8074
Langmuir 2007, 23, 8074-8078
Decomposition Reaction of Organophosphorus Nerve Agents on Solid Surfaces with Atmospheric Radio Frequency Plasma Generated Gaseous Species Seong H. Kim,*,† Jeong Hoon Kim,‡ and Bang-Kwon Kang§ Department of Chemical Engineering, The PennsylVania State UniVersity, UniVersity Park, PennsylVania 16802, APPLASMA, 111 Gibbons Street, State College, PennsylVania 16801, and APPLASMA Company, Ltd., 211 Sanhakwon, Ajou UniVersity, San 5 Wonchun-dong Yeongtong-gu, Suwon, Korea ReceiVed March 9, 2007. In Final Form: May 13, 2007 The decomposition and detoxification of compounds are of great interest in environmental protection and defenserelated areas. We report the generation of gaseous excited species by scanning atmospheric radio frequency (rf) plasma and their reactions with two representative organophosphorus nerve agents, paraoxon and parathion, deposited on solid surfaces. The excited gaseous species generated in the Ar and Ar/O2 plasma were identified as atomic oxygen, OH radical, and excited nitrogen molecule from optical emission spectroscopy analysis. The reaction of these species with paraoxon and parathion was monitored with reflection-absorption infrared spectroscopy and compared with the decomposition by UV irradiation and UV/ozone treatments. The decomposition products of the atmospheric rf plasma treatment were similar to those of the UV/ozone treatment. The atomic oxygen and excited OH species generated by the plasma appear to be responsible for the oxidation of paraoxon and parathion. The plasma-induced decomposition process was much faster and more efficient than the UV/ozone process. The complete detoxification of paraoxon and parathion upon a short time exposure to the Ar/O2 plasma was confirmed by the Drosophila melanogaster culture test.
Introduction There is a great need for plasma-based surface-cleaning processes for pesticide decontamination, chemical and biological warfare agent decontamination, biomaterial sterilization, polymer film stripping, etc.1 Among these applications, the decontamination of the surface of delicate and large equipment at the contaminated field poses a special challenge since it requires large-area treatments without arcing damages under ambient conditions. A number of atmospheric plasma systems and their applications to surface cleaning have recently been developed,2,3 but the chemical natures and reactions of atmospheric plasma generated reactive species are not well understood yet. In this paper we describe the characterization of gaseous excited species in the atmospheric radio frequency (rf) plasma and their reactions with organophosphorus (OP) nerve agents adsorbed on solid surfaces. The majority of pesticides and chemical nerve agents are OPbased molecules.4 Many OP molecules can irreversibly inhibit the acetylcholine esterase enzyme in the nervous systems of insects, animals, and humans and paralyze them. In the military, the decontamination of OP compounds is typically done by washing with bleach or ethylene glycol monomethyl ether solutions.4 However, these solution-based processes may not be applicable to delicate surfaces. The bleach solutions are highly corrosive to metals. The ethylene glycol monomethyl ether is less corrosive, but it can damage paints, plastics, and leather * To whom correspondence should be addressed. E-mail: shkim@ engr.psu.edu. † The Pennsylvania State University. ‡ APPLASMA § APPLASMA Co., Ltd. and Ajou University. (1) Chu, P. K.; Chen, J. Y.; Wang, L. P.; Huang, N. Mater. Sci. Eng. 2002, R36, 143. (2) Moon, S. Y.; Choe, W.; Kang, B. K. Appl. Phys. Lett. 2004, 84, 188. (3) Zhu, W.-C.; Wang, B.-R.; Yao, Z.-X.; Pu, Y.-K. J. Phys. D: Appl. Phys. 2005, 38, 1396. (4) Yang, Y.-C.; Baker, J. A.; Ward, J. R. Chem. ReV. 1992, 92, 1729.
materials. As an alternative, enzyme-based decontamination methods have been studied extensively.5 Organophosphorus hydrolase (OPH) is one of the most promising enzymes studied for decomposition of OP nerve agents.5 Although OPH works very efficiently against paraoxon, its activity significantly decreases for other OP molecules. For example, its hydrolysis activity decreases by 30-fold for parathion, and by ∼1000-fold for chloropyrifos.6,7 In addition, the solution-based methods generate a large quantity of waste solutions that are still toxic. Finally, these methods cannot be applied to equipment or fabric that should not be wet. Photochemical decomposition or ozonolysis has been studied for decomposition of OP nerve agents in dry conditions, but their efficiency is somewhat limited. Recently, atmospheric nonthermal plasma processes have been developed for decontamination of OP molecules. The plasma can destroy a wide spectrum of OP nerve agents as well as other chemicals and biological pathogens.8-11 However, the requirements for a high voltage for plasma generation, strict plasma generation conditions, and/or a high gas flow rate of the atmospheric plasma sources have posed some limitations in practical applications. In addition, the lack of chemistry control and the failure of expeditious and complete decomposition and detoxification has been a major hurdle to overcome. To resolve these challenges, a new atmospheric rf plasma generation source was developed that can operate at low voltage and low power and is suitable for scanning (5) Russell, A. J.; Berberich, J. A.; Drevon, G. F.; Koepsel, R. R. Annu. ReV. Biomed. Eng. 2003, 5, 1. (6) Bushway, R. J.; Fan, Z. J. Food Prot. 1998, 61, 708. (7) Cho, C. M.-H.; Mulchandani, A.; Chen, W. Appl. EnViron. Microbiol. 2002, 68, 2026. (8) Herrmann, H. W.; Selwyn, G. S.; Henins, I.; Park, J.; Jeffery, M.; Williams, J. M. IEEE Trans. Plasma Sci. 2002, 30, 1460. (9) Rahul, R.; Stan, O.; Rahman, A.; Littlefield, E.; Hoshimiya, K.; Yalin, A. P.; Sharma, A.; Pruden, A.; Moore, C. A.; Yu. Z.; Collins, G. J. J. Phys. D: Appl. Phys. 2005, 38, 17501. (10) Birmingham, J. G. IEEE Trans. Plasma Sci. 2004, 32, 1526. (11) Laroussi, M. Plasma Processes Polym. 2005, 2, 391.
10.1021/la700692t CCC: $37.00 © 2007 American Chemical Society Published on Web 06/19/2007
Decomposition of Organophosphorus NerVe Agents
Figure 1. Schematic of the scanning atmospheric rf plasma system and molecular structures of parathion and paraoxon.
over large surface areas.12,13 This plasma source is a dielectric barrier discharge system that utilizes a cylindrical electrode and a dielectric barrier surrounding the electrode. The plasma is generated in the open space between the dielectric barrier and the ground electrode and pulled to the substrate surface of interest at atmospheric pressure. In this paper we report the characterization of the reactive species in the Ar/O2 plasma generated in air and their reactions of paraoxon and parathion films deposited on metal and glass substrates. The excited gaseous species in the plasma were analyzed with optical emission spectroscopy. The decomposition reactions of parathion and paraoxon by the plasma were studied with infrared spectroscopy. It appears that atomic oxygen, excited hydroxyl radical, and ozone species are produced in the atmospheric rf plasma and are responsible for decomposition of paraoxon and parathion. The detoxification efficiency was demonstrated by comparing the survival rate of Drosophila melanogaster (fruit fly) in containers containing parathion and paraoxon films treated with the UV, UV/ozone, and plasma processes. Experimental Details
Langmuir, Vol. 23, No. 15, 2007 8075
Figure 2. Optical emission spectra of (a) Ar and (b) Ar/O2 plasma generated using the atmospheric rf plasma source (rf power 200 W). The optical emission from the excited gaseous species in the Ar and Ar/O2 plasma was analyzed using a spectrophotometer (Varian, Cary Eclipse) with a 1 nm wavelength resolution. A clean blank substrate was placed ∼5 mm below the plasma head, and the optical emission was analyzed from the side of the plasma generated between the head and the substrate. The center of the plasma was positioned at the focal point of the spectrometer. The decomposition of parathion and paraoxon films on the gold substrate was monitored with polarization-modulation reflectionabsorption infrared spectroscopy (PM-RAIRS). Gaseous reaction products were not analyzed since the treatments were done in an open space. The plasma-induced decomposition was compared with photodecomposition and ozonolysis. The collimated UV light from a 200 W high-pressure mercury lamp was used for the photodecomposition study. The ozonolysis study was done with a commercial UV/ozone cleaner (Bioforce Nanosciences). The detoxification efficiency of three different treatmentssplasma, UV, UV/ozoneswas tested by culturing D. melanogaster in a controlled container. The same number of Drosophila (20 counts) were enclosed in 300 mL glass jars (culture cells) containing fly food and slide glass test samples which were initially coated with paraoxon or parathion and treated with different cleaning methods. All Drosophila used in the test were newly hatched and not fertilized. The culture cells were sealed with a polyethylene film. and a few holes were made in the film to allow air diffusion into the cell.
The schematic of the scanning atmospheric rf plasma system is shown in Figure 1. A stainless steel cylindrical electrode (1.5 cm diameter, 12 cm long) is shielded inside a quartz tube (inner diameter 1.5 cm, wall thickness ∼2 mm) which separates the electrode from the grounded cover. The plasma was generated with Ar gas (flow rate 4000 sccm) by applying a 13.56 MHz rf electric field across the center electrode and the grounded cover. The typical operating rf power was 150-250 W. Oxygen (flow rate 20 sccm) was used as a reactive gas. The bottom of the plasma source is open to the substrate. The effective plasma area was ∼0.9 cm wide and 10 cm long. Since the plasma was operated in a glow discharge mode, it could be directly applied to metallic substrates as well as nonconducting substrates without arc or streamer damages.2 The plasma exposure to the sample was made by automated scanning of the sample under the fixed plasma source. The separation between the plasma source and the sample was ∼4 mm. The typical scan speed was varied within the 3-10 mm/s range. The substrates tested in this work were gold films deposited on Si wafers and glass slides. Onto these substrates, parathion and paraoxon were spin-coated from a 0.023 mM solution in ethanol. The loadings of parathion and paraoxon were determined with a quartz crystal microbalance (QCM). The deposited amounts of parathion and paraoxon films were ∼8.5 and ∼5.0 µg/cm2, respectively.
Characterization of Plasma-Generated Reactive Species. The plasma-generated gaseous species are typically in electronically excited states and emit characteristic wavelengths upon the electronic transition to lower electronic states. Figure 2a displays a typical optical emission spectrum of the Ar plasma produced in air.2 The characteristic emission peaks from the excited Ar molecules are clearly seen in the region from 696 to 812 nm.14 Since the plasma is generated in the air, nitrogen, oxygen, and water molecules in the air are also excited. The peaks at 337 and 674 nm originate from the C3Πu-B3Πg and B3Πg-A3Σu+ transitions of the excited N2 molecule, respectively.15 The peak at 309 nm is due to the A2Σ+-X2Π transition of the OH produced by dissociation of water vapor in the plasma.15 The peaks at 777 and 845 nm are due to the excited atomic oxygen.2 When oxygen is added into the plasma, the optical emission spectrum is drastically changed. First, the ratio of the atomic oxygen peaks to the Ar peaks increases. Second, the electronic transitions involving vibrationally excited states are observed for the OH A2Σ+-X2Π and N2 C3Πu-B3Πg peaks.15 Third, the
(12) Kim, J.-H.; Liu, G.; Kim, S. H. J. Mater. Chem. 2006, 16, 977. (13) Kim, S. H.; Kim, J.-H.; Kang, B.-K.; Uhm, H. S. Langmuir 2005, 21, 12213.
(14) Goleb, J. A. Anal. Chem. 1966, 38, 1059. (15) Pearse, R. W. B.; Gaydon, A. G. The Identification of Molecular Spectra, 4th ed.; John Wiley & Sons, Inc.: New York, 1976.
Results and Discussion
8076 Langmuir, Vol. 23, No. 15, 2007
Kim et al.
Figure 3. PM-RAIRS analysis of the surface decontamination progress for (a) plasma, (b) UV, and (c) UV/ozone treatments of parathion films coated on gold films. The plasma exposure time is the plasma width (9 mm) divided by the scan speed (3 mm/s). The UV and UV/ozone treatment time is the total exposure time over the whole surface.
Figure 4. PM-RAIRS analysis of the surface decontamination progress for (a) plasma, (b) UV, and (c) UV/ozone treatments of paraoxon films coated on gold films. The plasma exposure time is the plasma width (9 mm) divided by the scan speed (3 mm/s). The UV and UV/ozone treatment time is the total exposure time over the whole surface.
peaks corresponding to the OH B2Σ+-A2Σ+ and C2Σ+-A2Σ+ transitions grow significantly.15 The drastic increase of the atomic oxygen, OH, and N2 peaks relative to the Ar carrier gas peaks indicates a high efficiency of the O2 activation in the plasma. Although the optical emission from the excited ozone molecules (1B2-X1A1 peak at ∼322 nm) is not detected, it is very likely that the atmospheric rf plasma also produces ozone. It is wellknown that the atomic oxygen and other electronically excited OH, N2, and Ar species undergo a three-body reaction with molecular oxygen to produce ozone.16,17 The ozone species in the ground electronic state cannot be detected in optical emission spectroscopy. For the same reason, all other products that are not in the electronically excited states cannot be detected in optical emission analysis. Even if the electronically exited reaction products are produced, they cannot be detected unless their concentration in the gas phase is not high enough. OP Decomposition by Plasma-Generated Reactive Species. The UV, excited atoms and radicals, and ozone generated by the plasma can potentially decompose OP films deposited on solid surfaces. The decomposition products of parathion and paraoxon after the atmospheric rf plasma treatment were analyzed with PM-RAIRS. To find whether photolysis or oxidation is responsible for the decomposition reaction, the plasma-induced decomposition products were compared with the surface species remaining on the surface after UV and UV/ozone treatments. Although the electrons and ions in the plasma can be involved in the decomposition reaction, we did not perform any control experiment to study their effects separately since their lifetime in air is extremely short and the controlled production and dose of these charged particles in air are difficult. Figures 3 and 4 display PM-RAIRS spectra of parathion and paraoxon films deposited on the gold surface and treated with atmospheric rf plasma, UV, and UV/ozone. The vibrational peak assignments for the fresh parathion and paraoxon films were done through comparison with the literature for nitrobenzene, phospholipids, and hydroxyapatites (Table 1).18-20 The peaks at 1099, 1111, 1163, 1350, 1490, 1525, 1591, and 1614 cm-1 are due to the nitrobenzyl group. These peaks are present at the same
Table 1. Infrared Vibrational Peak Assignments for Parathion and Paraoxon18-20
(16) Eliasson, B.; Hirth, M.; Kogelschatz, U. J. Phys. D: Appl. Phys. 1987, 20, 1421. (17) Ahn, H.-S.; Hayashi, N.; Ihara, S.; Yamabe, C. Jpn. J. Appl. Phys. 2003, 42, 6578. (18) Dziri, L.; Desbat, B.; Leblanc, R. M. J. Am. Chem. Soc. 1999, 121, 9618. (19) Bocharov, S.; Teplyakov, A. V. Surf. Sci. 2004, 573, 403. (20) Clarkson, J.; Smith, W. E. J. Mol. Struct. 2003, 655, 413.
peak position (cm-1) 930 1032, 1053 1039, 1059 1099, 1111 1163 1236-1240 1282, 1296 1350 1392 1442 1490 1525 1591 1614
parathion
paraoxon
PdS C-O-P(S) C-O-P(O) ring breathing C-H bend P-O P-O PdO -NO2 stretch CH3 and CH2 bending CH3 and CH2 bending ring stretch + C-N stretch ring stretch + NO2 assymetric stretch ring stretch ring stretch + NO2 assymetric stretch
positions in both parathion and paraoxon spectra. The 1350 cm-1 peak is characteristic of the -NO2 group. Two weak peaks at 1392 and 1446 cm-1 are assigned to the C-C-H bending vibration in the ethyl group. The peaks at 1282 and 1299 cm-1 are associated with the PdO double bond of paraoxon. These peaks are not observed in parathion. In the parathion spectrum, the peak at 930 cm-1 is quite strong, which is not observed in the paraoxon spectrum, so this peak can be assigned to the PdS double bond. The peaks at 1032-1059 cm-1 are attributed to the C-O(P) ether bond. Their peak positions slightly shift depending on whether P is bonded to S or O. The peak at ∼1240 cm-1 is due to the P-O vibration. Upon a single pass of the Ar/O2 plasma (rf power 200 W, scan rate 3 mm/s) over the OP film, the nitrobenzyl and alkyl peaks disappear almost completely and broad peaks at 1055, 1245, 1650, and 1750 cm-1 (Figures 3a and 4a) appear. The peaks at 1055 and 1245 cm-1 are due to the formation of amorphous phosphorus oxide. These peaks are commonly observed in hydroxyapatites and phosphate glasses.21-23 The 1750 cm-1 peak is due to the carbonyl group. This peak intensity is strong after the first plasma treatment and then decreases as the plasma treatment is repeated. This indicates that carbonyl-containing (21) Masui, T.; Hirai, H.; Imanaka, N.; Adachi, G. Phys. Status Solidi A 2003, 198, 364. (22) Panda, R. N.; Hsieh, M. F.; Chung, R. J.; Chin, T. S. J. Phys. Chem. Solids 2003, 64, 193. (23) Tkalcec, E.; Sauer, M.; Nonninger, R.; Schmidt, H. J. Mater. Sci. 2001, 36, 5253.
Decomposition of Organophosphorus NerVe Agents
Figure 5. Decomposition rate comparison for the plasma, UV/ ozone, and UV processes. The remaining amounts of (a) parathion and (b) paraoxon were estimated from the PM-RAIRS intensities of the 1350, 1490, 1525, and 1591 cm-1 peaks corresponding to the nitrobenzyl group. The error bars were calculated from these four peaks of three different samples.
species are initially produced via oxidation of the hydrocarbon part of the OP molecules. Although the IR peaks of amorphous phosphorus oxide and carbonyl species are grown in conjunction with the decomposition of OP molecules, the production yield of these species cannot be obtained from the IR intensity alone due to uncertainty in the IR absorbance and baseline correction of these peaks. Upon repeated treatments with plasma, these species are further oxidized and removed from the remaining film. The peak at 1650 cm-1 is due to the bending vibration of water adsorbed on or in the amorphous phosphorus oxide film. The broad O-H stretching peak is also observed in the 32003500 cm-1 region (data not shown). The spectral changes of parathion and paraoxon films upon oxygen plasma exposure are not due to the irradiation of highflux UV photons from the plasma source. The PM-RAIRS data for the OP films irradiated with the 200 W UV lamp (Figures 3b and 4b) show the monotonic and slow decrease of the nitrobenzyl peaks, but the P-O and PdO peaks are not changed at all. This implies that the UV absorption and decomposition take place in the nitrobenzyl group. In the case of parathion, the decrease of the PdS peak at 930 cm-1 appears to be accompanied by the growth of the PdO peak at 1250 cm-1, indicating the conversion of the thiophosphorus ester group to the oxophosphorus ester group. The growth of the carbonyl peak at 1750 cm-1 is negligible even after a significant decomposition of the nitrobenzyl group. The plasma-induced reaction products appear to be similar to the ozonolysis product (Figures 3c and 4c). The decrease of the nitrobenzyl characteristic peaks is accompanied by the appearance
Langmuir, Vol. 23, No. 15, 2007 8077
Figure 6. Drosophila culture results for plasma-, UV/ozone-, and UV-treated (a) parathion and (b) paraoxon test samples.
of broad peaks at 1055, 1245, and 1750 cm-1. In the case of parathion, the 930 cm-1 peak intensity decreases faster than that of other peaks, indicating the high susceptibility of the PdS bond to the oxidation reaction. The decomposition rates of parathion and paraoxon upon plasma, UV, and UV/ozone treatments are estimated by monitoring the decrease of the nitrobenzyl peak intensities. Figure 5 shows the remaining amounts of parathion and paraoxon estimated from the PM-RAIRS intensities of the 1350, 1490, 1525, and 1591 cm-1 peaks. After baseline correction, the peak intensity was read from the baseline to the peak maximum. More than 95% of the deposited parathion and paraoxon films are decomposed by a single scan under the Ar/O2 plasma at a scan speed of ∼3 mm/s. The decomposition product is mainly amorphous phosphorus oxide. After the second scan with the Ar/O2 plasma, there are no discernible peaks of intact parathion and paraoxon in PM-RAIRS. From the QCM measurements of the initial spin-coated films, the initial thicknesses of the paraoxon and parathion films are estimated to be ∼40 and ∼60 nm, respectively. This means that the flux of the reactive species (excited state of radicals and atoms as well as ozone) generated by the Ar/O2 plasma is high enough to decompose a >38 nm thick paraoxon film and a >57 nm thick parathion film in a single plasma scan at a 3 mm/s speed. These decomposition rates of the Ar/O2 plasma process are much faster than those of the photodecomposition and ozonolysis processes. The decomposition yield for parathion is only 75% after UV irradiation (200 W Hg lamp) for 600 s and 90% after UV/ozone exposure for 600 s. The paraoxon decomposition by UV and UV/ozone is even slower. About 66% of the initial paraoxon film still remains on the substrate even after 600 s of irradiation with the 200 W UV lamp, and 20% remains after 600 s of exposure to UV/ozone.
8078 Langmuir, Vol. 23, No. 15, 2007
The significantly faster photodecomposition of parathion than paraoxon seems to be related to the presence of the PdS bond. Detoxification Test of the Plasma-Treated OP Films. The effectiveness of the OP decomposition and the nontoxicity of the decomposition reaction product were tested by culturing Drosophila in a confined space containing the substrates which were initially coated with OP films and treated with the Ar/O2 plasma, UV, and UV/ozone. Drosophila were chosen for this test since they are very sensitive to OP pesticides due to their small mass. Figure 6 compares the number of live Drosophila as a function of culture time. The UV- and UV/ozone-treated parathion and paraoxon samples were still toxic enough to kill all Drosophila in less than 12 h in the culture bottle. In contrast, the Drosophila in the plasma-treated sample bottles lived their full life cycle and laid eggs. Even the newly hatched larvae were able to develop into adult flies. These culture test results qualitatively demonstrate the efficacy of the atmospheric rf plasma process for the complete decomposition and detoxification of OP nerve agents. It should be noted that the complete decomposition and detoxification of the OP films can be attained with only two scans of the Ar/O2 plasma at a scan rate of 4 mm/s (total ∼12 s treatment time for a 1 in. wide slide glass substrate), while the OP films treated with UV and UV/ozone for 5 min are still highly toxic.
Conclusions The plasma-induced decomposition reactions of organophosphorus nerve agent molecules are demonstrated. An atmospheric
Kim et al.
rf plasma system was constructed for scanning treatments over large sample surfaces. The optical emission spectroscopy analysis found the oxygen atom and the excited OH and N2 species in the Ar/O2 plasma generated in the air. These reactive species in the plasma were able to completely decompose thick layers of parathion and paraoxon on the substrate surface in a very short exposure time. The decomposition reaction product is amorphous phosphorus oxides, which is similar to the product of the ozonolysis reaction of parathion and paraoxon. Since the decomposition reactions are based on excited atoms, radicals, and molecules, a wide range of chemical contaminants can be treated with the atmospheric rf plasma. Acknowledgment. This work was partially supported by a 3M Nontenured Faculty Grant. The Drosophila flies used in this study were kindly provided by Mr. Hyun-Gwan Lee and Professor Kyung-An Han. We are also grateful to Professor Qing Wang for allowing us to use the spectrophotometer for optical emission analysis. Supporting Information Available: Photographs of the Drosophila culture test. This material is available free of charge via the Internet at http://pubs.acs.org. LA700692T
