Ionization


Development of a Gas-Cylinder-Free Plasma Desorption/Ionization...

3 downloads 159 Views 1MB Size

Article pubs.acs.org/ac

Development of a Gas-Cylinder-Free Plasma Desorption/Ionization System for On-Site Detection of Chemical Warfare Agents Takahiro Iwai,†,∥ Ken Kakegawa,‡ Mari Aida,‡ Hisayuki Nagashima,† Tomoki Nagoya,† Mieko Kanamori-Kataoka,† Hidekazu Miyahara,‡ Yasuo Seto,*,† and Akitoshi Okino*,‡ †

National Research Institute of Police Science, 6-3-1, Kashiwanoha, Kashiwa, Chiba, Japan Department of Energy Sciences, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama, Kanagawa Japan



S Supporting Information *

ABSTRACT: A gas-cylinder-free plasma desorption/ionization system was developed to realize a mobile on-site analytical device for detection of chemical warfare agents (CWAs). In this system, the plasma source was directly connected to the inlet of a mass spectrometer. The plasma can be generated with ambient air, which is drawn into the discharge region by negative pressure in the mass spectrometer. High-power density pulsed plasma of 100 kW could be generated by using a microhollow cathode and a laboratory-built highintensity pulsed power supply (pulse width: 10−20 μs; repetition frequency: 50 Hz). CWAs were desorbed and protonated in the enclosed space adjacent to the plasma source. Protonated sample molecules were introduced to the mass spectrometer by airflow through the discharge region. To evaluate the analytical performance of this device, helium and air plasma were directly irradiated to CWAs in the gas-cylinder-free plasma desorption/ionization system and the protonated molecules were analyzed by using an ion-trap mass spectrometer. A blister agent (nitrogen mustard 3) and nerve gases [cyclohexylsarin (GF), tabun (GA), and O-ethyl S-2-N,Ndiisopropylaminoethyl methylphosphonothiolate (VX)] in solution in n-hexane were applied to the Teflon rod and used as test samples, after solvent evaporation. As a result, protonated molecules of CWAs were successfully observed as the characteristic ion peaks at m/z 204, 181, 163, and 268, respectively. In air plasma, the limits of detection were estimated to be 22, 20, 4.8, and 1.0 pmol, respectively, which were lower than those obtained with helium plasma. To achieve quantitative analysis, calibration curves were made by using CWA stimulant dipinacolyl methylphosphonate as an internal standard; straight correlation lines (R2 = 0.9998) of the peak intensity ratios (target per internal standard) were obtained. Remarkably, GA and GF gave protonated dimer ions, and the ratios of the protonated dimer ions to the protonated monomers increased with the amount of GA and GF applied.

C

hemical warfare agents (CWAs)1,2 are highly hazardous substances that have been used to incapacitate or decimate people by military forces or terrorists since World War I. Until now, a number of countries have ratified the Chemical Weapons Convention,3 which was established in 1997, that prohibits the development, production, stockpiling, and use of chemical weapons. However, chemical warfare and terrorism are ongoing threats, witnessed by the fact that CWAs were used in a series of serious crimes by Japanese cult Aum Shinrikyo in the middle of the 1990s4 and in the civil war in Syria in 2013.5 In case of a chemical terror attack, there is an urgent need to detect and identify causative agents.6 In particular, rapid on-site detection of CWAs is important to rescue victims and prevent secondary damage. Thus, hand-held or portable detecting equipment, such as gas-detection tubes,7 surface acoustic wave (SAW) sensors,8 gas chromatography (GC)-based instruments, such as GC/MS instruments,9−14 and ion-mobility spectrometry (IMS) instruments15−27 have been developed and used. However, gas-detection tubes are difficult © 2015 American Chemical Society

to operate while wearing protective clothing and take a long time to respond.28 SAW sensors have low sensitivity and are prone to frequent false positives.28 Thus, these devices have some problems for on-site use. GC-based methods have specificity and relatively high sensitivity (sub μg/m3)29 depending on the performance of the detector. However, it takes a long time to perform the measurements, and pretreatment or derivatization is needed for the analysis of some CWAs.30,31 Also, expertise is needed to operate the instrument. The size of IMS devices can be minimized because of ambient pressure hardware, and hand-held instruments are now commercialized. They are relatively easy to use and highly sensitive, and the measurement is very rapid, although they lack accuracy and specificity.28 Due to these characteristics, portable Received: March 5, 2015 Accepted: May 9, 2015 Published: May 9, 2015 5707

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry Table 1. Chemical Structures, Molecular Weights, Target Ions, Volatilities, and Toxicities of the Examined CWAs

a Values taken from ref 33. bFifty percent lethal concentration for 1 min of exposure in humans (ref 1). The LCt50 of GF is assumed to be the same as that for GD in ref 1.

ionization source for mass spectrometric analysis of materials on surfaces.48 The gas temperature of this plasma is about room temperature (approximately 30 °C), and this plasma source consumes very low power (below 3 W); thus, the device can be downsized. Hendricks et al. applied LTP to an ADI source for a portable mass spectrometer, and the detection of CWA simulants, illicit drugs, and explosives was demonstrated.49 However, in most cases, a gas cylinder needed to be carried with the analytical device because helium was used as a discharge gas. Thus, there are some problems in terms of portability and running cost. Therefore, a gas-cylinder-free plasma desorption/ionization system for on-site mass analysis of CWAs was developed in this study. In a previous study, a high-power pulsed microplasma was developed for on-site analysis of materials on human skin and was used as the plasma source for ADI-MS analysis.50 This plasma source has a low gas temperature and gives no electrical shock to object materials. Thus, the plasma can be irradiated for heat-sensitive or valuable materials. In addition, an air-stable plasma can be generated with this plasma source. We tried to develop this system based on ADI-MS by using high-power pulsed microplasma. In this system, outside air is drawn into an enclosed space in the plasma cell and the sample is desorbed and ionized in this space by plasma irradiation. The ionized sample is introduced into the mass spectrometer without diffusion into the outside atmosphere. In this paper, the design of the developed system is introduced and analytical performance of this system for various CWAs was evaluated by using real refractory CWAs.

GC-based instruments and IMS devices are frequently used in real detection situations. For GC-based and IMS methods, CWAs have to be introduced into the analytical instruments in the gas phase. Thus, in principle, nonvolatile CWAs cannot be detected with high sensitivity.32 For example, nerve gas O-ethyl S-2-N,Ndiisopropylaminoethyl methylphosphonothiolate (VX) is hard to evaporate. Once VX is spread, the molecules remain on a materials surface for a long time.1,33 VX that remains on any surface must be detected rapidly on-site because the molecules are absorbed through the skin into the body and manifest severe toxicity. It is difficult to detect VX by normal GC-based or IMS methods. Thus, VX must be swabbed with paper and then vaporized by heating for introduction into the IMS instrument. However, it is possible that the collection efficiency of swabbing is low and is not expected to have enough sensitivity.34 If CWAs are detected, false-positive results are occasionally shown due to the relatively low analytical precision of the method.26 Moreover, the heat source for sample vaporization can potentially initiate an explosion, which might occur at the same location in a terrorist incident. For these reasons, development of a new mobile detector, which is easy to use and highly sensitive and has high precision, for analyzing refractory CWAs on-site is needed. We focused on ambient desorption/ionization mass spectrometry (ADI-MS) as an analytical method for detection of nonvolatile CWAs. In this technique, solid and liquid samples on surfaces can be directly analyzed under atmospheric pressure without sample preparation. Desorption electrospray ionization (DESI),35 the first reported ADI method, and direct analysis in real time (DART),36 which utilizes atmospheric plasma as the source of desorption/ionization, were developed in 2004 and 2005, respectively. Currently, a wide variety of ADI-MS techniques have been developed, triggered by development of the DESI and DART methods,37,38 DESI and DART have been both commercialized and actively applied to forensic and homeland security missions as the most reliable ADI-MS methods.39−41 Applications for the detection of CWAs were also researched using real agents by D’Agostino et al.42−45 and Nilles et al.,46 where CWAs were extracted by a headspace manner from CWA-contaminated materials using solid phase microextraction (SPME) fiber and the resulting SPME fibers were subjected to DESI analysis with MS/MS and IMS/MS. However, these studies remain within the confines of laboratory application and, apart from the report by Mulligan et al., studies of mobile CWA-detection equipment using DESI or DART suitable for practical use are rare.47 In 2008, Harper et al. applied a kind of dielectric barrier discharge (DBD) jet, termed a low-temperature plasma (LTP) probe, as a desorption/



INSTRUMENTAL DESIGN AND EXPERIMENTAL SECTION Chemicals and Reagents. Tris(2-chloroethyl)amine (nitrogen mustard 3, HN3), cyclohexyl methylphosphonofluoridate (GF), O-ethyl-N,N-dimethylaminophosphonocyanidate (tabun, GA), VX, and dipinacolyl methylphosphonate (DPMP) were synthesized in the laboratories of the National Research Institute of Police Science (Kashiwa, Japan). All of these compounds were >98% pure by GC/MS analysis. Chemical structures, molecular weights, target ions, volatilities, and median lethal dosages (LCt50) of the target CWAs are given in Table 1. Solutions of CWAs in n-hexane (Wako Pure Chemical Industries, Osaka, Japan) were prepared at the desired concentrations. Methanol (Wako Pure Chemical Industries, Osaka, Japan) was used for cleaning the sample CWAs from the surface of a Teflon rod, which was used for sample introduction. 5708

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry

Operating Conditions for High-Power Pulsed Microplasma. A custom-built pulsed high-power supply, used in our previous study,50 was adopted for plasma generation. This power supply can work even with a dry-cell battery because the average energy consumption is a few W. Thus, this power supply can be miniaturized and is suitable for a mobile device. The operating parameters for the high-power pulsed microplasma were: capacitance of the main discharge capacitor, 1 μF; charging voltage, 300 V; repetition frequency of the plasma, 50 Hz. Because a high-power pulsed microplasma jet can be generated with air, gas-cylinder-free analysis can be realized by using the surrounding air as the discharge gas. In this study, helium supplied from a gas cylinder and the laboratory air were used as discharge gases. The helium discharge gas (99.999%) was obtained from Jyoto Gas Co. Ltd. (Tokyo, Japan). Air in the laboratory contains dust particles and other contaminating compounds, and it is highly probable that the inside of the mass spectrometer would become dirty if laboratory air was introduced into the mass spectrometer as is. For this reason, a filter for suppression of contaminants was installed at the gas introduction path when the air plasma was used to generate the plasma. The filter consisted of a serially connected column filled with silica gel and a 0.2 μm membrane filter. The airflow rate was 500 mL/min; hence, the helium-flow rate was adjusted to the same rate by using a mass-flow controller (MC-2 SLPMD, Alicat Scientific, Inc., Tucson, AZ, USA). The direction of the gas flow was toward the inlet of the mass spectrometer (i.e., the opposite direction to the sample surface). Thus, we anticipated that the plasma would not expand toward the sample surface. However, when the gas-flow rate was 500 mL/ min, the gas moved only about 340 μm within about 20 μs of pulsed plasma generation time, even in the smallest hole (800 μm in diameter) in the plasma cell. Pulsed plasma was extended toward both directions regardless of the polarity of the electrode, and the expansion rate of the plasma was sufficiently quicker than the gas-flow rate. Thus, plasma extension is almost independent from the gas flow. Therefore, the sample can be irradiated by the plasma if the sample is placed within approximately 5 mm from the plasma source. As described above, the rising gas temperature of the plasma is suppressed because pulsed plasma is generated and disappears within a very short time (10−20 μs). Moreover, the plasma does not cause electrical damage to the material of the object being tested because the electrode on the sample side is earthed; therefore, the plasma can be touched by conditioning the repetition frequency of discharge or charging voltage to the capacitor for the main discharge. Mass Spectrometric Analysis. All measurements were carried out by coupling the gas-cylinder-free plasma desorption/ionization cell with an ion-trap mass spectrometer (1100 series LC/MSD trap, Agilent Technologies, Tokyo, Japan). The principal parameters of the mass spectrometer were: capillary voltage, 0 V; capillary exit voltage, 111.1 V; octapole RF voltage, 146.3 Vpp; m/z range, 20−550; ion-accumulation time, 20 ms. All the measurements were carried out in positive-ion mode, and ion charge control was not used. Instrumental tunings, such as ion lens voltages, were adjusted to obtain the maximum signal intensity of the analyte target ions for the respective samples. In the case of MS2 analysis, the value of the fragmentation amplitude was adjusted to observe a certain amount of parent ions. To prevent the adsorption of samples and contaminant compounds to the inlet capillary, the dry gas temperature was set at 310−365 °C, but the outside head of the

Gas-Cylinder-Free Plasma Desorption/Ionization Cell. A schematic of a proof-of-concept gas-cylinder-free plasma desorption/ionization system is shown in Figure 1. This new

Figure 1. Schematic of a proof-of-concept gas-cylinder-free plasma desorption/ionization system.

plasma cell consisted of the electrodes for plasma generation and two Duracon parts, A and B, for pinching and holding the electrodes. A detailed description of the cylindrical microhollow cathode can be found in a previous report.50 The actual highpower pulsed microplasma jet used was slightly modified from that description. The thickness of the molybdenum plate, the thickness of the quartz plate for electric insulation, and the diameter of the small hole for plasma generation were 2 mm, 1 mm, and 800 μm, respectively. Part A had connected cylindrical holes, which had diameters of 14 and 16 mm in the central axis. Part A also had a hole 3 mm in diameter connected to the central axis hole from the radial direction. Gas tubing was connected to this hole for introducing gas into the plasma cell. Part B had a hole 2 mm in diameter in the central axis, the end of which was connected to the inlet capillary of the mass spectrometer. The cylinder-type electrode was pinched by these two Duracon parts. To make the joint between the Duracon parts and the electrode airtight, O-rings were installed on the contact surface. Part B had a circular shallow (1 mm deep) dent, and the electrode fit exactly with this hole so that the center axes of the small hole on the electrode and parts A and B were exactly aligned. This means that the discharge region was directly connected to the inlet of the mass spectrometer through the hole in part B at a gap of 17 mm. If the hole of part A is covered by the sample attached to the surface, gas can be introduced to the inside cell through the gas-introduction hall by negative pressure in the mass spectrometer because at this point only the gas-introduction hole is opened. The gas flow is indicated by arrows in Figure 1. This gas was used as the discharge gas, and the sample was desorbed and ionized in the quasi-closed space, which consisted of the sample surface and the hole in front of the plasma source. Ionized samples are anticipated to pass with the gas flow through the discharge region and part B and finally be introduced into the mass spectrometer without diffusion into the outside atmosphere. This is the mechanism of surface sample analysis by the gascylinder-free plasma desorption/ionization system. In this study, to optimize the analytical parameters, the sample was placed on the edge of a Teflon rod, which was inserted into the plasma cell, as shown in Figure 1. 5709

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry

Figure 2. Mass spectra of CWAs ionized by the gas-cylinder-free plasma desorption/ionization system. (a) HN3, (b) GF, (c) GA, and (d) VX in positive-ion mode. The mass spectra in helium plasma and air plasma are shown in the upper and middle traces; the corresponding product-ion mass spectra are shown in the lower traces.

Safety Considerations. CWAs were carefully handled by specially trained personnel. All experiments that used CWAs were performed in a specific facility. Sample vapor was prepared in a fume hood equipped with an alkaline scrubber system to prevent laboratory personnel from being exposed to vapors and CWAs from being released outside the facility. The use and synthesis of CWAs for this research was approved by the Minister of Economy, Trade and Industry of Japan (http:// www.meti.go.jp/polycy/chemical_management/cwc/ 200kokunai/202horitu_gaiyo.htm).

capillary was not heated to the same temperature because the spray shield was dismounted in this study. All the sample preparations were carried out in a fume hood. A 5 μL drop of a solution of the sample in n-hexane (10−1000 ppm) was put in the shallow dent (diameter: 3 mm; depth: 1 mm) on the edge of a Teflon rod, and the sample was attached by drying the sample solution for 1 min. The 100 ppm sample solutions of HN3, GF, GA, and VX contain the absolute amounts 3.0, 3.1, 3.3, and 1.9 nmol, respectively. The Teflon rod that supported the sample was inserted into the cell to place the sample 2 mm from the discharge region, and the analysis was started just 1.5 min after the sample instillation. After each analysis, the Teflon rod was rinsed with methanol to remove the residual CWA and dried by a nitrogen purge. In the practical analysis, various kinds of material will become samples for CWA analysis. Thus, the analysis of CWAs on other materials will be investigated in future studies.



RESULTS AND DISCUSSION Analytical Performance for Detection of CWAs. Helium and air plasma were directly irradiated to the edge of the Teflon rod in the absence of a sample, and the background mass spectra were measured. Automated instrumental tuning was set to the target ion of VX (m/z = 268). The results are 5710

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry

Figure 3. Comparison of integrated signals derived from CWAs in helium and air plasma. (a) HN3, (b) GF, (c) GA, and (d) VX. Measurements were performed in triplicate. The average value ± the standard deviation (error bars) are shown.

63, respectively. GF provided an MH+ base peak at m/z = 181, and the fragment ion [CH3P(OH)2F]+ was detected at m/z = 99 in both helium and air plasmas (Figure 2b). The peak at m/z = 279 was subjected to MS2 analysis and provided fragment ions of m/z = 181 and 99. Thus, the peak at m/z = 279 is assumed to be derived from a complex of the GF molecule and the fragment CH3P(OH)2F. A dimer species (2MH+) was also observed at m/z = 361. The base peak at m/z = 181 was subjected to MS2 analysis and provided the CH3P(OH)2F+ ion at m/z = 99. The results for GA are shown in Figure 2c. Signal peaks for the monomer and dimer species were observed at m/ z = 163 and 325, respectively in both helium and air plasmas. The base peak at m/z = 163 was subjected to MS2 analysis and provided the [(CH3)2NP(OH)2(CN)]+ fragment at m/z = 135. VX gave m/z = 268 (MH+) as the base peak (Figure 2d). When air plasma was used, the fragment [(C3H7)2NHCOH]+ was observed at m/z = 130. The base peak at m/z = 268 was subjected to MS2 analysis; peaks for [(−CH2CH2−)N(CH(CH3)2)2]+ and [(−CH2CH2−)NHCH(CH3)2]+ were observed at m/z = 128 and 86, respectively. In the MS2 analysis of the CWAs used in this study, the spectra obtained had almost identical fragmentation patterns, just as the report of Seto et al.51 In the gas-cylinder-free plasma desorption/ ionization system, the sample molecules should pass through the generation area of high-power pulsed microplasma, which has high density and a high excitation temperature and can be used for inorganic analysis of ultrasmall samples.52 However, it was found that prominent fragmentation did not occur for the CWAs examined. This may be because pulsed plasma is repeatedly generated and then disappears. When the pulse plasma generation is repeated at 50 Hz intervals, the interval of plasma generation is 20 ms and the plasma absent time is much longer than the plasma lighting time. Desorbed and ionized sample molecules produced by pulsed plasma, for which the

shown in Figure S-1, Supporting Information. By using helium as the discharge gas, the signal intensity of the total ions was about 2.4 times higher relative to when filtered air was used; therefore, a lower background signal can be obtained when filtered air is used as the plasma gas. Actual CWAs were analyzed by using the gas-cylinder-free plasma desorption/ionization system developed in this study. The purpose of this study includes realizing gas-cylinder-free analysis by using ambient air. First, we examined whether analysis of CWAs was possible by using helium plasma, which is frequently used as discharge gas for ionization. Then, CWAs were analyzed without using a gas cylinder by drawing ambient air into the plasma cell. These sets of results were compared. To confirm that the detected ions were derived from the sample CWA molecules, MS2 analysis was also carried out. In the MS and MS2 analyses, solutions of CWA in n-hexane (100 and 1000 ppm, respectively) were examined. The 100 ppm sample solutions of HN3, GF, GA, and VX contain the absolute amounts 3.0, 3.1, 3.3, and 1.9 nmol, respectively. Typical spectra obtained by analysis of CWAs by using both helium and air plasmas and typical MS2 spectra, for which the protonated molecules were selected as the base peaks (HN3: m/z = 204; GF: m/z = 181; GA: m/z = 163; VX: m/z = 268), when air plasma was used, are shown in Figure 2. The chlorine isotopic ratio indicated that HN3 provided [C 6 H 12 35 Cl 3 NH] + , [C 6 H 1 2 3 5 Cl 2 3 7 ClNH] + , [C 6 H 1 2 3 5 Cl 3 7 Cl 2 NH] + , an d [C6H1237Cl3NH]+ ions at m/z = 204, 206, 208, and 210, respectively, in both helium and air plasma. An uncertain peak was measured at m/z = 140 in helium plasma (Figure 2a). The base peak at m/z = 204 was subjected to MS2 analysis, which showed [C 4 H 9 N 35 Cl 3 ] + , [( 35 ClC 2 H 4 ) 2 N(−CH 2 CH 2 −)] + , [C4H10N35Cl2]+, [(35ClC2H4)NH(−CH2CH2−)]+, [NH 2 (−CH 2 C 3 5 ClH−)] + , [N(−CH 2 CH 2 −) 2 ] + , and [C2H435Cl]+ ions at m/z = 176, 168, 142, 106, 78, 70, and 5711

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry duration time is about 20 μs, should travel the distance of about 340 mm between the electrodes in 20 ms with the gas flow if the gas-flow rate is 500 mL/min. Actually, the gas-flow speed might be faster than this estimation because of negative pressure in the cell. Most of the sample ions are introduced into the mass spectrometer during the period when pulsed plasma is absent, even if the plasma expands. This might be the reason why this system has less incidences of the thermal fragmentation caused by high-power plasma. To compare the signals obtained in helium and air plasmas, the intensities of the signals derived from the sample molecules were summed (Figure 3). In the HN3 analyses, the signal intensity of m/z = 204 was decreased about 52% in air plasma relative to that in helium plasma. Similarly, the signal intensities of m/z = 181, 163, and 268 in the GF, GA, and VX analyses, respectively, were decreased about 66%, 79%, and 8%, respectively. The tendency of higher intensity signals for analysis in helium plasma than in air plasma seems common to the CWAs examined, although the data fluctuate due to hand manipulation. The decreasing signal rate of CWAs with higher volatility seemed to be larger than that of CWAs with lower volatility (HN3:120 mg/m3; GF: 659 mg/m3; GA: 611.3 mg/ m3; VX: 10.07 mg/m3 at 25 °C).33 To ascertain this result, the gas temperatures of helium and air plasma were measured with a sheath-type thermocouple (CT-1200D, CUSTOM, Tokyo, Japan) 2 mm away from the discharge region with a gas-flow rate of 500 mL/min: both were about 27 °C (room temperature = 25 °C). We conclude that the difference in signal decreasing rates of the CWAs are not caused by differences in the degree of thermal desorption, which is believed to heavily mediate sample desorption. The ionization process is assumed to be proton attachment from a protonwater cluster,53 and this mechanism might be valid for all the chemicals irrespective of target species and plasma gas species. Instead, we think that the difference in the signal decreasing rates of the CWAs are affected by the difference in the plasmarelated desorption process, the mechanism of which is poorly understood and should be investigated further. Limits of detection (LODs) for the CWAs were calculated in helium and air plasma. Generally, the LOD is evaluated from a calibration curve. However, the LODs could not be evaluated because of the large error derived from unstable sensitivity of the mass spectrometer and hand manipulation of the instrument. Thus, the LODs were calculated on the basis of a signal-to-noise ratio (S/N) of 3, assuming that the dose− response was sufficiently linear. S and N are the sample signal intensity at a concentration of 100 ppm and the standard deviation of the background signal, respectively, as reported by Iwai et al.50 The ion-accumulation time for the calculation was manually set at the time interval from the signal rising to being buried in noise. Calculated LODs for the CWAs are shown in Table 2. LODs at the pmol level were obtained for the analysis of all the CWAs. As shown in Figure 3, the signal intensity was higher in helium plasma than air plasma. However, comparison of the LODs for the monomer signals of the respective CWAs revealed that lower LODs could be obtained in air plasma, due to its lower background noise. Comparison of the LODs for dimers of GF and GA showed that lower LODs were obtained in helium plasma. In the analysis of VX, which has the lowest volatility and highest toxicity presented by the lowest lethal concentration (LCLo) of percutaneous exposure (89 μg/kg),54 the lowest LOD (1 pmol) was achieved in air plasma. According to the LCLo of VX, a person (body weight: 60

Table 2. Monitored Ions and LODs of the CWAs ion monitored agent HN3

GF GA VX

m/z 204 206 208 181 361 163 325 268

LOD (pmol)

assignment +

MH MH+ MH+ MH+ 2MH+ MH+ 2MH+ MH+

He

air

30 42 86 23 11 6.8 18 1.5

22 21 12 20 15 4.8 19 1.0

kg) could die if they touched 20 μmol of VX. LODs required for on-site detection are 1/100 of the lethal concentration in 1 min55 and, in this case, the required LOD in the chemical terrorism countermeasure aspect is estimated at 200 nmol. From this LOD evaluation, we have shown that our system can satisfy the sensitivity requirements for practical use in on-site detection, whereas sensitivity is dependent on the mass spectrometer, target chemicals, interfering materials, and ionaccumulation time, and could be adjusted by digital data processing after data acquisition in this study. For the field-deployability, the ion trap mass spectrometer used for counter-flow introduction atmospheric pressure chemical ionization MS51 could be attached to this gascylinder-free plasma desorption/ionization system. Instead, a miniature mass spectrometer, such as the device used in Hendricks et al.,49 could be considered. In practical use, as shown in Figure 1, by changing the depth of part A, the sample ions could be introduced into the mass spectrometer by contact of the surface to be analyzed with the axial opening space of part A. In this system, chemicals on the surface could be desorbed by direct plasma irradiation. This system is suitable for a mobile CWAs detection device because it does not need a gas cylinder and the plasma source is handheld size. Quantitative Analysis. By using a gas-cylinder-free plasma desorption ionization system, the analysis of CWAs on surfaces could be achieved with sufficient sensitivity. Next, the quantitative performance of our system was evaluated. As described in the previous section, the calibration curve was not linear due to fluctuation of signal intensity. Thus, DPMP, a phosphonate compound like the nerve agents, was used as internal standard, and the calibration curve was attempted. DPMP is chemically stable, nontoxic, and less volatile. DPMP (1.9 nmol) and VX (4.2 nmol) were mixed, and the results are shown in Figure 4a. The MH+ signal was observed at m/z = 265 for DPMP (molecular weight: 264.35). To evaluate the effect of the DPMP internal standard, data analysis was performed such that the peak area of the VX signal was accumulated 10 s after the beginning of plasma irradiation (Figure 4b). Similarly, the peak areas of monomer (m/z = 265) and dimer (m/z = 529) signals for DPMP were accumulated for 10 s. The peak area for the monomer and doubled peak area of the dimer were summed, and the ratio of the peak area of the VX monomer to the total peak area for DPMP (VX/DPMP) was calculated. The ratio VX/DPMP was used for construction of the calibration curve (Figure 4c). The calibration curve made by only the VX signal was not linear, but a sufficiently linear calibration curve (R2 = 0.9998) could be obtained in the range of 0.84−12.7 nmol when DPMP was used as an internal standard, even though it had a very short measurement time (10 s). 5712

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry

idate (sarin, GB), O-pinacolyl methylphosphonofluoridate (Soman, GD), and GF in a concentration-dependent manner in GC-electron ionization cylindrical ion trap MS.13 The observation of the GA and GA dimer ions in this experiment may be partially ascribed to the Self-CI mechanism. Instead, we rather consider that the formation of GF and GA dimer ions occurred in the atmospheric pressure condition through the plasma desorption/ionization, and the produced monomer and dimer ions of GA and GF were transported to the ion trap through the elimination of the neutral compounds. We observed dimer ions from GB, GD, GD, GA, and GF under corona discharge ionization in IMS analysis in addition to monomer ions with a concentration-dependent manner.57 Using the same ion trap MS instrument, we could not observe any dimer ion from GB, GD, GD, GA, nor GF in atmospheric pressure chemical ionization MS, irrespective of a clear observation of the protonated monomer ions.51 However, it is not difficult to discriminate strictly the dimer ion formation through atmospheric pressure ionization from ion trap Self-CI under a vacuumed condition. Thus, further investigation is needed in the future.



CONCLUSION In this study, a gas-cylinder-free plasma desorption/ionization system was developed and first applied to analysis of nonvolatile CWAs. In this system, CWAs are desorbed and ionized by high-power pulsed microplasma in a quasi-closed space. Protonated sample molecules were introduced into the mass spectrometer by gas flow through the discharge region. Gas-cylinder-free analysis can be achieved because ambient air can be used as the plasma gas. By using this system, HN3, GF, GA, and VX (introduced on a Teflon rod) were successfully measured with LODs at the pmol level. The sample signal intensity was higher in helium plasma, but the LODs were lower in air plasma due to its low background noise. To achieve the quantitative analysis, calibration curves were made by using DPMP as an internal standard and a linear plot (R2 = 0.9998) was obtained. In the analysis of GA and GF, we found a proportional relationship between the absolute amount of sample and the dimer/monomer signal ratio, and this might be used for estimating the amount of analyte. These investigations demonstrated the potential of the gas-cylinder-free plasma desorption/ionization system as a mobile on-site analytical method to detect nonvolatile CWAs on potentially contaminated surfaces. Future studies will focus on the influence of surface materials, matrix interference, and the development of a mobile instrument.

Figure 4. Quantification of VX using internal standard. (a) Mass spectrum of the mixture of DPMP (1.9 nmol) and VX (4.2 nmol) in air plasma. (b) Calibration curve from the peak area of the VX signals summed for 10 s from the beginning of plasma irradiation. (c) Calibration plot of the peak area ratio of VX/DPMP; signals were summed for 10 s from the beginning of plasma irradiation.

In the analysis of GF and GA, which showed relatively strong dimer signals, dependence of the dimer/monomer ratio on the sample concentration was investigated in air plasma. Solutions of 10, 100, and 1000 ppm concentrations were used for analysis. The absolute amount of sample molecules (x-axis) was plotted against the ratio of dimer/monomer peak areas (y-axis), and the results are shown in Figure S-2, Supporting Information. For both GF and GA, there is a proportional relationship between the absolute amount of sample molecules and the dimer/monomer signal ratio. From this result, rough quantitative analysis can be achieved by measuring the dimer/ monomer signal ratio. Of course, the existence of interfering substances or species other than the target CWAs might strongly affect the signal ratio values. In ion trap MS, selfchemical ionization (Self-CI) would occur through ion/ molecule reaction.56 Indeed, Smith et al. reported that dimer ions were observed from O-isopropyl methylphosphonofluor-



ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b00874.



AUTHOR INFORMATION

Corresponding Authors

*Tel: +81-4-7135-8001. Fax: +81-4-7133-9173. E-mail: seto@ nrips.go.jp. *Tel: +81-45-924-5689. Fax: +81-45-924-5689. E-mail: [email protected]. 5713

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry Present Address

(23) Sekioka, R.; Takayama, Y.; Seto, Y.; Urasaki, Y.; Shinzawa, H. Bunseki Kagaku 2007, 56, 17−22. (24) Zimmermann, S.; Abel, N.; Baether, W.; Barth, S. Sens. Actuators, B 2007, 125, 428−434. (25) Yamaguchi, S.; Asada, R.; Kishi, S.; Sekioka, R.; Kitagawa, N.; Tokita, K.; Yamamoto, S.; Seto, Y. Forensic Toxicol. 2010, 28, 84−95. (26) Kishi, S.; Sekioka, R.; Sodeyama, M.; Shiga, M.; Seto, Y. Bunseki Kagaku 2010, 59, 65−76. (27) Mäkinen, M. A.; Anttalainen, O. A.; Sillanpäa,̈ E. T. Anal. Chem. 2010, 82, 9594−9600. (28) Seto, Y.; Kanamori-Kataoka, M.; Tsuge, K.; Ohsawa, I.; Matsushita, K.; Sekiguchi, H.; Itoi, T.; Iura, K.; Sano, Y.; Yamashiro, S. Sens. Actuators, B 2005, 108, 193−197. (29) Seto, Y.; Kanamori-Kataoka, M.; Tsuge, K.; Ohsawa, I.; Maruko, H.; Sekiguchi, H.; Sano, Y.; Yamashiro, S.; Matsushita, K.; Sekiguchi, H.; Itoi, T.; Iura, K. Toxin Rev. 2007, 29, 299−312. (30) Haas, R. Environ. Sci. Pollut. Res. 1998, 5, 2−3. (31) Haas, R.; Schmidt, T. C.; Steinbach, K.; von Löw, E. Fresenius’ J. Anal. Chem. 1998, 361, 313−318. (32) Seto, Y.; Kanamori-Kataoka, M.; Tsuge, K. J. Mass Spectrom. Soc. Jpn. 2008, 56, 91−115. (33) Marrs, T. C.; Maynard, R. L.; Sidell, F. R. Chemical Warfare Agents: Toxicology and Treatment; John Wiley & Sons: Chichester, 1996. (34) Behringer, D. L.; Smith, D. L.; Katona, V. R.; Lewis, A. T., Jr.; Hernon-Kenny, L. A.; Crenshaw, M. D. Forensic Sci. Int. 2014, 241, 7− 14. (35) Takáts, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471−473. (36) Cody, R. B.; Laramée, J. A.; Durst, H. D. Anal. Chem. 2005, 77, 2297−2302. (37) Monge, M. E.; Harris, G. A.; Dwivedi, P.; Fernandez, F. M. Chem. Rev. 2013, 113, 2269−2308. (38) Harris, G. A.; Galhena, A. S.; Fernandez, F. M. Anal. Chem. 2011, 83, 4508−4538. (39) Morelato, M.; Beavis, A.; Kirkbride, P.; Roux, C. Forensic Sci. Int. 2013, 226, 10−21. (40) Lesiak, A. D.; Shepard, J. R. E. Bioanalysis 2014, 6, 819−842. (41) Laramée, J. A.; Cody, R. B.; Nilles, J. M.; Durst, H. D. In Forensic Analysis on the Cutting Edge; Blackledge, R. D., Ed.; John Wiley and Sons, Inc.: Hoboken, NJ, 2007; pp 175−195. (42) D’Agostino, P. A.; Chenier, C. L.; Hancock, J. R.; Lepage, C. R. J. Rapid Commun. Mass Spectrom. 2007, 21, 543−549. (43) D’Agostino, P. A.; Hancock, J. R.; Chenier, C. L.; Lepage, C. R. J. J. Chromatogr. A 2006, 1110, 86−94. (44) D’Agostino, P. A.; Chenier, C. L. Rapid Commun. Mass Spectrom. 2010, 24, 1617−1624. (45) D’Agostino, P. A. In DESI-MS/MS of Chemical Warfare Agents and Related Compounds Detection of Biological Agents for the Prevention of Bioterrorism; Banoub, J., Ed.; Springer: Netherlands, 2011; pp 163− 179. (46) Nilles, J. M.; Connell, T. R.; Durst, H. D. Anal. Chem. 2009, 81, 6744−6749. (47) Mulligan, C. C.; Talaty, N.; Cooks, R. G. Chem. Commun. 2006, 16, 1709−1711. (48) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2008, 80, 9097−9104. (49) Hendricks, P. I.; Dalgleish, J. K.; Shelley, J. T.; Kirleis, M. A.; McNicholas, M. T.; Chen, T. C.; Chen, C.-H.; Duncan, J. S.; Boudreau, F.; Noll, R. J.; Denton, J. P.; Roach, T. A.; Ouyang, Z.; Cooks, R. G. Anal. Chem. 2014, 86, 2900−2909. (50) Iwai, T.; Kakegawa, K.; Okumura, K.; Kanamori-Kataoka, M.; Miyahara, H.; Seto, Y.; Okino, A. J. Mass Spectrom. 2014, 49, 522−528. (51) Seto, Y.; Sekiguchi, H.; Maruko, H.; Yamashiro, S.; Sano, Y.; Takayama, Y.; Sekioka, R.; Yamaguchi, S.; Kishi, S.; Satoh, T.; Sekiguchi, H.; Iura, K.; Nagashima, H.; Nagoya, T.; Tsuge, K.; Ohsawa, I.; Okumura, A.; Takada, Y.; Ezawa, N.; Watanabe, S.; Hashimoto, H. Anal. Chem. 2014, 86, 4316−4326.



T.I.: Department of Applied Chemistry for Environment, Kwansei Gakuin University, 2-1 Gakuen, Sanda, Hyogo, Japan. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We would like to thank the Plasma Concept Tokyo, Inc. for their support and collaboration.



REFERENCES

(1) Stewart, C. E.; Sullivan, J. B., Jr.. In Hazardous Materials Toxicology - Clinical Principles of Environmental Health; Sullivan, J. B., Jr., Krieger, G. R., Eds.;Williams & Wilkins: Baltimore, MD, 1992; Chapter 100, pp 986−1014. (2) Somani, S. M. Chemical Warfare Agents; Academic Press: San Diego, CA, 1992. (3) Organization of the Prohibition of Chemical Weapons. Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons on their Destruction. http://www.opcw.org (accessed May 7, 2015). (4) Seto, Y.; Tsunoda, N.; Kataoka, M.; Tsuge, K.; Nagano, T. In Natural and Selected Synthetic Toxins Biological Implications; Tu, A. T., Gaffield, W., Eds.; ACS Symposium Series 745; American Chemical Society, Washington, DC, 1999; pp 318−332. (5) United Nations. United Nations mission to investigate allegations of the use of chemical waeapons in the Syrian Arabic Republic, Report if the alleged use of a chemical weapons in the Ghouta area of Damascus on 21 August 2013, http://www.un.org/disarmament/ content/slideshow/Secretary_General_Report_of_CW_Investigation. pdf (accessed May 7, 2015). (6) Society for Countermeasure against Chemical, Biological, Radiological, Nuclear, and Explosive Terrorism. Nuclear, Biological, and Chemical Terrorism Countermeasure Handbook; Shindan To Chiryo Sha: Tokyo, 2008. (7) Takayama, Y.; Sekioka, R.; Sekiguchi, H.; Maruko, H.; Ohmori, T.; Seto, Y. Bunseki Kagaku 2007, 56, 355−362. (8) Matsushita, K.; Sekiguchi, H.; Seto, Y. Bunseki Kagaku 2005, 54, 83−88. (9) Hook, G. L.; Kimm, G.; Betsinger, G.; Savage, P. B.; Swift, A.; Logan, T.; Smith, P. A. J. Sep. Sci. 2003, 26, 1091−1096. (10) Smith, P. A.; Koch, D.; Hook, G. L.; Erickson, R. P.; Lepage, C. R. J.; Wyatt, R. P. Trends Anal. Chem. 2004, 23, 296−306. (11) Sekiguchi, H.; Matsushita, K.; Yamashiro, S.; Sano, Y.; Seto, Y.; Okuda, T.; Sato, A. Forensic Toxicol. 2006, 24, 17−22. (12) Contreras, J. A.; Murray, J. A.; Tolley, S. E.; Oliphant, J. L.; Tolley, H. D.; Lammert, S. A.; Lee, E. D.; Later, D. W.; Lee, M. L. J. Am. Soc. Mass Spectrom. 2008, 19, 1425−1434. (13) Smith, P. A.; Lepage, C. J.; Lukacs, M.; Martin, N.; Shufutinsky, A.; Savage, P. B. Int. J. Mass Spectrom. 2010, 295, 113−118. (14) Smith, P. A.; Lepage, C. J.; Savage, P. B.; Bowerbank, C. R.; Lee, E. D.; Lukacs, M. J. Anal. Chim. Acta 2011, 690, 215−220. (15) Sohn, H.; Steinhanses, J. Int. J. Ion Mobility Spectrom. 1998, 1, 1−14. (16) Tayler, S. J.; Piper, L. J.; Connor, J. A.; Fitzgerald, J.; Adams, J. H.; Harden, C. S.; Shoff, D. S.; Davis, D. M.; Ewing, R. G. Int. J. Ion Mobility Spectrom. 1998, 1, 58−63. (17) Baumbach, J. I.; Eiceman, G. A. Appl. Spectrosc. 1999, 53, 338A− 355A. (18) Cottingham, K. Anal. Chem. 2003, 75, 435A−439A. (19) Utriainen, M.; Karpanoja, E.; Paakkanen, H. Sens. Actuators, B 2003, 93, 17−24. (20) Eiceman, G. A.; Stone, J. A. Anal. Chem. 2004, 76, 390A−397A. (21) Maruko, H.; Sekiguchi, H.; Seto, Y.; Satoh, A. Bunseki Kagaku 2006, 55, 191−197. (22) Keller, T.; Keller, A.; Tutsch-Bauer, E.; Monticelli, F. Forensic Sci. Int. 2006, 161, 130−140. 5714

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715

Article

Analytical Chemistry (52) Iwai, T.; Okumura, K.; Kakegawa, K.; Miyahara, H.; Okino, A. J. Anal. At. Spectrom. 2014, 29, 2108−2113. (53) Chan, G. C.-Y.; Shelley, J. T.; Wiley, J. S.; Engelhard, C.; Jackson, A. U.; Cooks, R. G.; Hieftje, G. M. Anal. Chem. 2011, 83, 3675−3686. (54) World Health Organization. Health Aspects of Chemical and Biological Weapons, 1st ed.; WHO: Geneva, 1970. http://www.who. int/csr/delibepidemics/biochem1stenglish/en/ (accessed May 7, 2015). (55) Seto, Y. Forensic Sci. Rev. 2014, 26, 23−51. (56) McLuckey, S. A.; Glish, G. L.; Asano, K. G.; Van Berkel, G. J. Anal. Chem. 1988, 60, 2312−2314. (57) Satoh, T.; Kishi, S.; Nagashima, H.; Tachikawa, M.; KanamoriKataoka, M.; Nakagawa, T.; Kitagawa, N.; Tokita, K.; Yamamoto, S.; Seto, S. Anal. Chim. Acta 2015, 865, 39−52.

5715

DOI: 10.1021/acs.analchem.5b00874 Anal. Chem. 2015, 87, 5707−5715