Differential Mobility Spectrometry for Inorganic ... - ACS Publications

Article pubs.acs.org/ac

Differential Mobility Spectrometry for Inorganic Filtration in Nuclear Forensics Spiros Manolakos,† Francy Sinatra,† Leila Albers,† Kevin Hufford,† James Alberti,† Erkinjon Nazarov,† and Theresa Evans-Nguyen*,†,‡ †

Draper Laboratory, 3802 Spectrum Boulevard, Suite 201, Tampa, Florida 33612, United States The University of South Florida, Department of Chemistry, 4202 East Fowler Avenue, CHE 205, Tampa, Florida 33620, United States

‡

S Supporting Information *

ABSTRACT: Differential mobility spectrometry (DMS) is applied to the analysis of inorganic mixtures relevant to nuclear forensics. Three primary components of potential radiological dispersal devices (RDDs), cobalt, cesium, and strontium, were studied by DMS to demonstrate rapid sample cleanup when coupled to mass spectrometry. Nanosprayed salt solutions comprised of stable analogs, as proxies to these radioisotopes, and isobaric interferents were introduced to DMS. The DMS effluent was directly coupled to a mass spectrometer to confirm the elemental identity of the separated clusters. DMS dispersion plots demonstrated distinctive elemental separation from both atomic and molecular interferents. These results support the potential use of DMS as a means of rapid separation for inorganic analyses prior to analysis in a field portable mass spectrometer. The mechanism for this process is speculated to involve dynamics of solvent cluster formation under the influence of the alternating high and low electric fields of the DMS.

F

metal cations.13 Thus, DMS may feasibly enhance the operation of a portable mass spectrometer by performing rapid on-site sample cleanup of isobaric interferences in other inorganic analyses. Herein, we describe DMS-MS analysis of cobalt, strontium, and cesium cations from simple salt mixtures with isobaric interferent analogs. A nanoelectrospray ionization source is used to probe the role of clustering phenomena in DMS separation of these representative inorganic species. The nature of the ESI process involves successive desolvation events and has thus been used to study gaseous metal ion solvation chemistry extensively in the literature.14−19 A model explaining the separative power of DMS suggests that neutral clustering and declustering dictate collision cross section and thus ion behavior in the alternating low and high fields, respectively.20−26 We apply this model to demonstrate the utility of DMS in the measurement of metal cations by mass analysis in a fieldable setting.

or elemental mass analysis, sample preparation, and chromatographic separation are time-consuming but crucial steps to mitigate sample “matrix” effects such as isobaric interferences.1 Furthermore, in a laboratory setting, commercial instrumentation employ dedicated collision-reaction cells using various selective ion−molecule reaction chemistries to reduce chemical noise.2 In forensic applications, resulting precise isotopic abundance ratio measurements serve confirmatory purposes but are impractical for initial on-site screening of evidence. In the detonation of improvised nuclear devices or radiological dispersal devices (RDDs, also known as “dirty bombs”), rapid characterization of radiological components can significantly aid in attribution. Identification of the primary agent (e.g., cobalt-60, strontium-90, and cesium-137) and its relative abundance would additionally serve to assess the environment for the immediate threat to human life. While portable mass spectrometry is an ideal screening tool, there is a lack of accompanying sample cleanup methods that focus on elemental analysis.3 Ion mobility spectrometry (IMS) is a readily fieldable analytical technique. IMS has been used to separate atomic ions and clusters4,5 and even toward nuclear applications (e.g., uranyl acetate and strontium chloride).6 A dynamic variant of IMS, differential mobility spectrometry (DMS) has been used widely for rapid front-end filtration prior to mass analysis.7−10 On the order of milliseconds, DMS selectively passes ions based on differences in mobility observed between high and low electric fields. However, DMS applications for atomic ions have primarily focused on halogens.11,12 We have reported earlier work exploring the use of DMS to distinguish small © 2016 American Chemical Society

■

METHODS The planar DMS sensor was fabricated from two parallel ceramic plates with rectangular electrodes produced by metal deposition featuring dimensions of 3.0 mm × 10.0 mm. The plates were separated by Teflon spacers 0.5 mm thick to maintain the filter gap defining the flow channel. As shown in Figure 1, the plates were housed in an interface designed to Received: April 12, 2016 Accepted: October 7, 2016 Published: October 7, 2016 11399

DOI: 10.1021/acs.analchem.6b01441 Anal. Chem. 2016, 88, 11399−11405

Article

Analytical Chemistry

flow of 1250 mL/min across the DMS achieved by the MS vacuum. The inlet was first designed to mimic the existing electrospray interface (Agilent Jetstream). To mitigate problems of saturation, the lower flow of a nanoelectrospray ionization needle was ultimately implemented. The heated nitrogen curtain gas from the TOF-MS was set to 200 °C, 4 L/ min, and routed through multiple channels in the Vespel body surrounding the DMS channel to passively heat the system. A cone-shaped inlet was fabricated in stainless steel to cap the DMS channel ahead of the nanospray. An additional aluminum spacer was also fabricated as a standoff for the ionization source housing to accommodate the added path length of ions transported through the DMS. RF dispersion and DC compensation voltages were supplied to the analytical plates using pogo-pin connectors and electronics from an SVAC model DMS unit (Sionex Corporation). Because the design of the Agilent electrospray source holds the inlet capillary at high voltage, modification of the SVAC electronics was necessary. The entire DMS system electronics were floated using isolating components including fiber optic couplers to the control electronics and a power transformer. Thus, typically, the entire SVAC unit was floated at −1.8 kV with an external supply. The biasing scheme outlined in Figure 2 allowed us to overcome a field discontinuity along the DMS channel from the ion source to the mass spectrometer inlet. Solutions of CsCl, BaCl2, RbCl, SrCl2, CoCl2, CuCl2, NiCl2, and Ca(OH)2 (Fisher Scientific) were prepared at concentrations of 10 μM in 75/25 HPLC grade acetonitrile/water (Thermo Fisher). Mixture solution samples were infused into the DMS-MS platform at the rate of 300 nL/min using a Harvard Apparatus (Hamden, CT, U.S.A.) syringe pump. A coated, 10 μm PicoTip emitter (New Objective, Woburn, MA) was held at instrument ground, while the nominal mass spectrometer nanoelectrospray voltage differential was typically set at −1.8 kV. Successful

Figure 1. DMS-MS (a) inlet design, (b) exploded diagram, and (c) final implementation. The last iteration of the inlet design was made to mimic the commercial nanospray configuration.

accommodate a commercial atmospheric sampling time-offlight mass spectrometer, the Agilent 6230. The interface was machined from chemically inert Vespel featuring pogo-pin connectors and housing demountable ceramic chips to facilitate cleaning. Thermally stable Kelrez o-rings maintained a leak free

Figure 2. RF voltage supplied from the SVAC required an offset bias voltage to operate with the Agilent MS. Using an optical isolator and transformer to decouple the large RF voltages, the entire SVAC unit was raised to the voltage of the inlet capillary to avoid electric failure. 11400


Article


Figure 3. Extracted ion dispersion plots of a mixture containing Ca(OH)2, CoCl2, and NiCl2: (a) calcium hydroxide, extracted mass of m/z 57; (b) singly charged cobalt, extracted mass of m/z 59; (c) singly charged nickel, extracted mass m/z 58. (d) Normalized, extracted ion (transmission plots) DMS spectra mapped to applied compensation voltages at 1000 Vrf.

Vc was scanned from −30 to +10 V, in 0.2 V increments over a period of 50 s. Extracted ion chromatograms for rubidium and strontium were aligned according to the Vc scan sequence (implemented by the Expert software) in postacquisition processing in OriginPro.

transmission of ions from the DMS to the mass analyzer required the use of a postinlet acceleration voltage, otherwise known as the “fragmentor” for Agilent instruments. This voltage was typically held over 200 V to allow for dissociation of clusters prior to detection in the TOF analyzer region. Sionex Expert software was used to control the DMS electronics to acquire comprehensive dispersion plots (Vrf × Vc). For the mixtures containing cesium and cobalt, the dispersion voltages (Vrf) were scanned from 500 to the maximum value of 1500 V amplitude, while the compensation voltage (Vc) was scanned from −43 to +15 V to determine the voltage apex for the DMS-separated species. For each run, the Sionex control software was synchronized to the Agilent Mass Hunter acquisition software. Extracted ion signals were collected from relevant masses with 0.1 m/z windows around the ion of interest. Dispersion and transmission plots of extracted ions were aligned according to the applied Vrf and Vc sequence parameters dictated by the Sionex software using Labview 2013. These plots are presented using OriginPro 9.0 (Northampton, MA). For the specific mixture of strontium with rubidium, a constant Vrf scan was conducted. While Vrf was held at 1350 V,

■

RESULTS AND DISCUSSION Cobalt. Possible chemical interference in the identification of cobalt-60, when analyzed by ICP-MS, include the atomic ion of nickel-60 and the molecular hydroxide ion of calcium-43. Therefore, we collected experimental dispersion plots through the DMS-MS for the nanosprayed solution mixture of CoCl2, NiCl2, and Ca(OH)2, as shown in Figure 3. Extracted ion chromatogram signals were compiled from ions of nominal masses 57, 58, and 59, corresponding to the most abundant isotopes of calcium hydroxide, nickel, and cobalt species, respectively. Two immediately distinct features of each dispersion plot are apparent. First, one ion track veers to slightly positive Vc at higher Vrf starting at (500 Vrf, 0 Vc). Review of the MS spectra presented in Figure S-2 suggests that the identities of these unfiltered ions in DMS may be attributed to nanosprayed clusters ( ion−molecule complexes) of 11401


Article


necessary dynamic range to 1:1.0) of radioactive cobalt-60 over naturally occurring cobalt-59. However, the natural abundances of cobalt-59, nickel-60, and calcium-43 hydroxide already yield a background ratio of ∼1.0:4.5. Without active cobalt enrichment, we cannot unequivocally attribute the entire m/z 60 signal to radioactive cobalt-60. Implementing the DMS conditions above at 1100 Vrf to reduce the calcium interference by a factor of 7.5 and the nickel interference by a factor of 1.6, ultimately yields a background ratio of 1.0:2.8. This moves toward a threshold control ratio of 1:1, but more importantly imparts greater confidence in attribution to and quantitation of the targeted cobalt-60. Furthermore, hypothetical background calcium-40 hydroxide ion signal for m/z 57 is over 500 times more abundant than that of the background m/z 60 signal. The demonstrated 7.5-fold reduction of calcium signal reduces the

Figure 4. Extracted and normalized barium/cesium mixture transmission plots at 1400 Vrf.

dispersion voltage of 1400 Vrf and shows cesium and barium separating in the mixture with a calculated resolving power of 0.53. The lack of a residual ion track may be partially explained by the use of a lower fragmentor voltage (200 V) which would mask detection of the atomic ions among the indeterminate clusters. The associated mass spectra from select Vc values are shown in Figure 5. The extracted ion signals were derived from m/z 133.89 assigned to the stable cesium isotope and m/z 11402


Article


Figure 5. Mass spectra at (a) 1400 Vrf, −16 Vc, showing selectivity for cesium and (b) 1400 Vrf, −12 Vc, showing selectivity for barium.

species in this system, as shown in Figure 6, with a resolving power of 1.18. In this experiment, unidentified clusters are

68.94 for the stable barium isotope. When compared to unfiltered spectra at (500 Vrf, 0 Vc), the cesium/barium ratio is enhanced at (1400 Vrf, −8.2 Vc), by 4.5 times. Interestingly, the lower fragmentor voltage revealed distinct barium isotope clusters with mixed water and acetonitrile adducts in the mass spectra (Figure 5b) which we identify and tabulate in Table S-1. From these intensity values, we estimate the fragmentor efficiency to release the bare Ba2+ ion at ∼38%. The maximum number of solvent adducts we observe for the Ba2+ ion is 5, and the average number of solvent adducts observed is 2.9 based on signal intensity. In contrast, we observe no solvent adducts in the corresponding cesium peak spectra. Furthermore, hydration spheres of Cs+ and Ba2+ in condensed phases are estimated to contain ∼6−8 and 9.5 waters, respectively.28 However, though more absolute evidence of clustering is observed in the mass spectra and greater numbers of solvent molecules are predicted from the literature, barium is experimentally observed with less dispersive behavior than cesium. Rather, aqueous phase ion radii are 0.29 nm for Ba2+ and 0.31 nm for Cs+,28,33 so that barium’s less dispersive track is associated with a higher mobility, hydrated ion cluster in the low field limit. Thus, DMS behavior for atomic ions is more nuanced than sheer number of ligand interactions and may reflect the binding energy of the ion−ligand interaction, as suggested by others.19,21,23,34 Strontium. For our final target species, strontium, we analyzed a mixture with rubidium to demonstrate separation from a near neighbor. Rb/Sr ratios are commonly useful in geochronology (by the beta decay of 87Rb to 87Sr). However, such measurements require mitigation of their isobaric interferences typically using dynamic reaction cells.35 Using DMS, we observed our most distinct separation of atomic

Figure 6. Transmission plots derived from the mixture of strontium and rubidium at 1350 Vrf.

dominant at the lower compensation voltages as we had observed with our cobalt mixtures. The relative peak order is analogous to the cesium and barium in that the alkali ion (rubidium) is dispersed more than the alkali earth metal ion (strontium). The corresponding mass spectra are shown in Figure 7. With a compensation voltage of −22.5 Vc, the peak mass spectrum 11403


Article


Figure 7. Mass spectra at 1350 Vrf and (a) −22.5 Vc, showing selective passage of rubidium; and (b) −17 Vc, showing selective passage of strontium.

for the rubidium is marked by the higher abundance isotope m/ z 84.90 and residual acetonitrile clusters at m/z 125.92 and 166.94. At the compensation voltage of −17 Vc, the mass spectrum is dominated by up to six acetonitrile adducts for the doubly charged Sr2+ ion (m/z 167.01). Filtration of the Rb+ signal in this spectrum at −17 Vc relative to peak rubidium intensity at −22.5 Vc suggests that the DMS suppresses 90% of the Rb+ signal (10% transmission). Neither peak spectra demonstrate the survival of water adducts at this fragmentor voltage of 325 V, though the acetonitrile adducts are clearly preserved. While the relative dispersion follows that of the cesium system, the separation is clearly better for the strontium target compared to the cesium target and begs more rigorous investigation of DMS using acetonitrile solvation.

■

used to selectively transmit one or another transition metal is encouraging to observe particularly for isobaric interferents. One might then extrapolate DMS as potentially capable of separating out other transition metals such as f-block elements. Such a separation capability on the short DMS time scales is attractive as a forensic screening tool for radiological and nuclear analyses.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b01441. Figure S-1 depicting dispersion plot tracks of the cobalt mixture and Figure S-2 representing mass spectra of the cobalt mixture after DMS separation. Additionally, calculated transmission plot peak resolutions are tabulated in Table S-1; identification of barium cluster components are tabulated in Table S-2 (PDF).

CONCLUSION

We have investigated the phenomenon of gas phase ion separation by differential mobility spectrometry for three target systems of interest: cobalt, cesium, and strontium. In each case, we demonstrated enrichment of the target species in a mixture with a near neighbor element. The cobalt system is interesting for the fact that the nickel demonstrates two DMS-stable tracks. In the cesium and strontium systems, evidence of the metal ion solvation as the primary chemistry responsible for the separation by DMS reflects the clustering/declustering model. The data suggests that the simple embodiment of DMS prefiltration into a portable mass spectrometer should improve identification and may even enhance quantitation of metal cations. Ultimately, the proof-of-concept that DMS may be

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 813-974-9633. Notes

The authors declare no competing financial interest. 11404


Article


■

(30) Ohno, T.; Muramatsu, Y. J. Anal. At. Spectrom. 2014, 29, 347− 351. (31) Liezers, M.; Farmer, O. T.; Thomas, M. L. J. Radioanal. Nucl. Chem. 2009, 282, 309−313. (32) Ames, F.; Audi, G.; Beck, D.; Bollen, G.; de Saint Simon, M.; Jertz, R.; Kluge, H.-J.; Kohl, A.; König, M.; Lunney, D.; Martel, I.; Moore, R. B.; Otto, T.; Patyk, Z.; Raimbault-Hartmann, H.; Rouleau, G.; Savard, G.; Schark, E.; Schwarz, S.; Schweikhard, L.; Stolzenberg, H.; Szerypo, J. Nucl. Phys. A 1999, 651, 3−30. (33) Marcus, Y. Chem. Rev. 1988, 88, 1475−1498. (34) Glendening, E. D.; Feller, D. J. Phys. Chem. 1996, 100, 4790− 4797. (35) Moens, L. J.; Vanhaecke, F. F.; Bandura, D. R.; Baranov, V. I.; Tanner, S. D. J. Anal. At. Spectrom. 2001, 16, 991−994.

ACKNOWLEDGMENTS This work was supported by Contract Number HDTRA-11-010012 by the Defense Threat Reduction Agency Basic Research Program in Nuclear Forensics.

■

REFERENCES

(1) Barshick, C.; Duckworth, D.; Smith, D. Inorganic Mass Spectrometry: Fundamentals and Applications; CRC Press: New York, 2000. (2) D’Ilio, S.; Violante, N.; Majorani, C.; Petrucci, F. Anal. Chim. Acta 2011, 698, 6−13. (3) Hart, G. L.; Hager, G. J.; Barinaga, C. J.; Duckworth, D. C. Market Research Survey of Commercial Off-the-shelf (COTS) Portable MS System for IAEA Safeguards Applications; Pacific Northwest National Laboratory, 2013. (4) Bowers, M. T.; Kemper, P. R.; von Helden, G.; Koppen, P. A. M. van. Science 1993, 260, 1446−1451. (5) Kemper, P. R.; Bowers, M. T. J. Phys. Chem. 1991, 95, 5134− 5146. (6) Dion, H. M.; Ackerman, L. K.; Hill, H. H. Talanta 2002, 57, 1161−1171. (7) Blagojevic, V.; Chramow, A.; Schneider, B. B.; Covey, T. R.; Bohme, D. K. Anal. Chem. 2011, 83, 3470−3476. (8) Kolakowski, B. M.; Mester, Z. Analyst 2007, 132, 842−864. (9) Coy, S. L.; Krylov, E. V.; Schneider, B. B.; Covey, T. R.; Brenner, D. J.; Tyburski, J. B.; Patterson, A. D.; Krausz, K. W.; Fornace, A. J.; Nazarov, E. G. Int. J. Mass Spectrom. 2010, 291, 108−117. (10) Schneider, B. B.; Nazarov, E. G.; Londry, F.; Vouros, P.; Covey, T. R. Mass Spectrom. Rev. 2016, 35, 687−737. (11) Barnett, D. A.; Guevremont, R.; Purves, R. W. Appl. Spectrosc. 1999, 53, 1367−1374. (12) Ells, B.; Barnett, D. A.; Froese, K.; Purves, R. W.; Hrudey, S.; Guevremont, R. Anal. Chem. 1999, 71, 4747−4752. (13) Sinatra, F. L.; Wu, T.; Manolakos, S.; Wang, J.; Evans-Nguyen, T. G. Anal. Chem. 2015, 87, 1685−1693. (14) Blades, A. T.; Jayaweera, P.; Ikonomou, M. G.; Kebarle, P. J. Chem. Phys. 1990, 92, 5900−5906. (15) Blades, A. T.; Jayaweera, P.; Ikonomou, M. G.; Kebarle, P. Int. J. Mass Spectrom. Ion Processes 1990, 102, 251−267. (16) Peschke, M.; Blades, A. T.; Kebarle, P. J. Phys. Chem. A 1998, 102, 9978−9985. (17) Peschke, M.; Blades, A. T.; Kebarle, P. Int. J. Mass Spectrom. 1999, 185−187, 685−699. (18) Shvartsburg, A. A.; Siu, K. W. M. J. Am. Chem. Soc. 2001, 123, 10071−10075. (19) Beyer, M.; Williams, E. R.; Bondybey, V. E. J. Am. Chem. Soc. 1999, 121, 1565−1573. (20) Krylova, N.; Krylov, E.; Eiceman, G. A.; Stone, J. A. J. Phys. Chem. A 2003, 107, 3648−3654. (21) Liu, C.; Le Blanc, J. C. Y.; Shields, J.; Janiszewski, J. S.; Ieritano, C.; Ye, G. F.; Hawes, G. F.; Hopkins, W. S.; Campbell, J. L. Analyst 2015, 140, 6897−6903. (22) Campbell, J. L.; Le Blanc, J. C. Y.; Schneider, B. B. Anal. Chem. 2012, 84, 7857−7864. (23) Campbell, J. L.; Zhu, M.; Hopkins, W. S. J. Am. Soc. Mass Spectrom. 2014, 25, 1583−1591. (24) Kafle, A.; Coy, S. L.; Wong, B. M.; Fornace, A. J.; Glick, J. J.; Vouros, P. J. Am. Soc. Mass Spectrom. 2014, 25, 1098−1113. (25) Schneider, B. B.; Covey, T. R.; Coy, S. L.; Krylov, E. V.; Nazarov, E. G. Eur. Mass Spectrom. 2010, 16, 57−71. (26) Schneider, B. B.; Covey, T. R.; Coy, S. L.; Krylov, E. V.; Nazarov, E. G. Anal. Chem. 2010, 82, 1867−1880. (27) Hall, A. B.; Coy, S. L.; Kafle, A.; Glick, J.; Nazarov, E.; Vouros, P. J. Am. Soc. Mass Spectrom. 2013, 24, 1428−1436. (28) Ohtaki, H.; Radnai, T. Chem. Rev. 1993, 93, 1157−1204. (29) Shvartsburg, A. A.; Wilkes, J. G.; Lay, J. O.; Michael Siu, K. W. Chem. Phys. Lett. 2001, 350, 216−224. 11405
