Vapor phase detection of hydrogen peroxide with ambient sampling chemi/chemical ionization mass spectrometry

Abstract

In this paper, we demonstrate a mass spectrometric method for sensitive and rapid detection of gaseous hydrogen peroxide (H₂O₂) using an ambient sampling chemi/chemical ionization source. Metastable helium atoms (He*) were used as the primary ionizing agents and were generated from a dielectric barrier discharge (DBD) source. A gaseous sample of ambient air was drawn into an enclosed ionization chamber and subsequently mixed with metastable species at sub-atmospheric pressure. O₂⁻ ions produced from the ionization of ambient air readily react with H₂O₂ to form the stable [H₂O₂ + O₂]⁻ cluster ion. A detection limit of 0.8 ppbv was achieved by coupling the ion source to a commercial time-of-flight mass spectrometer. The construction of the ion source is simple, and vapors from bulky objects can also be analyzed directly by placing the sample close to the sampling nozzle. The versatility of this ion source for the application in rapid security screening is demonstrated by non-invasive detection of an aqueous solution of hydrogen peroxide concealed in closed beverage bottles.

---

Introduction

Due to the increasing incidence of terrorist and criminal activities using home-made explosives which can be easily synthesized from concentrated hydrogen peroxide, a rapid and sensitive detection method for the vapor phase hydrogen peroxide is in urgent need for safety and security screening. Effective and selective detection of hydrogen peroxide vapor is also important for the real time industrial monitoring of toxicity, and the pathological diagnosis of respiratory disease.¹ Standard detection methods for hydrogen peroxide (H₂O₂) are based on the electrochemical reactions at electrodes (amperometric or potentiometric) and their uses are mostly limited to the samples in liquid phase.^2–4 Although it is possible to collect H₂O₂ vapor with flowing liquid, and detect it with conventional electrochemical sensors,^5,6 the detection process is rather slow for real time response. Electrochemical sensors based on agarose-coated Prussian-blue modified carbon transducer,⁷ and phthalocyanine chemiresistors⁸ have been developed recently for the direct detection of hydrogen peroxide vapor. A certain degree of selectivity could be achieved with these sensors due to the reduction and oxidation reaction of H₂O₂ at different electrodes. However, because these methods are sensitive to other oxidizing agents, there is always some possibility of false-positive results. In this respect, mass spectrometric detection of gaseous H₂O₂ is an alternative approach which avoids this problem.

In fact, there is a rich body of literature on the detection of explosives with mass spectrometry,^9–15 however, only very limited work has been done on the detection of gaseous H₂O₂. The first experimental detection of protonated H₂O₂, i.e., [H₂O₂ + H]⁺, dates back to 1975, when Lindinger et al. proposed the presence of gas phase [H₂O₂ + H]⁺ using a flowing afterglow apparatus.¹⁶ Fujii et al. confirmed the existence of [H₂O₂ + H]⁺ in the microwave discharge plasma of the mixed CH₄/O₂ gases but they found that [H₂O₂ + H]⁺ was only a minor product, compared to other ionic products.¹⁷ The detection of [H₂O₂ + H]⁺ with other positive-mode chemical ionization mass spectrometry is not effective,¹⁸ because of the low proton affinity of H₂O₂ (674.5 kJ mol⁻¹).¹⁹

With an atmospheric pressure chemical ion source using dielectric barrier discharge (DBD) as the excitation source,^20,21 we have shown that O₂⁻ ions produced from the ionization of ambient air could react readily with H₂O₂ to form stable [H₂O₂ + O₂]⁻ adduct ions.²¹ A new version of DBD chemical ionization source has also been recently developed in our laboratory, in which, unlike the other methods that use flowing metastable helium in the open space,^12,22–25 the ambient air containing the gaseous sample is drawn into an enclosed ionization chamber, and is subsequently mixed with metastable helium at a sub-atmospheric pressure in front of the ion inlet of the mass spectrometer.²⁶ Volatile compounds evaporated from a sample can easily be detected in real time by positioning the sample close to the sampling nozzle, this can be extend to non-volatile samples by passing a hot carrier gas over the sample, which is then introduced to the sampling nozzle. In this paper, we present the application of this ion source which was optimized for the rapid and sensitive detection of H₂O₂ vapor in the negative ion mode.

Experimental section

Mass spectrometer

The experiments were performed using a commercial orthogonal-acceleration time-of-flight mass spectrometer for LC/MS (AccuTOF, JEOL, Akishima, Japan). The original electrospray ion source was removed and the cone-shape ion-sampling orifice of the AccuTOF was replaced by a custom made vacuum flange with a threaded hole situated at the center, which was coaxial with the skimmer of the mass spectrometer (see Fig. 1a). The configuration of the internal electrodes of the mass spectrometer remained unchanged. The potential differences between the vacuum flange and the skimmer and between the skimmer and ion guide were both set to be 1 V. The pressure in the vacuum stage before the skimmer was 200 Pa.

Fig. 1 a) Schematic of the ambient sampling chemi/chemical ion source using dielectric barrier discharge (DBD) for the generation of metastable helium (He*) atoms. The ionization chamber formed by the 1/4′′ stainless steel T-union (Swagelok) is electrically connected to the vacuum flange of the mass spectrometer. The inner diameters for the sampling nozzle (d) and ion transport channel (D) are 0.5 mm and 4 mm, respectively. The AC H.V. applied to the DBD source is 2 kV peak-to-peak. The pressure in the ionization chamber is 1330 Pa, and the pressure in the first vacuum stage of the mass spectrometer before the skimmer is 200 Pa. b) Photograph of the ion source taken during the application of non-invasive screening of PET bottle for the trace hydrogen peroxide. The leaked hydrogen peroxide was collected using a funnel-like gas collector attached to the sampling nozzle of the ion source.

Ion source

As shown in Fig. 1a, the ion source was consisted of a three-way stainless steel T-union (Nippon Swagelok FST Inc., Hyogo, Japan) that connected the dielectric barrier discharge source and the sampling nozzle to the mass spectrometer. The sampling nozzle was made of polyether ether ketone (PEEK) tube (I.D. 0.5 mm) and was obtained from GL Scientific (Tokyo, Japan). Metastable helium atoms (He*) generated from the DBD were reacted with sample and ambient gases in the junction of the stainless steel T-union. The mixture of gases and ions was transferred into the high vacuum via a metallic channel with an inner diameter (D) of 4 mm.

The dielectric barrier discharge tube was made of high grade quartz tube (Vitreosil) with 6 mm O.D. and 4 mm I.D. The electrodes for the DBD were consisted of a stainless steel connector (inner electrode), and a copper strip (outer electrode) which was adhered on the outer surface of the quartz tube. The sinusoidal signal of 20 kHz generated from a function generator was amplified by a power amplifier and was stepped up by a transformer to an AC high voltage (AC H.V.). The AC H.V. was applied to the outer electrode, and the inner electrode was connected to ground potential. An insulator was used to separate the discharge source from the body of the ion source, and the resultant dielectric barrier discharge was confined within the discharge tube and was not extended into the ionization chamber. Throughout the experiment, the AC H.V. was set at 2000 V peak-to-peak.

The flow rate of helium gas (99.999%) was monitored and controlled by a mass flow controller (Horiba, Kyoto, Japan). During the start-up of the ion source, helium was first flowed at 400 mL min⁻¹ for a few seconds to purge unwanted impurities, and after that, the flow rate was reduced to 100 mL min⁻¹ for typical operation. For the measurement of the operating pressure of the ion source, the T-union was replaced by a four-way cross-union, and the additional channel was connected to a Baratron capacitance manometer (MKS Instruments, Andover, MA). The operating pressure of the ion source was 1330 Pa and the suction rate for ambient and sample gases was 1 L min⁻¹. Direct and real time detection of volatile compounds was conducted by placing the sample close to the sampling nozzle. Optionally, a funnel-like gas collector could also be attached to the sampling nozzle as shown in Fig. 1b to improve the collection of sample gas.

Chemicals and sample preparation

Aqueous solution of hydrogen peroxide (30% w/w) was purchased from Kanto Chemical (Tokyo, Japan). Pure water was prepared using Milli-Q system (Millipore, Bedford, MA, USA). To generate the hydrogen peroxide vapor of known concentration, aliquots (0.5 μL) of diluted H₂O₂ aqueous solutions were loaded to the gas sampling bag (Tedlar, ASONE, Osaka, Japan). The gas sampling bag was filled with nitrogen gas, and the samples were allowed to evaporate and equilibrate in the sealed environment for ∼30 min.

Results and discussion

The operation mechanism of the present ion source is similar to other atmospheric pressure ion sources that utilize flowing He* atoms. In the positive ion mode, the Penning ionization (chemi-ionization) of the ambient gases such as oxygen and water molecules is the primary ionization process to produce reactant ions (e.g. O₂⁺ and protonated water clusters, [(H₂O)_n + H]⁺ for the subsequent chemical ionization process. In the negative ion mode, electrons generated from the reactions of He* with ambient gases and metallic wall (e.g. secondary electron emission) are thermalized by collisions with gas molecules and are subject to electron capture by oxygen to produce oxygen anion O₂⁻. The oxygen anions can further react with other ambient and analyte gases via chemical ionization processes such as proton abstraction, charge transfer, and clustering reactions.

Fig. 2a shows the typical negative mode mass spectrum for the background ions produced by the present ion source. Major peaks are due to ions of O₂⁻, and [H₂O + O₂]⁻. Fig. 2b is the mass spectrum obtained by approaching a plastic beaker containing aqueous solution of 0.3% (w/w) hydrogen peroxide to the sampling nozzle of the ion source. The dominant peak at m/z 66 is originated from the adduct ions of [H₂O₂ + O₂]⁻, and the peak at m/z 100 is due to the dimer of hydrogen peroxide [(H₂O₂)₂ + O₂]⁻.

Fig. 2 a) Negative mode mass spectrum of the background ions produced by the ambient sampling chemi/chemical ionization source. b) Mass spectrum of H₂O₂ obtained by approaching a plastic beaker containing 0.3% (w/w) liquid hydrogen peroxide close to the sampling nozzle of the ion source.

Fig. 3 shows the real time response of the ion source towards the hydrogen peroxide vapor of different concentrations (0.8 ppbv, 4 ppbv, and 100 ppbv). The measurement was performed by inserting the nozzle of the gas sampling bag to the sampling nozzle of the ion source. The extracted ion signal for [H₂O₂ + O₂]⁻ (m/z 66) responded almost instantaneously upon the exposure to hydrogen peroxide vapor and reached the steady state level in less than 2 s. When the gas sampling bag was removed from the sampling nozzle, the ion signal restored to its baseline level in ∼2 s. At the present stage using our AccuTOF time-of-flight mass spectrometer, the detection limit for H₂O₂ vapor is about 0.8 ppbv or less.

Fig. 3 Extracted ion chronogram for [H₂O₂ + O₂]⁻ (m/z 66) ion obtained for hydrogen peroxide vapor of different concentrations: 0.8 ppbv, 4 ppbv, and 100 ppbv. Inset shows the magnified chronogram for the concentrations of 0.8 ppbv and 4 ppbv.

A sensitive and non-invasive method for the detection of hydrogen peroxide concealed within a liquid container or beverage bottle is now urgently needed, particularly for the screening of the liquids at airports. Spectroscopic methods such as spatially offset Raman spectroscopy and infra-red absorption spectroscopy have been proposed for this purpose.^27,28 Alternatively, here, we demonstrate that mass spectrometric method with an appropriately modified ion source may also find application in the rapid and non-invasive detection of liquid hydrogen peroxide contained in the liquid bottle without opening the bottle cap.

Fig. 4a shows six samples of PET (polyethylene terephthalate) bottle drinks in which an aqueous solution of hydrogen peroxide was concealed in the bottles labelled C and E. The H₂O₂ concentration in sample C was 30% w/w, and in E, it was disguised as orange juice by mixing the H₂O₂ (30%) with the original carbonated orange juice at a ratio of 1/1 v/v. During the preparation, the liquid H₂O₂ was carefully poured into the bottle with the help of funnel and the lip was cleaned with wet tissue paper. After sealing tightly, the bottles were thoroughly washed with tap water, and wiped again with clean tissue paper. The samples were allowed to dry for about 5 min and were tested together with other PET bottles containing genuine drinks. A funnel-like gas collector was connected to the sampling nozzle of the ion source, this was constructed to collect the vapor emitted from the tightly sealed bottles more efficiently. The measurement was conducted by placing the top of the bottle into the gas collector as shown in Fig. 1b.

Fig. 4 Non-invasive screening of beverages contained in the PET bottle for the trace of hydrogen peroxide without opening the bottle cap. a) Photograph of six samples (A–F) used in the experiment. Sample C contained 30% w/w of aqueous hydrogen peroxide, and E is a mixture of the original carbonated orange juice with the H₂O₂ (30%) at a ratio of 1/1 v/v. Other bottles contain genuine drinks. b) Chronogram for the total ion count (TIC) recorded during the screening process. c) Extracted ion chronogram for [H₂O₂ + O₂]⁻ at m/z of 66.

Total ion chronogram acquired during a single run of the screening process is depicted in Fig. 4b. Positions of the samples could be identified from small fluctuations of total ion current taking place during the interchange of samples. The extracted ion chronogram for [H₂O₂ + O₂]⁻ is depicted in Fig. 4c, and the samples containing hydrogen peroxide solution can be easily identified from other samples with good contrast. The same measurement cycles were repeated several times and no false positive or negative result was observed.

Conclusion

In summary, we have shown that hydrogen peroxide vapor can be effectively detected with mass spectrometry as [H₂O₂ + O₂]⁻ cluster ion by ambient sampling chemi/chemical ionization. Compared to our previous ion source that used flowing metastable helium in the open space, the sensitivity of the present ion source for the detection of vapor phase hydrogen peroxide was ∼10× higher, and the required flow rate of helium for maintaining stable discharge and sufficient ionization was 5–10× lower. One possibility for the sensitivity enhancement is that, since the chemical ionization takes place in a sub-atmospheric pressure environment right in front of the inlet of the mass spectrometer, the transmission of the generated ions into the detector was expected to be improved. With a detection limit of <1 ppbv, rapid and non-invasive screening of a type of liquid container that is commonly used for hydrogen peroxide was possible, based on the detection of vapor phase hydrogen peroxide. Beside hydrogen peroxide, the ion source can also be applied to detection of other volatile compounds,²⁶ and could be used for the simultaneous detection of anions and cations when interfaced with a mass spectrometer that allows high speed switching of ion detection modes.

Acknowledgements

This work was funded by the Safety/Security S & T Project from the Japanese Ministry of Education, Culture, Sports, Science and Technology. L. C. Chen was supported by the Japan Society for the Promotion of Science (JSPS).

References

1. A. W. Dohlman, H. R. Black and J. A. Royall, Am. Rev. Respir. Dis, 1993, 148, 955–960.
2. M. Shao, Y. Shan, N. Wong and S. Lee, Adv. Funct. Mater., 2005, 15, 1478–1482.
3. A. Salimi, L. Miranzadeh, R. Hallaj and H. Mamkhezri, Electroanalysis, 2008, 20, 1760–1768.
4. H. Liu, M. Wen, F. Zhang, D. Liu and Y. Tian, Anal. Methods, 2010, 2, 143–148.
5. H. Huang, P. K. Dasgupta, Z. Genfa and J. Wang, Anal. Chem., 1996, 68, 2062–2066.
6. H. Huang and P. K. Dasgupta, Talanta, 1997, 44, 605–615.
7. J. Benedet, D. Lu, K. Cizek, J. La Belle and J. Wang, Anal. Bioanal. Chem., 2009, 395, 371–376.
8. F. I. Bohrer, C. N. Colesniuc, J. Park, I. K. Schuller, A. C. Kummel and W. C. Trogler, J. Am. Chem. Soc., 2008, 130, 3712–3713.
9. S. A. McLuckey, G. L. Glish, K. G. Asano and B. C. Grant, Anal. Chem., 1988, 60, 2220–2227.
10. S. Boumsellek, S. H. Alajajian and A. Chutjian, J. Am. Soc. Mass Spectrom., 1992, 3, 243–247.
11. Z. Takats, I. Cotte-Rodriguez, N. Talaty, H. Chen and R. G. Cooks, Chem. Commun., 2005, 1950–1952.
12. R. B. Cody, J. A. Laramee and H. D. Durst, Anal. Chem., 2005, 77, 2297–2302.
13. N. Na, C. Zhang, M. Zhao, S. Zhang, C. Yang, X. Fang and X. Zhang, J. Mass Spectrom., 2007, 42, 1079–1085.
14. C. Mullen, M. J. Coggiola and H. Oser, J. Am. Soc. Mass Spectrom., 2009, 20, 419–429.
15. C. Mayhew, P. Sulzer, F. Petersson, S. Haidacher, A. Jordan, L. Märk, P. Watts and T. Märk, Int. J. Mass Spectrom., 2010, 289, 58–63.
16. W. Lindinger, D. L. Albritton, C. J. Howard, F. C. Fehsenfeld and E. E. Ferguson, J. Chem. Phys., 1975, 63, 5220–5222.
17. T. Fujii, S. Iijima and K. Iwase, Chem. Phys. Lett., 2001, 341, 513–517.
18. P. Spanel, A. M. Diskin, T. Wang and D. Smith, Int. J. Mass Spectrom., 2003, 228, 269–283.
19. E. P. L. Hunter and S. G. Lias, J. Phys. Chem. Ref. Data, 1998, 27, 413–656.
20. L. C. Chen, Y. Hashimoto, H. Furuya, K. Takekawa, T. Kubota and K. Hiraoka, Rapid Commun. Mass Spectrom., 2009, 23, 333–339.
21. L. C. Chen, H. Suzuki, K. Mori, O. Ariyada and K. Hiraoka, Chem. Lett., 2009, 38, 520–521.
22. L. V. Ratcliffe, F. J. M. Rutten, D. A. Barrett, T. Whitmore, D. Seymour, C. Greenwood, Y. Aranda-Gonzalvo, S. Robinson and M. McCoustra, Anal. Chem., 2007, 79, 6094–6101.
23. F. J. Andrade, J. T. Shelley, W. C. Wetzel, M. R. Webb, G. Gamez, S. J. Ray and G. M. Hieftje, Anal. Chem., 2008, 80, 2646–2653.
24. J. D. Harper, N. A. Charipar, C. C. Mulligan, X. Zhang, R. G. Cooks and Z. Ouyang, Anal. Chem., 2008, 80, 9097–9104.
25. Y. Zhang, X. Ma, S. Zhang, C. Yang, Z. Ouyang and X. Zhang, Analyst, 2009, 134, 176–181.
26. L. C. Chen, Z. Yu, H. Furuya, Y. Hashimoto, K. Takekawa, H. Suzuki, O. Ariyada, K. Hiraoka. Submitted to J. Mass Spectrom., 2010.
27. C. Eliasson, N. A. Macleod and P. Matousek, Anal. Chem., 2007, 79, 8185–8189.
28. H. Itozaki, Y. Yamauchi. Liquid explosive detection from outside of the bottle by IR. In Infrared Technology and Applications XXXV. SPIE: Orlando, FL, USA, 2009.

---