Machine-learning assisted suspect screening in passive sampler extracts and rapid direct injection LC-MS/MS of river water from the River Thames, London, UK in 2018/9.

The application of new sorbent-film coated passive samplers for capture of bulk commercial and military explosives vapours in operationally relevant spaces such as luggage, rooms, vehicles and shipping containers is presented. Samplers were easily integrated with in-service detection technologies with little/no sample preparation required. Ethylene glycol dinitrate (EGDN) was detected within 4 h in a container holding a suitcase packed with 0.2 kg Perunit 28E. Within a 22,000 dm3 room, 1 kg of concealed Seguridad was detected within 24 h and in an adjoining room within 7 days. Exposed samplers also successfully captured components of 1 kg TNT after 72 h and 1 kg concealed Perunit 28E after 6 h in both a furnished room and a large, partially filled shipping container. For the latter, samplers captured detectable residues outside the container after 24 h and were stable during wet weather for 72 h. A one-week trial at three operationally relevant venues including a university, a theatre and a government building revealed a nuisance positive rate of <1.4% (n = 72). Finally, two alternative applications are presented for extraction of liquid samples and use a particulate contact swab showing flexibility for a range of different search activities.

No abstract