2015 年 4 巻 1 号 p. A0038
We previously reported on the development of a portable mass spectrometer for the onsite screening of illicit drugs, but our previous sampling system could only be used for liquid samples. In this study, we report on an attempt to develop a probe heating method that also permits solid samples to be analyzed using a portable mass spectrometer. An aluminum rod is used as the sampling probe. The powdered sample is affixed to the sampling probe or a droplet of sample solution is placed on the tip of the probe and dried. The probe is then placed on a heater to vaporize the sample. The vapor is then introduced into the portable mass spectrometer and analyzed. With the heater temperature set to 130°C, the developed system detected 1 ng of methamphetamine, 1 ng of amphetamine, 3 ng of 3,4-methylenedioxymethamphetamine, 1 ng of 3,4-methylenedioxyamphetamine, and 0.3 ng of cocaine. Even from mixtures consisting of clove powder and methamphetamine powder, methamphetamine ions were detected by tandem mass spectrometry. The developed probe heating method provides a simple method for the analysis of solid samples. A portable mass spectrometer incorporating this method would thus be useful for the onsite screening of illicit drugs.
The abuse of illicit drugs has become a global problem. A survey by the United Nations Office on Drugs and Crime (UNODC) revealed that over 300 million people worldwide used illicit drugs in 2009.1) One way to prevent such abuse is to conduct onsite screening for such drugs. Such screening is generally done using immunoassay kits,2,3) but the false positive rate for such kits is sometimes problematic due to cross reactions of antibodies.4,5) In addition, it is difficult to prevent the rapid spread of newly synthesized drugs such as synthetic cannabis because it can take as long as a year to generate the antibodies needed to recognize them. The physiological effect of synthetic cannabis is similar to marijuana, leading to an increased use.6) While mass spectrometers give more accurate results than conventional immunoassay kits, they are not generally used for onsite screening because they are not portable due to their size and weight.
Recent efforts to downscale mass spectrometers to make them more portable7–17) have focused on improving sensitivity. Because the pump dominates the weight of mass spectrometers, it would be necessary to use a smaller pump to reduce the weight of the overall system. However, a smaller pump has a lower pumping speed, resulting in a lower sensitivity in an analysis. Several technologies have been developed to overcome this lower sensitivity problem. Contreras et al. (2008) used a toroidal ion trap, a gas chromatograph, and an electron ionization source in a portable mass spectrometer.16) They concentrated the sample using a solid phase micro extraction method, thereby enhancing sensitivity. Ouyang et al. (2009) developed a portable mass spectrometer with a rectilinear ion trap and an electrospray ionization source.7) They developed a discontinuous atmospheric interface that improves the efficiency of ion transfer from an ionization source at atmospheric pressure to a mass analyzer in a vacuum.
Our group developed a prototype of a portable mass spectrometer for use in the screening of illicit drugs as an alternative to an immunoassay kit.18,19) It uses a highly sensitive ionization source, a low-pressure dielectric barrier discharge ionization (LP-DBDI) source. A discontinuous sample gas introduction technique and a vacuum headspace method are used to efficiently transport the sample vapor to the LP-DBDI source. This spectrometer can detect 0.1 ppm of methamphetamine in a liquid, making it sufficiently sensitive for screening illicit drugs. However, the portable mass spectrometer we developed can only be used with a liquid sample. In its Global Synthetic Drugs Assessment 2014, UNODC indicated that the unlawful import of illicit drugs to Asia was increasing.20) The detection of such imported illicit drugs before they fall into the hands of criminals is highly important for the security of our society. This requires a portable mass spectrometer that is capable of analyzing solid samples.
To date, many methods have been developed for the analysis of solid samples by mass spectrometry (MS). Ambient sampling/ionization MS is now widely accepted as having great potential.21,22) Desorption electrospray ionization and direct analysis in real time are the most frequently used techniques.23,24) The thermal desorption of the sample followed by ambient ionization has also been frequently applied. Several ambient ionization methods have been coupled with thermal desorption, such as electrospray ionization,25) atmospheric pressure chemical ionization (APCI),26,27) DBDI,28) and atmospheric pressure photoionization.29) Popov et al. (2005) developed a rapid method for the detection of explosives using thermal desorption and APCI. A cotton swab with explosives absorbed to it is placed in the heater of the APCI source. The vaporized explosive is mixed with air containing CHCl3 vapor and analyzed by APCI-MS.26) An atmospheric pressure solids analysis probe is typically used to analyze solid samples by MS.27) The sample, in a glass melting-point capillary, is inserted into a heated stream of N2 gas, which vaporizes the sample, thus permitting it to be analyzed. Usmanov et al. (2013) recently developed ambient flash desorption MS for solid sample analysis that uses a linearly driven heated metal filament.28) The heated filament is driven up and down so that it touches the sample. The vaporized sample is ionized by collision with metastable helium generated by a flow of helium gas through a DBDI source. Although these methods are useful for solid sample analysis, the setups are too bulky for a portable mass spectrometer.
The probe heating method described here can be easily incorporated in our portable mass spectrometer. It provides a simple method for the rapid analysis of solid samples. In the probe heating method, the user wipes off solid sample by the sampling probe and sets the probe on the heater. A mass spectrum can then be obtained within 1 min. The probe heating method does not require a gas supply and its power consumption is low. Those advantages suggest that the probe heating method is adequate for use in a portable mass spectrometer. Using the method, we measured five illicit drugs, namely, methamphetamine, amphetamine, 3,4-methylenedioxymethamphetamine (MDMA), 3,4-methylenedioxyamphetamine (MDA), and cocaine, because abuse of those drugs is a problem in Japan.
The portable mass spectrometer into which we incorporated the developed probe heating method is shown schematically in Fig. 1. We pressed the tip of the sampling probe, which is an aluminum rod with a diameter of 2 mm and a length of 20 mm, on a solid sample such as a powder or tablet in order to directly attach the sample to the tip. We used aluminum as the sampling probe because of its high thermal conductivity and low cost. The probe was placed on a heater, which was then heated to vaporize the sample. The increase in the heater temperature might result in an increase in signal intensity, but a large the amount of electricity would be needed for this, which is problematic for portable battery-driven instruments. Thus, the heater was typically set to 130°C. The measurement started 3 s after the probe was placed on the heater to allow the temperature of the probe to stabilize. After 3 s, the temperature of the probe was nearly the same as that of the heater. The sample vapor was introduced into the portable mass spectrometer and then analyzed. Because a solid sample stuck to the tip of the probe can cause memory effects, the sampling probe is disposable.
Sample affixed to the sampling probe is vaporized and introduced into the mass spectrometer. Sample introduction is regulated by pinch valve. Sample molecules are ionized by low-pressure dielectric barrier discharge ionization source and analyzed by linear ion trap mass analyzer.
The configuration of the portable mass spectrometer is described in detail in our previous study.19) Because the vacuum headspace method was used in our previous study, the pressure in the sample vial was lower than atmospheric pressure. In this study, the pressure around the sampling probe was at atmospheric pressure. To maintain the appropriate pressure for the LP-DBDI source, we lowered the conductance of the gas entrance port. A PEEK tube with an inner diameter of 0.17 mm and a length of 13 mm was used as a gas entrance port and was connected via a silicone tube and the PEEK tube with an inner diameter of 1.0 mm and a length of 18 mm to a glass tube, which was the LP-DBDI source. A gate valve was placed between the PEEK tube and the glass tube. By closing the gate valve, the PEEK tubes can be exchanged with a silicone tube with the vacuum state of the mass spectrometer being maintained. Between measurements, we exchanged those tubes to avoid sample carry-over. The opening and closing of the silicone tube was controlled by a pinch valve which was normally closed. The sample gas entered the LP-DBDI source only when the pinch valve was opened. Two discharge electrodes to which 3-kV and 10-kHz AC voltages were applied to generate a barrier discharge were placed inside and outside the glass tube. The ions generated by the LP-DBDI source were transported to the analyzer chamber through a 0.5 mm orifice. A compact linear ion trap (LIT) mass analyzer in the analyzer chamber was evacuated by means of a 10-L/s turbomolecular pump (HiPace10, Pfeiffer Vacuum, Asslar, Germany) and a 5-L/min diaphragm pump (MVP006-4, Pfeiffer Vacuum). The rod electrodes were 4.9 mm in diameter and 20 mm in length, and their inscribed diameter was 5.1 mm. Two of the electrodes had a 0.5 mm wide slit and a length of 18 mm to enable ejection of the trapped ions. When a 1.5-MHz RF voltage was applied to the rod electrodes, the ions were trapped in the axial direction by a pseudopotential due to the RF voltage and in the longitudinal direction by the potential due to the difference between the rod DC voltage and the incap–endcap voltage. Resonance phenomena were used to eject the ions by sweeping the frequency of the supplemental AC applied to the rod electrodes from 50 kHz to 500 kHz. The resulting ions were detected by a photomultiplier-tube-based detector.
The portable mass spectrometer operated on the basis of discontinuous sample gas introduction and intermittent ionization. One full mass spectrum could be obtained each 1 s, which means that the pinch valve opens once per about 1 s. Figures 2(A) and (B) show the instrument control sequence and the pressure change in the analyzer chamber, respectively. Such pressure was measured by a pressure gauge (925 Micro Pirani™ vacuum transducer, MKS Instruments, Inc., Andover, MA) connected to the analyzer chamber. The fluctuations in the pressure curve were due to power supply noise in the analog output from the pressure gauge. Electric power for the LP-DBDI was turned on simultaneously with the opening of the pinch valve. After a certain number of milliseconds, an RF voltage was applied to trap the produced ions. The ion accumulation time was when both the RF voltage and electric power were applied. The pressure increased linearly during the valve-open time. After the valve was closed, a cooling time, typically 750 ms, was needed for the pressure inside the analyzer chamber to drop to below 0.01 Pa for mass scanning. A supplemental AC voltage was applied to eject the trapped ions in accordance with their m/z. The incap–endcap voltage was normally −10 V, but if the number of ions entering the LIT needed to be controlled, the incap voltage was adjusted during the ion accumulation time. The rod DC voltage was set to –20 V. In the post-scan time, all of the ions in the LIT were ejected by setting all voltages to ground. The power consumption of our mass spectrometer is about 70 W. That spectrometer can operate with a battery for about 1 h.
Based on the control sequence, one full mass spectrum can be obtained in one second.
Methamphetamine hydrochloride was purchased from Dainippon Sumitomo Pharma (Osaka, Japan), and cocaine hydrochloride was purchased from Takeda Pharmaceutical Company Limited (Osaka, Japan). Amphetamine hydrosulfate, MDA hydrochloride, and MDMA hydrochloride were synthesized in our laboratory. The materials were dissolved in water at a concentration of 10 mg/mL and stored in a refrigerator or a freezer.
Since the sampling probe was heated by placing it on a heater, the temperature of the heater affected the signal intensity. We investigated the relationship between the temperature and intensity for three temperatures: 70°C, 100°C, and 130°C. In the wipe sampling, the amount of powder sample affixed to the probe tip was not constant between the measurements, and this variability obscured the effects of a change in temperature on the signal intensity. We therefore used a methanol solution containing methamphetamine as the sample for this test.
A 1-μL droplet of methanol containing 100 ng/μL of methamphetamine was placed on the tip of the sampling probe and dried. This resulted in a stable amount of sample on the probe. Figure 3(A) shows the mass spectrum and tandem mass spectrometry (MS2) spectrum of the 100 ng methamphetamine when the heater was set to 130°C. The protonated methamphetamine ions at m/z 150 were fragmented into m/z 91 and m/z 119 ions by collision induced dissociation (CID). Although, in this experiment, only methamphetamine was vaporized and introduced into the instrument, numerous small peaks not related to this molecule appeared in the mass spectrum. Similar noise peaks were also detected in our previous study, in which headspace sampling was used in conjunction with our portable mass spectrometer.19) Moreover, we observed those peaks even when only ambient gas was introduced into the instrument. Thus, we hypothesize that noise peaks are derived from molecules vaporized from the components of the instrument such as silicone, glass and PEEK tubes and the analyzer chamber.
Higher temperature induced stronger signal intensity.
Figure 3(B) shows an extracted ion chromatograph (EIC) for m/z 91 ions after CID. As mentioned in the experimental section, the measurement starts 3 s after placing the sampling probe on the heater. After 3 s, the temperature of the probe was nearly the same as that of the heater. However, the EIC intensity showed progressive increase. This must be due to the adsorption of sample molecules on the inner surface of the flow path, especially the silicone tube. As a silicone tube is a porous material, many adsorption spots are possible. Vaporized sample molecules gradually covered those spots, leading to an increase in the number of molecules passing through the silicone tube, resulting in an increase in the EIC intensity. Between measurements, we changed the silicone tube to avoid sample carry-over, and, thus, the sample adsorption process occurring in every measurement.
A higher temperature resulted in a greater signal intensity. This indicates that heating at over 130°C may lead to a highly sensitive measurement. However, heating at a temperature of over 130°C could lead to the thermal decomposition of the sample. As described above, the higher the heating temperature, the greater the amount of electricity required, which can be problematic for portable battery-driven instruments. Thus, we set the temperature of the sampling probe to 130°C in subsequent experiments.
Illicit drug detectionIn addition to the detection of methamphetamine described above, other illicit drugs were detected using the probe heating method, including 100 ng of amphetamine, 300 ng of MDMA, 100 ng of MDA, and 30 ng of cocaine. Methanol solutions, each containing an illicit drug, were used as the samples. Figure 4 depicts the mass spectra and MS2 spectra for these samples. All ions appeared as protonated ions in the mass spectrum. Amphetamine ions were observed at m/z 136, which fragmented into m/z 119 and m/z 91 ions. MDMA ions and MDA ions were detected at m/z 194 and m/z 180, respectively. Both ions fragmented into m/z 163 and m/z 135 ions. Cocaine ions appeared at m/z 304, and the fragment ions were recognized at m/z 182. The fragmentation patterns obtained for these drugs were identical to those obtained in previous studies.30,31) As is the case with methamphetamine, noise peaks appeared in the mass spectrum of amphetamine, MDMA and MDA, showing the existence of unknown molecules vaporized from the components of the instrument.
Selected precursor ions were m/z 136 ions in amphetamine, m/z 194 ions in MDMA, m/z 180 ions in MDA, and m/z 304 ions in cocaine. Obtained MS2 spectrum is shown in the inset.
Figure 5 shows calibration curves for the illicit drugs obtained using the EIC intensity of the fragment ions: m/z 91 ions for methamphetamine and amphetamine, m/z 135 ions for MDMA, m/z 163 ions for MDA, and m/z 182 ions for cocaine. The mean of the EIC intensity for 1 min from the start of the measurement was used for data points. In each concentration, the mean values were calculated from triplicate samples. Standard deviation (SD) is expressed by error bars. As mentioned above, we used a methanol solution of the illicit drugs as the sample to realize a stable amount of sample on the tip of the sampling probe. Thus, the SD is attributable to the instrument. We expect that the portable mass spectrometer coud be used for qualitative analyses of illicit drugs. In such applications, the SD of the instrument is considered to be sufficiently small. The limits of detection (LODs) for these drugs are listed in Table 1. The portable mass spectrometer with the probe heating method was the most sensitive for cocaine. Although the LOD for MDMA was the highest among the five drugs tested, a 3 ng sample was still detected. MDMA is mainly sold in the form of a tablet called ecstasy.32) The amount of stimulant drugs for one use is generally about several 10 mg. This indicates that the sensitivity of our instrument is sufficient to permit illicit drugs that are possessed by a user to be detected.
Curves were obtained from 1 to 100 ng for methamphetamine, amphetamine, and MDA, from 3 to 300 ng for MDMA, and from 0.3 to 30 ng for cocaine. Data are the means obtained from triplicate samples.
| Compound | Amphetamine | Methamphetamine | MDA | MDMA | Cocaine |
|---|---|---|---|---|---|
| m/z (MS1) | 136 | 150 | 180 | 194 | 304 |
| m/z (MS2) | 91, 119 | 91, 119 | 163, 135 | 163, 135 | 182 |
| LOD (ng) | 1 | 1 | 1 | 3 | 0.3 |
In this paper, we investigated the sensitivity of our mass spectrometer for methamphetamine, amphetamine, MDMA, MDA, and cocaine. To consider the use of our portable mass spectrometer in countries other than Japan, other drugs such as marijuana, opiates, and synthetic cannabis would need to be detected. We would therefore need to increase the variety of target drugs to enhance the usability of our instrument.
Evaluation of sample amount affixed to sampling probeIn the practical use of the portable mass spectrometer, the user wipes off the solid samples by the sampling probe. Using the portable mass spectrometer to detect illicit drugs that are in practical use depends on not only the sensitivity of the spectrometer, but also the efficiency of the wipe sampling. Thus, we evaluated the amount of sample affixed to the tip of the probe after wiping. One milligram of methamphetamine powder and that of cocaine powder were individually placed on a paper and the powder was wiped off by the probe. The tip was then dipped in 1.5 mL of methanol in a polypropylene tube. The amount of sample that was transferred to the methanol was quantified using a triple quadrupole mass spectrometer (Quattro Premier, Waters Corp., Milford, MA) with flow injection analysis. For the quantification of methamphetamine and cocaine, their calibration curves were obtained using standard samples.
Figure 6 shows the quantified amount of affixed sample powder: 300 μg for methamphetamine and 700 μg for cocaine. The large difference in the amount might be due to the difference in the grain size of the material. Because sample quantification is not required for onsite screening, this high coefficient of variation is not problematic. When we measured methamphetamine and cocaine powder using the portable mass spectrometer, we obtained mass spectra similar to those obtained with methanol solutions of the drugs (data not shown). The results shown in Fig. 6 indicate that, although the amount of sample affixed to the probe depends on the kind of sample, at least 1 μg is affixed for any drug. As described above, the LOD of our instrument was 3 ng even for MDMA. This means that our portable mass spectrometer with the probe heating method is sufficiently sensitive for the onsite screening of illicit drugs.
Data were obtained from triplicate samples.
In addition to using only high purity samples, we also used a mixture of samples, which must also be screened onsite. For example, illicit drugs are often mixed with plant material. To simulate this condition, we mixed methamphetamine powder with clove powder (Tree of Life, Tokyo, Japan) for use as a sample. The mixture ratio was 1 : 1. The sample was affixed to the sampling probe and analyzed.
Figure 7 shows the mass spectrum, the MS2 spectrum, and the EIC of the m/z 91 ions, which were fragment ions of m/z 150 ions. As shown in the mass spectra, there was a peak corresponding to methamphetamine ion at m/z 150 but also many impurities. However, isolation of the m/z 150 ions and subsequent CID enabled the fragment ions at m/z 119 and m/z 91 in the MS2 spectrum to be recognized. As shown in the EIC, our instrument detected methamphetamine in the mixture sample. This supports our conclusion that our portable mass spectrometer, when used in conjunction with the probe heating method, can be used for the onsite screening of illicit drugs.
Although many impurities were found in the mass spectra, fragment ions of methamphetamine at m/z 91 and 119 were clearly detected in the MS2 spectrum.
Our newly developed probe heating method enables our previously developed portable mass spectrometer to be used for the analysis of solid samples as well as liquid samples. It is demonstrated in this paper that the portable mass spectrometer with the probe heating method can detect methamphetamine, amphetamine, MDMA, MDA, and cocaine with the LODs ranging from 0.3 to 3 ng. It is also shown that illicit drug molecules can be detected in samples containing impurities. These results suggest that the portable mass spectrometer with the probe heating method is well suited for onsite screening of illicit drugs.
This work was supported in part by the “R&D Program for Implementation of Anti-Crime and Anti-Terrorism Technologies for a Safe and Secure Society,” Funds for Integrated Promotion of Social System Reform and Research and Development of the Ministry of Education, Culture, Sports, Science and Technology, Japan.
