Download File 1

Toxicological & Environmental Chemistry, Jan.–Dec. 2005; 87(1–4): 509–519

Gliding arc plasma processing for decomposition of chloroform

HYUNG KEUN SONG, JAE-WOOK CHOI, HWAUNG LEE, & ANTONIUS INDARTO Korea Institute of Science & Technology, Clean Technology Research Center, PO Box 131, Cheongryang, Seoul 130-650, Korea (Received 8 October 2004; revised 27 January 2005; accepted 23 September 2005)

Abstract The decomposition of chloroform (CHCl3) diluted in air was studied. The experiment was carried out by using a gliding arc plasma. Different values of initial concentrations of chloroform, total gas flow rates, and input power frequencies have been used to investigate this effects on the conversion reaction products both qualitatively and quantitatively. Experimental results indicate that the maximum conversion of chloroform was 97% at a total gas flow rate of 180 L h
1 containing 1% chloroform. Using air as carrier gas, decomposition of CHCl3 produces CCl4, CO2, CO, and Cl2 as the main products. Small amounts of HCl and COCl2 are also detected. Liquid products were also produced. Keywords: Plasma, gliding arc plasma, chloroform, decomposition

Introduction Chloroform (CHCl3) is classified as toxic and carcinogenic and it is an important task to remove it from indoor or outdoor ambient atmosphere [1,2]. Chloroform is used as a solvent in the production of various chemicals, and existed in some pesticide formulations. It is suspected to have a global warming potential (GWP) and to contribute to the destruction of the ozone layer in the stratosphere [3]. Two approaches are generally used to remove this pollutant from air, i.e. by thermal or catalytic oxidation. Several procedures of thermal oxidation have been developed and established [4, 5]. This process requires high burning temperature (between 760 and 980 C) and relatively long residence time to complete the oxidation reaction [4]. To avoid these problems, many researchers have proposed the use of catalysts. Although this could reduce the duration of residence time, the removal efficiency was found to be lower

Correspondence: Hyung Keun Song, Korea Institute of Science & Technology, Clean Technology Research Center, PO Box 131, Cheongryang, Seoul 130-650, Korea. E-mail: [email protected] ISSN 0277-2248 print/ISSN 1029-0486 online # 2005 Taylor & Francis DOI: 10.1080/02772240500382456

510

A. Indarto et al. Vent to atmosphere GC-FID bentone MFC-1 GC-TCD SK carbon Compressed air

Quadrupole mass spectrometer

Scrubber MFC-2

Plasma zone CHCl3

Capillary tube

Electrode

Water bath

Kl solution Plasma reactor

Figure 1. Schematic diagram of experimental setup.

than by thermal oxidation. At low temperatures, the catalyst’s lifetime is short [6] and varies with different catalysts when the target material consists of different VOC species. To overcome these problems, gliding arc discharge has a bright prospect, especially for industrial scale application. Gliding arc discharge is the subject of renewed interest for applications to various chemical reactions [7]. Therefore, the main recent improvements in utilizing gliding arc discharge are more powerful electrical discharges under thermal non-equilibrium conditions [8], short resident time, and concentration capacity. Kinetic destruction of chloroform using photosensitized oxidation [9], thermal oxidation [10], or photocatalysis [11, 12] have been studied previously but, unfortunately, the reaction pathway under gliding arc discharge condition was not studied in detail. In the present work, the reaction mechanism of decomposition of CHCl3 diluted in compressed air was studied by gliding arc discharge. The main products CO, CO2, Cl2, and CCl4 were the objects of analysis. Experimental setup A diagram of experimental setup is shown in Figure 1. Chloroform and compressed air are used as input gas. The setup is described in detail in the following section.

Plasma reactor and applied power system The reactor was made from a quartz–glass tube of an inner diameter of 45 mm and a length of 300 mm. Top and bottom of the reactor are equipped with teflon seals, the lower comprising two electrodes made of stainless steel, length of the electrodes being 100 mm. The distance of the electrodes in the narrowest section is 1.5 mm. The gas mixture is fed between the electrodes with a capillary of 0.8 mm inner diameter. A thermocouple,

Gliding arc plasma processing for decomposition of chloroform

511

6000

4000

Voltage (V)

2000

0

−2000

−4000

−6000 0.4

Current (A)

0.2

0.0

−0.2

−0.4

−0.6

−2e-5

−1e-5

0

1e-5

2e-5

Time (s)

Figure 2. Voltage and current profile.

located 10 cm above the electrode, is provided to measure the outlet gas temperature. A high frequency AC power supply (Auto electric, A1831, Korea) is connected to the gliding arc electrode to generate the plasma. In this experiment, the total power can not be adjusted; it is a function of the breakdown voltage and changes into stable equilibrium automatically when the arc is formed. The supplied power is calculated for equilibrium condition when the plasma has been operated for 30 min. Figure 2 shows the typical waveforms of voltage and discharge current used in these experiments.

512

A. Indarto et al.

Input gas Chloroform (CHCl3) (Junsei Chemical Co., Japan) having a purity of 99.0%, is varied in its concentration to 1, 3, 5, and 8% volume of total gas flow. Air is used as a carrier gas and is controlled by a mass flow controller (Tylan, FC-280S) at flow rates of 180, 240, and 300 L h
1. The air entering the reactor first passes through a scrubber and is mixed with CHCl3. The CHCl3 is introduced by passing a portion of the air through a bubble flash placed in a water bath. The concentration of CHCl3 is controlled both by the temperature of the water bath and the gas flow rate through the flask. The input lines are heated by means of heating tape to avoid the condensation. The composition of the mixture is analyzed at the outlet reactor with the plasma operation on and off.

System of measurements The outlet gas composition is examined by an electron-impact ionization quadrupole mass spectrometer (Balzers, QMS 200) with software Quadstar 421. For the qualitative and quantitative analysis of reactants and products, two GC’s are carried out. Chlorinated methane compounds in the product gas mixture are qualitatively determined by GC-FID (YoungLin, M600D, Korea, Column: 5 m  2.5 cm i.d. Bentone) and for CO and CO2 by GC-TCD (YoungLin, M600D, Korea, Column: 5 m  2.5 cm i.d. SK Carbon). Chlorine (Cl2) gas is determined by bubbling the reacting gas through 0.05 M aqueous KI for a given experiment time, followed by iodometric titration with 0.05 M Na2SO3 [13]. The evaluation of system as to its performance, selectivity and conversion is formulated as: Selectivity of ðCO þ CO2 Þ ¼

Selectivity of Cl2 ¼

Conversion of chloroform ðÞ ¼

moles of (CO þ CO2 Þ produced  100% moles of chloroform converted

ð1Þ

2  moles of Cl2 produced  100% 3  moles of chloroform converted

ð2Þ

moles of chloroform converted  100% moles of initial chloroform

ð3Þ

These parameters are used to study the effect of initial CHCl3 concentration, total gas flow rate, and input power frequency. The supplied power is shown as a function of voltage and current determinated by oscilloscope (Agilent, 54641A, USA) and calculated to power term by the following equation: Z Supplied power (W ) ¼ ðV ðtÞ  IðtÞÞdt  frequency ð4Þ In this study the experimental data were taken 30 minutes after the initiation of the gliding arc plasma referred to the stable temperature of the outlet bulk gas measured by thermocouple. Results and discussion With the gliding arc plasma, the destruction of chloroform (CHCl3) produces CO2, CO, Cl2, and CCl4, as the main products. The yields of CO and Cl2 are higher than that of CO2

Gliding arc plasma processing for decomposition of chloroform

513

and CCl4. The concentration of CO2 is 3–5 times lower than the concentration of CO. CCl4 is formed at yields between 8 and 15% by mol basis. HCl and COCl2, are also detected by QMS, but in low concentration. The destruction performance was studied as functions of initial CHCl3 concentration, total gas flow rate, and input frequency. The gaseous product analysis is focused on CO, CO2, and Cl2 compounds because these are the main products and can be measured precisely.

Effect of initial concentration Figure 3(a) shows the effect of various initial concentrations of chloroform (CHCl3) on its conversion at an input power frequency of 20 kHz. The maximum conversion reached 97% at the lowest concentration of chloroform, at 1%, and a total gas flow rate of 180 L h
1. Conversion of chloroform gradually decreases when the initial concentration increases. The decrement rate reaches 2.5% per 1% of increase of initial chloroform concentration in the inlet stream. One factor that contributes to the decrease is the power supplied to the reactor which decreases when the initial concentration of chloroform increases (Figure 3(b)). Thus, less energy is supplied for initiation of gas molecules. The supplied power is highly dependent on the breakdown process producing the arc by breaking bonds or exciting molecules [8]. Chloroform has less strong chemical bonds compared to components such as O2 and N2 that are mostly existing in the inlet stream. It means that the required energy to initiate the first arc plasma is also less. Higher concentrations of chloroform decrease the required breakdown energy which decreases the supplied power automatically. Therefore, as mentioned above, the stabilized equilibrium power is automatically adjusted by the system. Different values of initial concentration of chloroform have significant effects on the energy efficiency of the system. The energy consumed per converted chloroform is rising from 18.6 to 114 Watt-h L
1 when the concentration of chloroform is decreasing from 8% to 1%. At a higher initial concentration of chloroform the collisions are more effective although the percentage of conversion is lower. Small increments of gas temperature due to increasing initial chloroform concentration are measured, reaching 250 C and rising 2–3 C per 1% increment of initial chloroform. It reflects the energy released from the exothermic plasma reaction and calculated following formula 5 and 6. k
j
X
   _ ¼ m _  Cp products  Tg,t
Tg,0 _  Cp un-reacted species þ m ð5Þ Q _ _
H reaction ¼ Q
Wpower supply

ð6Þ

When the plasma reaction occurs, a portion of chemical energy is converted to thermal energy and raises the temperature. Compared to the supplied electrical energy, exothermic plasma reaction constitutes 70–80% of the energy increasing the product gas temperature (calculated using formula 5 and 6). Higher initial concentrations of chloroform produce more heat. By calculating the converted chloroform, the ratio leads to 97.03 (1%  180 L h
1  97.03%) and 274.44 (3%  180 L h
1  91.48%) for the chloroform concentration at 1% and 3% and total gas flow rate 180 L h
1. This means that the molecule collision probability at higher concentration of chloroform, i.e. 3%, is approximately 3 times higher than that at lower concentration, i.e. 1%.

514

A. Indarto et al.

CHCl3 conversion (%)

(a)

CHCl3 conversion

100 80 60

180 L h−1 240 L h−1 300 L h−1

40 20

(b)

300

Supplied power

Power (Watt)

280 260 240 220 200

180 L h−1 240 L h−1 300 L h−1

180 160

Molar selectivity (%)

(c)

60

(CO + CO2)

50 40 30 180 L h−1 240 L h−1 300 L h−1

20 10

Molar selectivity (%)

(d)

60

Cl2

50 40 30 180 L h−1 240 L h−1 300 L h−1

20 10 0

2

4 6 Initial CHCl3 concentration (%)

8

10

Figure 3. Effect of initial chloroform concentration on (a) chloroform conversion, (b) supplied power, (c) yields of (CO þ CO2), and (d) of Cl2.

Gliding arc plasma processing for decomposition of chloroform

515

The yields of gas products, (CO þ CO2) and Cl2, is shown in Figure 3(c) and (d). No significant differences in yields of (CO þ CO2) are observed at different initial concentrations of chloroform. The maximum yield of (CO þ CO2) is around 50%. It shows that almost half fraction of reaction produced CO or CO2. Conversion into CCl4 was detected not exceed than 15% in all experimental conditions. The remaining carbon possibly transforms into solid (shoot) or liquid products which are also produced during experiment. In the case of Cl2, the yields increase slightly with increasing chloroform concentration. Oxygen and nitrogen in the input stream should be considered as factors for determining the product selectivity. Oxygen and nitrogen can be transformed into high-energy state species or radicals by collision. Calculated from mass spectra data, the difference of oxygen concentration before and after the plasma reaction approaches 10% and for nitrogen 5%, independent of the initial chloroform concentration. The role of radical species, such as oxygen and oxygen radical to maintain the decomposition process, has been studied before [14, 15].

Effect of total gas flow rate The effect of total gas flow rate, determining the residence time of chloroform in the reactor, was examined. Figure 4(a) shows the chloroform conversion according to the various total gas flow rates at initial chloroform concentrations of 1, 3, 5% and power supply frequency of 20 kHz. The conversion of chloroform decreases with the increase of total flow rate. Raising the total gas flow rate reduces the residence time of chloroform in the reactor and the chance and time of a molecule to collide with electrons and other high-energy state species which have enough energy to destroy carbon–hydrogen or carbon–chloride bonds. Lower conversion is also caused by decreasing power as gas flow rate increased (Figure 4b). This phenomenon has been observed previously [16]. In terms of energy efficiency the effect of total gas flow rate is rather smaller. The consumed energy per converted chloroform decreases from 149 to 122 Watt-h L
1 when the flow rate increases from 180 to 300 L h
1. With increasing of total gas flow rate (Figure 4(b)), the (CO þ CO2) yield increases, Figure 4(c), while the yield of Cl2 tends to decrease.

Effect of applied frequency Decomposition of chloroform increases slightly from 92 to 97% when the frequency of input power is increased from 15 to 20 kHz at 180 L h
1 (Figure 5a). The same trend is observed at 300 L h
1 with yields between 42 and 60%. Radu et al. [17] mention that a change in frequency will change the basic breakdown mechanism. Higher frequency will increase the extent of sudden-fluctuating pulsed current and voltage peak per cycle. Furthermore, it will increase the total power supplied to the plasma (Figure 5b) which provides more energy for the gas ionization process. However, the ratio of power consumed versus converted chloroform is increased from 92 Watt-h L
1 at 15 kHz to 142 Watt-h L
1 at 20 kHz. It means that at lower frequency, although the conversion of chloroform is lower, the energy efficiency is better that at higher frequency. Increasing frequency decreases the yields of (CO þ CO2) (Figure 5c). However, the yield of Cl2 increases from 18 to 25% at 180 L h
1 and achieves a maximum of 38% at a total gas flow rate 300 L h
1 (Figure 5d).

516

A. Indarto et al. (a) CHCl3 conversion

CHCl3 conversion (%)

100

80

60 1% 3% 5%

40

(b)

Power (Watt)

300

Supplied power

250

200 1% 3% 5%

150

Molar selectivity (%)

(c)

60

(CO + CO2)

50 40 30 20

1% 3% 5%

10

Molar selectivity (%)

(d)

60

Cl2

50 40 30 20 10 160

1% 3% 5% 180

200

220

240

Total gas flow rate

260

280

300

320

(L h−1)

Figure 4. Effect of total gas flow rate on (a) chloroform conversion, (b) supplied power, (c) yields of (CO þ CO2), and (d) of Cl2.

Gliding arc plasma processing for decomposition of chloroform

CHCl3 conversion (%)

(a)

CHCl3 conversion

100 80 60 40

180 L h−1 300 L h−1

20

Power (Watt)

(b)

517

Supplied power

300 250 200 150

180 L h−1 300 L h−1

100

(c)

(CO + CO2)

Molar selectivity (%)

80

60

40 180 L h−1 300 L h−1

20

Molar selectivity (%)

(d)

60

Cl2

50 40 30 20

180 L h−1 300 L h−1

10 14

15

16

17

18

19

20

21

Frequency (kHz)

Figure 5. Effect of applied power frequency on (a) chloroform conversion, (b) supplied power, (c) yields of (CO þ CO2), and (d) of Cl2.

518

A. Indarto et al. 1.2e-9 1.0e-9

Intensity

8.0e-10 6.0e-10 4.0e-10 COCl2

2.0e-10

0

20

40

60

CHCl3 Cl2

CCl4

80

120

100

140

160

180

200

AMU (m/z)

Figure 6. Mass spectrum of chloroform decomposition (taken at applied frequency 20 kHz, 1% of chloroform, total gas flow rate 240 L h
1).

Mass spectra A mass spectrum of the chloroform decomposition mixture is shown in Figure 6. The main þ fragment of chloroform is m/z 82/84/86 (CClþ 2 ). CO2 has its molecular ion at m/z 44 (CO2 ) þ þ but overlaps with N2O (N2 O ) which may form as gas product. CO is at 28 (CO ) but overlaps minor N2 spectra 29 (Nþ 2 ). Some additional peaks are m/z 63/65, for COCl2 þ (COClþ ), m/z 70/72 for Cl2 (Clþ 2 ), m/z 117 for CCl4 (CCl3 ).

Conclusions The performance of chloroform (CHCl3) conversion in gliding arc plasma at atmospheric pressure, in respect to initial concentration of chloroform, total gas flow rate, and input power frequency, was studied. Gliding plasma generates enough active species to fragment chloroform molecules. The maximum conversion of 97% is achieved for a feed gas stream containing 1% of chloroform and a total air flow rate of 180 L h
1. Four gaseous compounds, CO, CO2, CCl4 and Cl2, are the major products. Cl2 yields are relatively high and reached maximum of 50%. Yields of CO and CO2 reach up to 78% mol at 15 kHz of frequency and total gas flow rate at 180 L h
1. Conversion into CCl4 was detected in the case of chloroform decomposition, but not more than 15%. COCl2 has an important role as intermediate species to produce Cl2 and CO. Development is still needed to remove some traces of unwanted compounds in the final product, such as: COCl2 and CCl4 which are also classified as toxic gases.

Acknowledgement This work was supported by the National Research Laboratory of the Ministry of Science and Technology of Korea.

Gliding arc plasma processing for decomposition of chloroform

519

References 1. Meek ME, Beauchamp R, Long G, Moir D, Tuner L, Walker M. J. Toxicol. Environ. Health, Part B. 2002;5:283. 2. Toxic Release Inventory (TRI) Basis of OSHA Carcinogens, Technical Update, February 2005. 3. U.S. Environmental Protection Agency. Greenhouse gases and global warming potential values. April 2002. 4. Urashima K, Chang JS. IEEE Trans. Diel. Elect. Ins. 2000;7(5):602. 5. Ryan JV, Lemieux PM, International Incineration Conference, Savana GA. May 6–10, 1996. 6. Lahousse C, Bernier A, Grange P, Delmon B, Papaefthimiou P, Ioannides T, Verykios X. J. Catal. 1998;178:214. 7. Czernichowski A. Pure & Appl. Chem. 1994;66(6):1301. 8. Fridman A, Nester S, Kennedy LA, Saveliev A, Mustaf-Yardimci O. Prog. Energy Combust. Sci. 1999;25:211. 9. Olbregts J. J. Photochem. 1980;14:19. 10. Lou JC, Chang YS. Combust. Flame. 1997;109:188. 11. Alberici RM, Jardim WF. Appl. Catal B: Environ. 1997;14:55. 12. Feiyan C, Pehkonen SO, Ray MB. Water Res. 2002;36:4203. 13. Skoog DA, West DM, Holler FJ, Crouch SR. Analytical chemistry, an introduction. Saunders College Publishing: Philadelphia; 2000. 14. Davidson TJ, Schiff HI, Brown TJ, Howard CJ. J. Chem. Phys. 1978;69:4277. 15. Aker PM, Niefer BI, Sloan JJ, Heydtmann H. J. Chem. Phys. 1987;87(1):203. 16. Cobine JD. Gaseous Conductor Theory and Engineering Application. Dover Publications Inc. New York, 1958. pp 160–177. 17. Radu I, Bartnikas R, Wertheimer MR. IEEE Trans. Plasma Sci. 2003;31(6):1363.