 4a and a0 . From Table 3, we can conclude that: (1) the concen2 tration of SO2 mg m3) is at a neglected level 4 (9.5  10 in pure air gas stream after passing photolysis area, and thus not considered in H2S-contained gas stream; (2) SO2 4 –S accounts for ca. 81% of H2S destructed by MDEL-Hg; (3) SO2 4 –S accounts for ca. 51% of H2S destructed by MDEL-I2. Photodecomposition of H2S via MDELs is quite complete, and the main product is SO2 4 . Sn =SOx =SO2 3 maybe formed as intermediates during photolysis (Kovalenko et al., 2001; Kataoka et al., 2005). These species are less harmful and are easy to be removed. They can be further decomposed to SO2 4 by extended discharge through the following reactions: 2 Sn =SOx =SO2 3 þ H2 O=O2 !    ! SO4

2

9.5  10 25.0 16.6

10 0.6 0.6

magnitude of 0.1 mm Hg1 cm1 in the region of 180– 206 nm and with the magnitude of 0.001 mm Hg1 cm1 at 253.7 nm, with an absorption cross-sections of ca. 5  1018 cm2 (Hideo, 1978). Thus 183/253.7 nm irradiated by MDEL-Hg and 178.3/180.1/183/184.4/187.6/206.2 nm irradiated by MDEL-I2 are believed to conduct initial photodissociation of H2S. Another possible mechanism that cannot be ruled out is the indirect oxidation of H2S by rad   icals such as O 2 =HO2 =O =OH , which are produced by O2/H2O etc. absorbing energetic hm (Hideo, 1978; Canela et al., 1998; Zuo et al., 2003). These active radicals can also be absorbed and further to destruct H2S molecule. As such, the photolysis reaction of H2S with MDELs is likely to be as follows:

3 SO2 4 (mg m )

SO2 4 –S=H2 S–S destructed (%) – 81 51

4. Conclusions (1) Low concentration of H2S can be removed effectively by MDELs. In the tube with 36 mm inner diameter, removal efficiency of H2S (initial concentrations ranged from 0 to 25 mg m3) is higher than 50% at gas residence time as short as 1.5 s. Shorten gas residence time confers reduced equipment space, which consequently reduces capital investment. (2) Within Tube-A (36 mm of inner diameter), when the gas flow rate is 0.42 standard l s1, absolute removal amount of H2S photodecomposed by MDEL-Hg is 4.30 lg s1, and energy yield is 77.3 mg kW h1 with initial [H2S] of 19.5 mg m3, while the absolute removal amount is 4.48 lg s1 and energy yield is 80.5 mg kW h1 by MDEL-I2 with initial [H2S] is 18.9 mg m3. This excellent performance of energy efficiency suggests that MDELs would be an alternative technology to remove H2S from those places where low concentration of H2S poses a problem. Parallel multi-lamps are recommended to improve the removal efficiency and increase energy yield simultaneously. (3) The main photolysis product of low concentration H2S via MDELs is SO2 4 , indicating that the MDEL technique can decompose H2S to less harmful products which are also easier to be treated.

References Ai, Z.H., Yang, P., Lu, X., 2005. Degradation of 4-chlorophenol by a microwave assisted photocatalysis method. J. Hazard. Mater. 124, 147–152. Al-Shamma’a, A.I., Pandithas, I., Lucas, J., 2001. Low pressure microwave plasma ultraviolet lamp for water purification and ozone production. J. Phys. D: Appl. Phys. 34, 2775–2781. Brad, R.W., Harold, B.L., James, J.V., Baronavski, A.P., 1989. Ultraviolet photodissociation dynamics of H2S and D2S. J. Chem. Phys. 90, 1403–1414. Canela, M.C., Rosana, M.A., Wilson, F.J., 1998. Gas-phase destruction of H2S using TiO2/UV–VIS. J. Photochem. Photobiol. A 112, 73– 80. Cassella, R.J., Oliveira, L.G., Santelli, R.E., 1999. On line dissolution of ZnS for sulfide determination in stabilized water samples with zinc acetate, using spectrophotometry by methylene blue method. Spectrosc. Lett. 32, 469–484. Chang, M.B., Tseng, T.D., 1996. Gas-phase removal of H2S and NH3 with dielectric barrier discharges. J. Environ. Eng. 122, 41–46. De Zwart, J.M.M., Keunen, J.G., 1992. C1-cycle of sulfur compounds. Biodegradation 3, 37–59.

1780

L.-Y. Xia et al. / Chemosphere 71 (2008) 1774–1780

Gao, Z.Q., Yang, S.G., Ta, N., Sun, C., 2007. Microwave assisted rapid and complete degradation of atrazine using TiO2 nanotube photocatalyst suspensions. J. Hazard. Mater. 145, 424–430. Gross, U., Ubelis, A., Spietz, P., Burrows, J.P., 2000. Iodine and mercury resonance lamps for kinetics experiments and their spectra in the far ultraviolet. J. Phys. D: Appl. Phys. 33, 1588–1591. Hassal, S.B., Ballik, E.A., 1991. Observation of continuous KrCl excimer fluorescence in microwave discharge employing binary mixtures of Kr and Cl2. Can. J. Phys. 69, 699–701. Hideo, O., 1978. Photochemistry of Small Molecules. John Wiley, New York, p. 168, 204, 208. Hong, J., Sun, C., Yang, S., Liu, Y., 2006. Photocatalytic degradation of methylene blue in TiO2 aqueous suspensions using microwave powered electrodeless discharge lamps. J. Hazard. Mater. 133, 162–166. Horikoshi, S., Hidaka, H., Serpone, N., 2002. Environmental remediation by an integrated microwave/UV-illumination technique 3. A microwave-powered plasma light source and photoreactor to degrade pollutants in aqueous dispersions of TiO2 illuminated by the emitted UV/Visible radiation. Environ. Sci. Technol. 36, 5229–5237. Horikoshi, S., Hidaka, H., Serpone, N., 2004. Environmental remediation by an integrated Microwave/UV illumination technique VI. A simple modified domestic microwave oven integrating an electrodeless UV– Vis lamp to photodegrade environmental pollutants in aqueous media. J. Photochem. Photobiol. A 161, 221–225. Iwaguch, S., Matsumura, K., Tokuoka, Y., Wakui, S., Kawashima, N., 2002. Sterilization system using microwave and UV light. Colloid. Surface. B 25, 299–304. Kataoka, S., Lee, E., Tejedor, I.T., Anderson, M.A., 2005. Photocatalytic degradation of hydrogen sulfide and in situ FT-IR analysis of reaction products on surface of TiO2. Appl. Catal. B-Environ. 61, 159–163. Kerst, C., Lein, I., Potzinger, P., 1998. Gas-phase photolysis of pentamethyldisilane at 206 nm. J. Photochem. Photobiol. A 113, 9–21. Kla´n, P., Litera´k, J., Ha´jek, M., 1999. The electrodeless discharge lamp: a prospective tool for photochemistry. J. Photochem. Photobiol. A 128, 145–149. Koda, S., Koga, K., Takizawa, K., Ihara, Y., Takami, A., 2001. Photodissociation of H2S doped in low temperature rare gas solids under UV irradiation. Chem. Phys. 274, 283–289. Kovalenko, O.N., Kundo, N.N., Kalinkin, P.N., 2001. Kinetics and mechanism of low-temperature oxidation of H2S with oxygen in the gas phase. React. Kinet. Catal. Lett. 72, 139–145. Limbeck, A., 2006. Microwave-assisted UV-digestion procedure for the accurate determination of Pd in natural waters. Anal. Chim. Acta 575, 114–119.

Liu, X., Hwang, D.W., Yang, X.F., Harich, S., Lin, J.J., Yang, X., 1999. Photodissociation of hydrogen sulfide at 157.6 nm: observation of SH bimodal rotational distribution. J. Chem. Phys. 111, 3940–3945. Martinez-Haya, B., Zapater, I., Quintana, P., Menendez, M., Verdasco, E., Santamaria, J., Banares, L., Aoiz, F.J., 1999. Photodissociation of dimethyl sulfide at 227.5 nm: resonance-enhanced multiphoton ionization of the methyl fragment. Chem. Phys. Lett. 311, 159–166. Ma, Y.L., Zhao, J.L., Yang, B.L., 2006. Removal of H2S in waste gases by an activated carbon bioreactor. Int. Biodeter. Biodegr. 57, 93–98. Muller, P., Loupy, A., Klan, P., 2005. The electrodeless discharge lamp: a prospective tool for photochemistry Part 6. Photochemistry of valerophenone and 4-nitroanisole in high-temperature water. J. Photochem. Photobiol. A 172, 146–150. Pandithas, I., Brown, K., Al-Shamma’a A.I., Lucas, J., Lowke, J.J., 2003. Biological applications of a low pressure microwave plasma UV lamp. In: 14th IEEE International Pulsed Power Conference, Digest of Technical Papers, pp. 1112–1115. Rogers, L.J., Ashfold, M.N.R., Matsumi, Y., Kawasaki, M., Whitaker, B.J., 1996. Photolysis of CH3SH and H2S at 243.1 nm studied by photofragment ion imaging. J. Chem. Soc. Faraday Trans. 92, 5181– 5183. Santos, J.C.C., Korn, M., 2006. Exploiting sulphide generation and gas diffusion separation in a flow system for indirect sulphite determination in wines and fruit juices. Microchim. Acta 153, 87–94. Spietz, P., Gross, U., Smalins, E., Orphal, J., Burrous, J.P., 2001. Estimation of the emission temperature of an electrodeless discharge lamp and determination of the oscillator strength for the I(2P3/2) 180.038 nm resonance transition. Spectrochim. Acta B 56, 2465–2478. Wilson, S.H.S., Howe, J.D., Ashfold, M.N.R., 1996. On the near ultraviolet photodissociation of hydrogen sulphide. Mol. Phys. 88, 841–858. Zhang, J., Boyd, I.W., 2000. Multi-wavelength excimer ultraviolet sources from a mixture of krypton and iodine in a dielectric barrier discharge. Appl. Phys. B 71, 177–179. Zhang, X.W., Li, G.T., Wang, Y.Z., Qu, J.H., 2006a. Microwave electrodeless lamp photolytic degradation of acid orange 7. J. Photochem. Photobiol. A 184, 26–33. Zhang, X.W., Li, G.T., Wang, Y.Z., 2006b. Microwave assisted photocatalytic degradation of high concentration azo dye Reactive Brilliant Red X-3B with microwave electrodeless lamp as light source. Dyes Pigments 74, 536–544. Zuo, L., Hou, L., Yang, L.S., 2003. The experimental investigations of dielectric barrier discharge and pulse corona discharge in air cleaning. Plasma Sci. Technol. 5, 1961–1964.