1
Detecting explosive vapours in large volume freight: 2016 improvements G. Fernández de la Mora, 1 D. Zamora, 1 M. Amo 1 , J. Fernández de la Mora 2 1 SEADM, Boecillo, Spain; 2 Yale University, Mech. Eng. Dept., New Haven, CT 06520-8286, USA Workshop on Spectrometry for security applications, Birmingham December 15th, 2016 MOBILITY RESOLUTION CONCLUSIONS The following relevant results have been obtained during 2016 in the ACES explosives vapour detector: (i) The DMA resolution has increased significantly. (ii) The sensitivity gain has increased markedly, in particular in the TNT and EGDN channels. (iii) The PoD for RDX is close to 100% for 20 fg captured on the filter. (iv) LDL for some explosives is limited by interferents. ACKNOWLEDGEMENTS We are grateful to the EU H2020 SME Phase 2 Program “ACES”, grant number 672001. +1 +4 +1 +3 +4 -2 -3 -4 -5 -6 -7 -8 -9 +1 +2 +3 +4 +5 OVERVIEW Figure 3: Mobility plots of the tetraheptylammonion+ ion standard. X-Axis presents electrical mobility (Volts), Y-Axis presents signal intensity, also in Volts. The red plot presents a mobility peak with the previous DMA; the mobility peak in blue presents the mobility plot when the DMA has been fitted with one prelaminarizer, and the orange plot presents the mobility peak when the DMA has been fitted with two prelaminarizers Vapors desorbed from filter ↓ Ionization (SESI) ↓ Analyzer (Ion mobility Front-End + triple quad MS) SAMPLER Filter loaded with vapours Cargo ↓ Sample gas ↓ Vapors trapped in filter Figure 1: SEADM’s cargo screening concept Desorber Ionizer Triple quadrupole MS Sciex Triple Quad 5500 2.6 Lit/min sample flow Figure 2: Schematic of the analyzer INSTRUMENT SEDET’s vapor explosive detector ACES 2.1 (Figures 1-2) based on a triple quadrupole mass spectrometer preceded by a differential mobility analyzer (DMA) has demonstrated the capability to separate gas phase species present in the atmosphere at concentrations below 0.01 ppq (10 -17 atmospheres) from explosive vapours present in large volume freight. This lower detection limit (LDL) is partially determined by the finite sensitivity of the analyzer, and partially by competition from other interfering vapours (with the same mobility, precursor and product ion masses as the target explosive) also present in ambient air at ~0.01 ppq. The present poster describes in general terms the explosive screener architecture, and presents in detail the main results achieved in 2016. ACES operation can be described as follows: between 0.6 and 1 m 3 of air from the container or truck to be analyzed are collected on a sampling filter, and brought to the Front-End. The Front-End receives the filter in a variable temperature thermal desorber and leads the vapours during a desorbtion cycle of 5 minutes into SEADM´s D-LFSESI (Desolvating Low Flow SESI ionizer based upon secondary ionization through Cl - ), which injects the ions into the ion mobility filter (SEADM’s P5 DMA). At the Front- End exit, a triple quadrupole mass spectrometer(Sciex Triple Quad API 5500) completes the analysis. Figure 4: Increase in resolving power with DMA voltage. The resolving power now increases continuously with DMA voltage, while before it reached a plateau around 78 SENSITIVITY INCREASE DETECTION PROBABILITY Table 2: Results for 314 loaded tests for RDX. The Probability of Detection (PoD) is shown as a function of the femtograms (fg) captured on the filter. The detection probability for signals between 10 and 20 fg is 30%, and reaches nearly 100% for signals above 20 fg. In order to detect 0.01 ppq of RDX with a PoD above 90%, 20 fg are needed in the filter, equivalent to a sampling volume of 1,075 liters Experiments carried out in SEADM´s DMA workbench have demonstrated that SEADM´s DMA can reach resolving powers above 100 for m/z 410 Da. Design improvements have been the following: • Prelaminarization stage: Eliminates turbulences from the DMA recirculation circuit. • Laminarizer stage: New laminarizers with improved aerodynamic characteristics. • Two DMA blower modules enable higher Reynolds numbers in the DMA separation channel. In the future we expect to obtain the same Reynolds number improvement through a more efficient diffuser. 40 50 60 70 80 90 100 110 120 1000 2000 3000 4000 5000 6000 DMA Resolving Power DMA Voltage (V) DMA Resolving Power for THA + ion 2 DMA Blowers working in series Sensitivity (counts/fg) EGDN-62 RDX-46 PETN-46 NG-46 TNT-46 November 2016 1.3 83 19 31 324 December 2015 0.4 54 18 17 73 Improvement Factor 3.6 1.5 1.1 1.9 4.4 Table 1: ACES 2.1 sensitivity has increased significantly in 2016, with improvement factors above 4 for TNT. Sensitivity is defined as the number of ion counts measured by the mass spectrometer when 1 femtogram is deposited on the filter. Present sensitivity has reached 324 counts per femtogram for TNT, delivering unprecedented detection capability in large volume freight The sensitivity gain has been achieved through chemical means: a solution of acids in minute quantities (pg/s) is continuously injected in the desorber. The best results have been obtained with nitric acid, while other compounds also show relevant effects. The reasons for the sensitivity gain are still unclear. Our main hypothesis is that the acids eliminate pollution traces present in the desorber, which when present, react with explosives and retain them before they could reach the Ionizer. The amount of acids injected is very small when compared to the HCl injected in the SESI, and does not interfere with ionization. -Q1: 0,000 to 2,834 min from Sample 1 (20161123 p33 Q1. Secundario 200ppb HNO3 MeOH-Agua 9-1 2000mbar) of 20161123 p33 Q1. Secunda... Max. 1,7e6 cps. 60 80 100 120 140 160 180 200 220 240 260 280 300 320 m/z, Da 0,0 1,0e5 2,0e5 3,0e5 4,0e5 5,0e5 6,0e5 7,0e5 8,0e5 9,0e5 1,0e6 1,1e6 1,2e6 1,3e6 Intensity, cps 62,1 71,0 73,1 197,8 199,9 95,1 75,0 151,1 81,1 171,0 98,1 131,1 125,2 161,0 248,1 179,3 113,1 263,1 135,1 201,8 233,1 207,1 83,1 64,2 46,0 312,1 321,1 279,2 229,1 289,1 69,1 295,2 158,1 55,0 Figure 5: Q1 plot of the background noise with and without injection of nitric acid in the desorber at a rate of 15 pg/s. The X- Axis shows ion masses between 50 and 330 m/z, while the Y-Axis presents ion counts/s between 0 and 1.3 million. The red line is the noise prior to nitric acid injection, while the blue line shows internal noise during injection. As expected, mass 62 increases significantly with nitric acid injection, while the relevant result is that masses between 160 and 180, and between 190 and 205, decrease considerably with nitric acid injection, delivering a much cleaner background. Signal level (fg) Number of tests Detections PoD 0-10 41 0 0.0% 10-20 67 20 29.9% 20-30 52 49 94.2% 30-40 30 30 100.0% >40 124 124 100.0% Total 314 223 71.0% Figure 6: NG262/62 versus NG262/46 for numerous nitroglycerin loaded tests. The inset is a zoom of the low concentration region, showing two contaminating species with intensities up to 17 ppq 5700 5500 5300 5100 4900 4700 4500 4300 4100 3900 3700 300 1300 2300 3300 4300 m/z 27 28 29 30 31 +4 +4 +4 +4 +4 39 40 41 42 43 44 +5 +5 +5 +5 +5 +5 54 55 56 57 +6 +6 +6 +6 z = 5 z = 6 z = 7 z = 4 (b)
