Letter pubs.acs.org/ac
Photoionization-Generated Dibromomethane Cation Chemical Ionization Source for Time-of-Flight Mass Spectrometry and Its Application on Sensitive Detection of Volatile Sulfur Compounds Jichun Jiang,†,‡ Yan Wang,†,‡ Keyong Hou,† Lei Hua,† Ping Chen,† Wei Liu,†,‡ Yuanyuan Xie,† and Haiyang Li*,† †
Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian, 116023, People’s Republic of China ‡ Graduate University of Chinese Academy of Sciences, 19 Yuquan Road, Beijing, 100039, People’s Republic of China S Supporting Information *
ABSTRACT: Soft ionization mass spectrometry is one of the key techniques for rapid detection of trace volatile organic compounds. In this work, a novel photoionization-generated dibromomethane cation chemical ionization (PDCI) source has been developed for time-of-flight mass spectrometry (TOFMS). Using a commercial VUV lamp, a stable flux of CH2Br2+ was generated with 1000 ppmv dibromomethane (CH2Br2) as the reagent gas, and the analytes were further ionized by reaction with CH2Br2+ cation via charge transfer and ion association. Five typical volatile sulfur compounds (VSCs) were chosen to evaluate the performance of the new ion source. The limits of detection (LOD), 0.01 ppbv for dimethyl sulfide and allyl methyl sulfide, 0.05 ppbv for carbon disulfide and methanthiol, and 0.2 ppbv for hydrogen sulfide were obtained. Compared to direct single photon ionization (SPI), the PDCI has two distinctive advantages: first, the signal intensities were greatly enhanced, for example more than 10-fold for CH3SH and CS2; second, H2S could be measured in PDCI by formation [H2S + CH2Br2]+ adduct ion and easy to recognize. Moreover, the rapid analytical capacity of this ion source was demonstrated by analysis of trace VSCs in breath gases of healthy volunteers and sewer gases.
O
applications in analysis of mixtures of polyatomic molecules.10 Therefore, new reactant ions, which have wider ionization ability, less fragmentations, and easier recognizable adduct ions are always demanded. Besides, simple and efficient method to produce high-purity reactant ions is also necessary. PI is another widely used soft ionization method for online mass spectrometry.11−22 It is suitable for pressure from vacuum to atmospheric pressure. Light sources, such as synchrotron radiation-based,23 laser-based,24 and low pressure discharge VUV lights,25 were mostly utilized. Among them, low pressure discharge VUV lamp is much preferable for online measurement of VOCs due to its compact nature and low cost. However, these VUV lamps suffer from poor sensitivity due to its low ionization cross sections for common VOCs, and the sensitivity is only at a few hundred parts per billion (ppb).26,27 To increase the sensitivity, Li et al.28 used photoelectron to generate O2+ to ionize the analytes by charge-transfer reactions (CTI). Compared to single photon ionization (SPI), the CTI
ver the past decade, soft ionization mass spectrometry is increasingly utilized as one of the most powerful techniques for rapid detection of trace volatile organic compounds (VOCs) in a complex matrix. The widely used soft ionization ion sources for mass spectrometry (MS) are chemical ionization (CI) and photoionization (PI). The most representative CI-MS are proton transfer reaction mass spectrometry (PTR-MS)1,2 and selected ion flow tube mass spectrometry (SIFT-MS).3,4 Both of these two techniques use reactant ions H3O+, NO+, or O2+ to ionize trace VOCs to achieve very high sensitivity and have been widely used in atmospheric chemistry,5 food science,6 and medical applications7 during the past decades. Although PTR-MS and SIFTMS are successful soft techniques, there are some shortcomings. The H3O+ reactant ions is difficult to ionize compounds with low proton affinity (PA).8 Only molecules with ionization energy (IE) less than 9.26 eV can be ionized by charge transfer collisions with NO+ (IE, 9.26 eV) reactant ions, in addition, the adduct product ion [M + NO]+ formed by association reaction is difficult to distinguish from other product ions.9 When using O2+ (IE, 12.07 eV) precursor ions, fragment ions usually generated, which limits its © 2016 American Chemical Society
Received: January 31, 2016 Accepted: April 23, 2016 Published: April 23, 2016 5028
DOI: 10.1021/acs.analchem.6b00428 Anal. Chem. 2016, 88, 5028−5032
Letter
Analytical Chemistry
Figure 1. Schematic diagram of the photoionization-generated dibromomethane cation chemical ionization time-of-fight mass spectrometer.
a central hole of 1 mm diameter). Four 0.5 mm thick PTFE insulation washers were used to separate these electrodes. All the electrodes from the repelling electrode to orifice electrode were connected with 1 MΩ resistances in turn. Direct current voltages of 13 V (V1) and 12 V (V2) were applied onto the repelling electrode and the orifice electrode, respectively, and form a uniform electric field along the axis direction of the 19 mm ionization region. Gas-phase analytes were directly introduced into the ionization region through a 250 μm i.d. 0.4 m long stainless steel capillary (SSC1), while the reaction reagent in a 10 mL headspace stainless steel bottle were carried by compressed air through a two-way ball valve into another 250 μm i.d. 0.2 m long stainless steel capillaries (SSC2). The gaseous analytes flowed through the window of the VUV lamp along the axis direction of the ionization region and vented out at the gas vent, which is located at the last transmitting electrode. The pressure in the ion source was set at 500 Pa by adjusting a flapper valve, which was fixed between a 3.5 L/s dry scroll pump (Agilent Technologies Inc., California) and the ion source chamber. The ion source could operate in two modes, i.e., SPI and PDCI, by simply switching the VA1. In SPI mode, VA1 was turned off, and the analytes were merely ionized by the VUV photons. In PDCI mode, VA1 was turned on, while reagent gas of dibromomethane entered into the ion source, and then photoionization-generated CH2Br2+ induced the PDCI. Of course, the PDCI cannot be operated independently in the presence of VUV lamp, few analytes were also ionized by the VUV photons, while the major part of the analytes were ionized through chemical reaction with CH2Br2+. rf-Only Quadruple TOFMS. The mass spectrometer used in the experiments was a home-built rf-only quadruple orthogonal acceleration reflectron TOF-MS (see the Supporting Information S-1). A mass resolution of 3000 (fwhm) at m/z 78 was achieved with a 0.4 m field-free drift tube. The TOF signals were recorded by a 100 ps time-to-digital converter (TDC) (model 9353, Ametec Inc., Oak Ridge) at a repetition rate of 50 kHz. All the mass spectra were accumulated for 50 s, and all the data were obtained by averaging results from five parallel measurements.
ion source obtained a higher sensitivity (3 ppbv for aromatics). However, the photoelectron emission efficiency will decrease due to oxidation of the electrode during long-term measurement, so a self-adjustment algorithm was used to stabilize the reactant ions for accurate quantitative analysis.29 Furthermore, they added a radio frequency (rf) field into the CTI source to enhance the ionization efficiency, and the sensitivity was further improved by more than 10-fold compared to the CTI ion source.30 Whereas, these CTI sources with O2+ precursor ions faced with the same problem as mentioned above. Volatile sulfur compounds (VSCs) is a kind of odor gas, and even a trace level could cause unpleasant smells and annoy people.31,32 Moreover, some of the VSCs usually are highly toxic, and ppbv level VSCs can cause health problems to humans.33,34 In addition, VSCs can be disease markers in exhaled breath gas for medical diagnostics (e.g., hydrogen sulfide for halitosis and dimethyl sulfide for hepatic coma, etc.).35−39 However, because of its high reactive and adsorptive properties, losses are associated within storage and analysis processes.40 PTR-MS and SIFT-MS methods have been developed for online analysis of VSCs at the ppb level. Different precursor ions, such as H3O+ for CH3SH, NO+ for C2H6S2, and O2+ for C2H6S, were required to reduce MS spectrum complexity. Even so, it was still not possible to quantify CS2 under 50 ppb in the presence of carbon dioxide and acetone.9 Therefore, an efficient method for rapid analysis of trace VSCs is highly demanding. In this study, we introduce a VUV photoionization-generated dibromomethane cation (CH2Br2+) source for further chemical ionization of VSCs and its analytical performances to measure VSCs in breath gas of healthy volunteers and sewer gas.
■
INSTRUMENTATION
VUV-Lamp-Based Photoionization-Generated Dibromomethane Cation Chemical Ionization (PDCI). The PDCI ion source as shown in Figure 1a includes a commercial available VUV krypton discharge lamp (PKS106, Cathodeon Ltd., Cambridge, U.K.) and a chemical reaction ionization region. The VUV lamp with the photon flux of about 1 × 1011 photons/s was installed on the top of the ionization, which was used to produce the reactant ions. The ionization region consisted of five stainless steel electrodes: a repelling electrode (4 mm length, 13 mm i.d., 42 mm o.d., and a 6 mm central hole in the front panel), three circular transmitting electrodes (4 mm length, 13 mm i.d., 30 mm o.d.), and an orifice electrode (with
■
EXPERIMENTAL SECTION Sample Preparation Method. The introduction of sample preparation methods were shown in the Supporting Information S-2. 5029
DOI: 10.1021/acs.analchem.6b00428 Anal. Chem. 2016, 88, 5028−5032
Letter
Analytical Chemistry
■
RESULTS AND DISCUSSION Ionization of VSCs with PDCI. Figure 2 illustrates the mass spectra of the reagent gas and gas mixture of 20 ppbv
While in SPI mode, as shown in Figure 2c, the signal intensities of CH3SH, C2H6S, CS2 ,and C4H8S were only 493, 1408, 142, and 1465 counts. Unfortunately, H 2S was undetected due to the low photoionization cross section and the mass discrimination of low mass range in the rf-only quadrupole transmission. Therefore, the signal intensities of molecular ion M+ of CH3SH+, C2H6S+ CS2+, and C4H8S+ in PDCI were 11.3, 6.9, 13.9, and 4.3-fold greater than those obtained in SPI with the same VUV lamp, respectively. Moreover, owing to the isotopic abundance pattern of element Br, the formation of adduct ion [M + CH2Br2]+ could not only avoid the mass discrimination effect to low mass ions due to the quadruple ion transmission but also help to distinguish adduct ions from other ions in the mass spectrum. LODs, Linear Dynamic Range, and Stability. The analytical performances of the PDCI ion source for the five VSCs were evaluated, and the results were summarized in Table 1. The linear range was almost three-orders of magnitude, 0.2− 200 ppbv for methanthiol, dimethyl sulfide, and allyl methyl sulfide, 0.4−500 ppbv for carbon disulfide, and 2−1000 ppbv for hydrogen sulfide, respectively. The correlation coefficients were in the range of 0.9977−0.9990. The LODs (S/N = 3) ranged from 0.01 to 0.2 ppbv. The stability of an ion source is important for online quantitative analysis and long-term monitoring, the flux of CH2Br2+ cation in PDCI was very stable, the relative standard deviations (RSD) of CH2Br2+ and [H2 S + CH 2 Br2 ]+ intensities were 0.90% and 5.08%, respectively, within 6 h of continuous measurement of H2S with PDCI ion source (see Figures S-4 and S-5 of the Supporting Information). Applications. VSCs are significant disease markers in the exhaled breath of human beings. PDCI-TOFMS was applied to measure VSCs in the breath gas of healthy persons. We used a 3L Tedlar bag to collect the total breath gas of healthy volunteers, all of them first took a deep breath, then exhaled completely into the bag. The background mass spectra of the sampling bag were acquired by filling with the room air. Figure 3a displays the mass spectra of exhaled breath of one of the volunteers, and four of the above VSCs, i.e., H2S, CH3SH, C2H6S and C4H8S, were observed. Moreover, 21 healthy volunteers were participated in this experiment and a statistical analysis of the concentrations of VSCs was obtained. It is seen that the breath concentration of H2S was relatively high at typically 10−40 ppbv while CH3SH, C2H6S and C4H8S were at lower levels of 10 ppbv, as depicted in Figure 3b. Besides, 3 L of sewer gas was collected in a Tedlar bag by sampling pump from one of the sewers in Dalian Institute of Chemical Physics. Figure 4a shows the mass spectra of sewer gas sample obtained by PDCI source; the concentration of VSCs were determined as 482, 46, 4, 2, and 4 ppbv for H2S, CH3SH, C2H6S, CS2, and C4H8S, respectively. The sewer gas samples at various times during 1 day were also analyzed, as shown in Figure 4b. The results show that H2S and CH3SH
Figure 2. Mass spectra of the reagent ions (a) and mass spectra in (b) PDCI mode and (c) SPI mode for gas mixture of 20 ppbv dimethyl sulfide and allyl methyl sulfide, 100 ppbv methanthiol and carbon disulfide, and 500 ppbv hydrogen sulfide.
dimethyl sulfide (C2H6S, IE = 8.69 eV) and allyl methyl sulfide (C4H8S, IE = 8.7 eV), 100 ppbv methanthiol (CH3SH, IE = 9.439 eV) and carbon disulfide (CS2, IE = 10.07 eV), and 500 ppbv hydrogen sulfide (H2S, IE = 10.457 eV). As shown in Figure 2a, 1000 ppmv dibromomethane was chosen as reagent gas (see the Supporting Information S-3), and CH2Br2+ and (CH2Br2)2+ were generated and induced the PDCI. Both of CH2Br2+ and (CH2Br2)2+ participated in the chemical reactions, while the latter provided less contribution due to lower concentration and decrease of ionization energy of dimer ions.41 In PDCI mode, as shown in Figure 2b, the molecular ion M+ peaks for CH3SH, C2H6S, CS2 and C4H8S were obtained through charge-transfer reactions with signal intensities of 5577, 9732, 1976, and 6346 counts, respectively. Besides, H2S, CH3SH, and CS2 can be ionized to generate [M + CH2Br2]+ through CH2Br2+ attachment. All of the mass spectra in PDCI mode were obtained by deducting the background while retain the peaks of reactant ions. CH 2Br2 + hv → CH 2Br2+
(1)
CH 2Br2 + CH 2Br2+ → (CH 2Br2)2+
(2)
M + hv → M+
(3)
+
+
M + CH 2Br2 → M + CH 2Br2 +
(4)
+
M + (CH 2Br2)2 → M + CH 2Br2
(5)
M + CH 2Br2+ → [M + CH 2Br2]+
(6)
Table 1. Characteristic Peaks, Linear Dynamic Range and LODs of Five Kinds of VSCs in PDCI Mode compounds hydrogen sulfide methanthiol dimethyl sulfide carbon disulfide allyl methyl sulfide
formula (M) H2S CH3SH C2H6S CS2 C4H8S
characteristic peaks +
[M + CH2Br2] M+/[M + CH2Br2]+ M+ M+/[M + CH2Br2]+ M+
linear range (ppbv)
correlation coefficient (R2)
LODs (S/N = 3) (ppbv)
2−1000 0.2−200 0.2−200 0.4−500 0.2−200
0.9990 0.9980 0.9987 0.9978 0.9977
0.2 0.046 0.010 0.054 0.013
5030
DOI: 10.1021/acs.analchem.6b00428 Anal. Chem. 2016, 88, 5028−5032
Letter
Analytical Chemistry
Figure 3. (a) Mass spectra of breath gas of a healthy volunteer (male, age 31). (b) Box-whisker plots of median concentrations of four common VSCs observed of 21 healthy volunteers.
Figure 4. (a) Mass spectra of sewer gas. (b) The concentrations of five VSCs in sewer gas at different times of 1 day (August 14, 2015, DICP, China).
■
were the main parts of VSCs in sewer gas, and the concentrations of these two VSCs were related to the behavior of the lab staffs, higher during work time while lower after working hours. It should be noted that the calibration curve of concentration was acquired under the condition of relative humidity of 100% for the two above-mentioned applications.
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b00428. rf-only quadruple TOFMS; sample preparation method; selection of reaction reagent; linear calibration curves for H2S, CH3SH, C2H6S, CS2, and C4H8S; and peak intensity of reactant ions CH2Br2 + and [H2S + CH2Br2]+ ions of 300 ppbv H2S vs time for PDCI ion source (PDF)
■
CONCLUSION In this work, a novel PDCI source based on a commercial VUV lamp was developed for an rf-only quadrupole TOFMS. A stable and high purity of reactant ion CH2Br2+ was produced by VUV photoionization of 1000 ppmv dibromomethane reagent gas. Compared to SPI, the sensitivity of PDCI is enhanced more than 13-fold for molecular ions of CS2. In addition, H2S, CH3SH, and CS2 could form [M + CH2Br2]+ ions, which is easy to distinguish from other ions in the mass spectrum by the apparent isotopic ratios of Br element. The limits of detection were about 0.01 ppbv for dimethyl sulfide and allyl methyl sulfide, 0.05 ppbv for carbon disulfide, and 0.2 ppbv for hydrogen sulfide. This source has been successfully applied to investigate the concentration ranges of different VSCs in breath gas of health volunteers and the daily concentration variations of VSCs in sewer gas. Its good sensitivity and rapid analysis speed imply that the PDCI ion source has a huge potential for medical diagnostics and environmental chemistry.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Fax: +86-411-84379517. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Financial support from the National Special Fund for Development of Major Research Equipment and Instrument (Grants 2011YQ05006904), NSF of China (Grants 21375129 and 21275143), and the National High-Tech Research and Development Plan (Grant No. 2014AA06A501) are gratefully acknowledged. 5031
DOI: 10.1021/acs.analchem.6b00428 Anal. Chem. 2016, 88, 5028−5032
Letter
Analytical Chemistry
■
(36) Whiteman, M.; Armstrong, J. S.; Chu, S. H.; Jia-Ling, S.; Wong, B. S.; Cheung, N. S.; Halliwell, B.; Moore, P. K. J. Neurochem. 2004, 90, 765−768. (37) Sehnert, S. S.; Jiang, L.; Burdick, J. F.; Risby, T. H. Biomarkers 2002, 7, 174−187. (38) Hisamura, M. Nippon Naika Gakkai Zasshi 1979, 68, 1284− 1292. (39) Chen, S.; Zieve, L.; Mahadevan, V. J. Lab. Clin. Med. 1970, 75, 628−635. (40) Nielsen, A. T.; Jonsson, S. Analyst 2002, 127, 1045−1049. (41) Kim, M.; Zoerb, M.; Campbell, N.; Zimmermann, K.; Blomquist, B.; Huebert, B.; Bertram, T. Atmos. Meas. Tech. 2016, 9, 1473−1484.
REFERENCES
(1) Lindinger, W.; Jordan, A. Chem. Soc. Rev. 1998, 27, 347−375. (2) Blake, R. S.; Monks, P. S.; Ellis, A. M. Chem. Rev. 2009, 109, 861−896. (3) Smith, D.; Španěl, P. Mass Spectrom. Rev. 2005, 24, 661−700. (4) Španěl, P.; Smith, D. Mass Spectrom. Rev. 2011, 30, 236−267. (5) de Gouw, J.; Warneke, C. Mass Spectrom. Rev. 2007, 26, 223− 257. (6) Biasioli, F.; Gasperi, F.; Yeretzian, C.; Märk, T. D. TrAC, Trends Anal. Chem. 2011, 30, 968−977. (7) Schwarz, K.; Filipiak, W.; Amann, A. J. Breath Res. 2009, 3, 027002. (8) Španěl, P.; Smith, D. Int. J. Mass Spectrom. 1998, 181, 1−10. (9) Pysanenko, A.; Španěl, P.; Smith, D. J. Breath Res. 2008, 2, 046004. (10) Diskin, A. M.; Wang, T.; Smith, D.; Španěl, P. Int. J. Mass Spectrom. 2002, 218, 87−101. (11) Hou, K.; Wang, J.; Li, H. Rapid Commun. Mass Spectrom. 2007, 21, 3554−3560. (12) Gamez, G.; Zhu, L.; Schmitz, T. A.; Zenobi, R. Anal. Chem. 2008, 80, 6791−6795. (13) Hanley, L.; Zimmermann, R. Anal. Chem. 2009, 81, 4174−4182. (14) Saraji-Bozorgzad, M. R.; Eschner, M.; Groeger, T. M.; Streibel, T.; Geissler, R.; Kaisersberger, E.; Denner, T.; Zimmermann, R. Anal. Chem. 2010, 82, 9644−9653. (15) Wu, Q.; Hua, L.; Hou, K.; Cui, H.; Chen, P.; Wang, W.; Li, J.; Li, H. Int. J. Mass Spectrom. 2010, 295, 60−64. (16) Eschner, M. S.; Selmani, I.; Gröger, T. M.; Zimmermann, R. Anal. Chem. 2011, 83, 6619−6627. (17) Wu, Q.; Hua, L.; Hou, K.; Cui, H.; Chen, W.; Chen, P.; Wang, W.; Li, J.; Li, H. Anal. Chem. 2011, 83, 8992−8998. (18) Cui, H.; Hua, L.; Hou, K.; Wu, J.; Chen, P.; Xie, Y.; Wang, W.; Li, J.; Li, H. Analyst 2012, 137, 513−518. (19) Hou, K.; Li, F.; Chen, W.; Chen, P.; Xie, Y.; Zhao, W.; Hua, L.; Pei, K.; Li, H. Analyst 2013, 138, 5826−5831. (20) He, S.; Cui, H.; Lai, Y.; Sun, C.; Luo, S.; Li, H.; Seshan, K. Catal. Sci. Technol. 2015, 5, 4959−4963. (21) Hua, L.; Hou, K.; Chen, P.; Xie, Y.; Jiang, J.; Wang, Y.; Wang, W.; Li, H. Anal. Chem. 2015, 87, 2427−2433. (22) Xie, Y.; Chen, P.; Hua, L.; Hou, K.; Wang, Y.; Wang, H.; Li, H. J. Am. Soc. Mass Spectrom. 2016, 27, 144−152. (23) Li, J.; Cai, J.; Yuan, T.; Guo, H.; Qi, F. Rapid Commun. Mass Spectrom. 2009, 23, 1269−1274. (24) Hanna, S.; Campuzano-Jost, P.; Simpson, E.; Robb, D.; Burak, I.; Blades, M.; Hepburn, J.; Bertram, A. Int. J. Mass Spectrom. 2009, 279, 134−146. (25) Kuribayashi, S.; Yamakoshi, H.; Danno, M.; Sakai, S.; Tsuruga, S.; Futami, H.; Morii, S. Anal. Chem. 2005, 77, 1007−1012. (26) Syage, J. A.; Hanning-Lee, M. A.; Hanold, K. A. Field Anal. Chem. Technol. 2000, 4, 204−215. (27) Syage, J. A.; Nies, B. J.; Evans, M. D.; Hanold, K. A. J. Am. Soc. Mass Spectrom. 2001, 12, 648−655. (28) Hua, L.; Wu, Q.; Hou, K.; Cui, H.; Chen, P.; Wang, W.; Li, J.; Li, H. Anal. Chem. 2011, 83, 5309−5316. (29) Xie, Y.; Hua, L.; Hou, K.; Chen, P.; Zhao, W.; Chen, W.; Ju, B.; Li, H. Anal. Chem. 2014, 86, 7681−7687. (30) Chen, P.; Hou, K.; Hua, L.; Xie, Y.; Zhao, W.; Chen, W.; Chen, C.; Li, H. Anal. Chem. 2014, 86, 1332−1336. (31) Vassilakos, C.; Papadopoulos, A.; Lahaniati, M.; Maggos, T.; Bartzis, J.; Papagianakopoulos, P. Fresenius Environ. Bull. 2002, 11, 516−521. (32) Li, K.-C.; Shooter, D. Int. J. Environ. Anal. Chem. 2004, 84, 749− 760. (33) Andersson, F. A. T.; Karlsson, A.; Svensson, B. H.; Ejlertsson, J. J. Air Waste Manage. Assoc. 2004, 54, 855−861. (34) Lestremau, F.; Andersson, F. A. T.; Desauziers, V.; Fanlo, J.-L. Anal. Chem. 2003, 75, 2626−2632. (35) Yaegaki, K. Japanese Dental Science Review 2008, 44, 100−108. 5032
DOI: 10.1021/acs.analchem.6b00428 Anal. Chem. 2016, 88, 5028−5032
