ExploringApplicabilityofDirectAnalysisinRealTime

withMassSpectrometry(DART-MS)toIdentifyHomemadeExplosiveResiduesPost-Blast

by

ChelseaElizabethBlack

AthesissubmittedtotheFacultyofGraduateandPostdoctoralAffairsinpartialfulfillmentoftherequirementsforthedegreeof

MasterofScienceIn

Chemistry

CarletonUniversityOttawa,Ontario

©2019ChelseaElizabethBlack

1

AbstractApplication of Direct-Analysis-in-Real-Time (DART) ionization with mass spectrometry

(DART-MS) to identify explosives from post-blast residues is presented. Explosives of

interest represent real current threatsencountered in criminal investigations inNorth

America and Europe: homemade organic peroxides, binary explosives and smokeless

powder.Aseriesof simulated improvisedexplosivedevices (IEDs)weremanufactured

usingtriacetonetriperoxide(TATP),hexamethylenetriperoxidediamine(HMTD),methyl

ethylketoneperoxide(MEKP),homemadebinaryexplosives(composedofafuel-oxidizer)

andsingleanddouble-basesmokelesspowders.EachIEDwasconfiguredtoyieldbomb

fragmentsrepresentativeofactualmaterialsrecoveredfrombombinginvestigations.The

goal of this study was to demonstrate the validity of DART-MS for identification of

homemade explosives using real world samples (i.e. not laboratory simulations) and

develop a quality assured method for use in accredited forensic laboratory settings.

Smokelesspowderwasofspecificinterestasthereiscurrentlynoreportedmethodto

identify nitrocellulose (NC) post-blast, unless unconsumed material is recovered.

Therefore, this study aimed to demonstrate the validity of DART-MS to characterize

thermalbreakdownproductsofNC.Allrecoveredfragmentswereanalyzeddirectlyand

in directly (i.e. cotton swabs and solvent extraction methods) using full scan high

resolutionmassspectrometry (HRMS).Thisworkdemonstrates the forensicvalidityof

DART-MStoproviderapidandqualityassuredidentificationofexplosiveresiduesfrom

realpost-blastIEDfragments.

2

Acknowledgements

TomysupervisorsDr.NigelHearnsandDr.JeffreySmith,Iwouldliketothankyou

bothfortheopportunitytoworkonthisprojectwithyouboth.Yourvision,supportand

guidancethroughoutourworkingyearstogetherwasparamountinthesuccessofthis

project.Jeff,Itrulyappreciatetheamazingopportunitiesyouhavegivenme.Forallthe

advice,supportandencouragement,Iamtrulythankful.Havingcometoyourofficein

bothmomentsofexcitementandstress,youalwaysknewexactlywhatIneededtohear

tofurthermotivatemeorhelpmegetthroughanystrugglesIwasexperiencing.Nigel,I

amforevergratefulfortheexperienceIhavehadworkingonthisprojectwithyou.The

developmentofmyknowledgeandskillsareadirectresultofthepatienceandtimeyou

havetakenteachingwhenmeaboutforensicscience,explosives,experimentaldesign,

professionalismandthelistgoeson.Thankyouforadvocatingforthisresearchwithinthe

labandtheorganization.Yourcontinuous,never-endingsupporthasmadethisthemost

rewardingandmemorableworkingexperienceofmylife.Thank-you!

To my friends and family – I would like to thank you all for your continuous

encouragementandsupportduringthecompletionofmyMasters.Dad,withtheturning

of each school year, and yes I know there have beenmany, your support has never

wavered.Asaparentchaperoneonpublicschoolfieldtrips,afaninthestandsofmany

sportsgamesandcomingtoOttawatositinasaspecialguestatmydefence-youhave

never missed any special moment in my educational career, big or small. Your

understandingandappreciationformyloveofschoolhasallowedmetopursueanyand

everydreamIhaveeverhad.Yourkindwords,immenseloveandthoughtfuladvicehas

3

helpedmereacheachdestinationinthepaththathasledmetotheendofgradschool.

Youaretrulymygreatestconfidant.Mom,havingsharedyourinnatelovefor learning

withme,Ihaveyoutothankformydriveanddeterminationtoobtainsuchafulsome

education.RyanandCourtney,youhavebeenthegreatestrole-modelsalittlesistercould

ask for. Your hard-work, determination, and commitment to your professions while

maintainingimportantrelationshipsandfriendshipsissomethingItrulyadmireinbothof

you. Thank you so much for always cheering me on, providing sound advice and

encouragingmetonevergiveup!TotheMacDonaldfamily,Icanneverthankyouenough

for includingme in your extremely loving and supportive family. Finally, tomy girls –

Christine,Stephanie,Annah,JillianandKylie.Ourfriendshipmeanstheworldtome.Iam

so thankful for all the memories made and your never-ending support and

encouragement.

BeingamemberoftheSmithLabhasbeensuchapleasure.Participatingingroup

meetings,celebratingtheholidaysatourepicChristmaspartiesandsharingaglassof

beer(ciderformeofcourseJ)attheendoftheweekwillforeverbesomeofmyfondest

memories of my time at Carleton University. To all Smith labmembers, current and

alumni,Iwanttothankyoufortheknowledge,experienceandadviceyouhaveshared

withmeduringmytimeinthelab.Karl,ourabilitytonaturallyseguefromexperiment

troubleshootingtoourlatestFridaynightadventureswithourfriendsandfamilieswill

neverbelostuponme.OurfriendshipisextremelyimportanttomeandIlookforwardto

continuousupdatesandcatchingupwhenyoucometovisitinKingston!

4

ToallthemembersoftheDepartmentofChemistryatCarletonUniversity,Ithank

you for your support, knowledge, ideas and friendships and the many learning and

teachingopportunitiesprovidedtomebyTApositionsandvolunteeropportunities.

To all the members of the Trace Evidence section of the Forensic Laboratory

Services and the RCMP at large; I thank you for your support, access to resources,

knowledgeandpatienceasInavigatedmywaythroughthisresearchstudy. Youhave

beensowelcoming,encouragingandsupportive.

5

TableofContentsAbstract...................................................................................................................................................................1

Acknowledgements.................................................................................................................................................2

ListofTables............................................................................................................................................................6

ListofFigures..........................................................................................................................................................7

ListofSchemes......................................................................................................................................................12

ListofAbbreviations..............................................................................................................................................13

Foreword...............................................................................................................................................................15

1.Introduction......................................................................................................................................................18

1.1.Explosives...................................................................................................................................................18

1.1.1.Homemadeexplosives.......................................................................................................................20

1.2.ExplosiveAnalysis......................................................................................................................................29

1.2.1.IonMobilitySpectrometry(IMS)........................................................................................................29

1.2.2.ChromatographyMethods.................................................................................................................30

1.2.3.MassSpectrometry(MS)....................................................................................................................35

2.MaterialsandMethods.....................................................................................................................................45

2.1.Consumables,ReagentsandStandardReferenceMaterials.....................................................................46

2.2.IEDConstruction,DetonationandFragmentCollection............................................................................47

2.3.ReferenceMaterialSamplePreparation...................................................................................................49

2.3.1.FuelandOxidizer................................................................................................................................49

2.3.2.OctanitrateCellobioseSynthesisandSamplePreparation................................................................49

2.3.3.SmokelessPowder.............................................................................................................................50

2.4.Post-BlastExtractPreparation...................................................................................................................50

2.5.DART-MSAnalysis......................................................................................................................................51

3.ResultsandDiscussion......................................................................................................................................52

3.1DART-MSParameterOptimization.............................................................................................................52

3.2.AnalysisofFragments................................................................................................................................53

3.2.1.OrganicPeroxideExplosives...............................................................................................................61

3.2.2.BinaryExplosives................................................................................................................................74

3.2.3.SmokelessPowders............................................................................................................................90

4.Conclusion.........................................................................................................................................................96

5.FutureWork......................................................................................................................................................97

References.............................................................................................................................................................98

Appendix1:SupplementaryInformation............................................................................................................102

6

ListofTablesTable1.Examplesofproductsthatsourcecomponentsforbinaryexplosives............................24

Table2.TypeandamountofexplosiveusedasmainchargeforeachIED...................................48

Table3.LODsobservedforexplosivesofinterestmeasuredusingin-housemethodandQuickStripcomparedtoliteraturevalues.............................................................................52

Table4.DepictionoftheOPBEidentifiedviaDART-MSdirectanalysisofamultitudeofdifferentpost-blastfragmentscomparedtoin-direct.........................................................................62

Table5.CharacteristicionsofHMTDpresentuponanalysisofresiduescollectedviaswabsdifferentiatedbysubstrate...................................................................................................71

Table6.Listofmassformulaefortheionscharacteristicofnitratedsugarthermalbreakdownproducts,withassociatedmassshift(amu)..........................................................................93

7

ListofFiguresFigure1.Triacetonetriperoxide(TATP)........................................................................................21

Figure2.Hexamethylenetriperoxide(HMTD)..............................................................................22

Figure3.Methylethylketone(MEKP)..........................................................................................22

Figure4.Fully-nitratednitrocellulose...........................................................................................26

Figure5.Nitroglycerin...................................................................................................................27

Figure6.Diphenylamine...............................................................................................................27

Figure7.Ethylcentralite...............................................................................................................27

Figure8.Schematicdiagramofanionmobilityspectrometer(IMS)............................................29

Figure9.Schematicdiagramoftheinstrumentationusedtoseparateanalytesofamixtureviahigh-performanceliquidchromatography(HPLC)................................................................31

Figure10.Schematicdiagramoftheinstrumentationusedtoseparatevolatilecomponentsinamixtureviagaschromatography(GC)...................................................................................32

Figure11.Schematicdiagramoftheinstrumentationusedforsimultaneousseparationofanionsandcationsinsolutionviaionchromatography(IC).............................................................33

Figure12.SchematicdiagramofDARTsource..............................................................................39

Figure13.ImagesofthedifferentorientationsoftheDARTsourcewithrespecttotheMSinterface:(a)surfacedesorptionmodeand(b)transmissionmode.....................................42

Figure14.SchematicdiagramofthedesignandengineeringoftheQ-Exactivehybridmassspectrometer.........................................................................................................................44

Figure15.Fragmentscollectedpost-blastfromthedetonationofdevicesutilizinghomemadeexplosivesasthemaincharge(IED#1-14)............................................................................47

Figure16.IonsobservedviaoperationoftheQExactiveinpositivefullscanmode,withouttheDARTsourceturnedon.Totalioncount103.........................................................................54

Figure17.EndogenousDART-MSions.Totalioncount104-105..................................................54

Figure18.Analysisofanunusedcottonswabinpositivemodeusingfullscan.Totalioncount104.........................................................................................................................................55

Figure19.AnunusedQuickStripanalyzedinpositivemodeusingfullscan.Totalioncount103-105.........................................................................................................................................55

8

Figure20.WaterdepositedontoaQuickStrip,analyzedinpositivemodeusingfullscan.Totalioncount104–105................................................................................................................56

Figure21.MethanoldepositedontoaQuickStrip,analyzedinpositivemodeusingfullscan.Totalioncount104–106................................................................................................................56

Figure22.AcetonedepositedontoaQuickStrip,analyzedinpositivemodeusingfullscan.Totalioncount104–106................................................................................................................57

Figure23.AcetonitriledepositedontoaQuickStrip,analzyedinpositivemodeusingfullscan.Totalioncount104–106.......................................................................................................57

Figure24.AnalysisofdichloromethanedepositedontoaQuickStrip,inpositivemodeusingfullscan,tobeusedforsolventassociatedionsubtraction.Totalioncount104–106..............58

Figure25.AnalysisofhexanedepositedontoaQuickStrip,inpositivemodeusingfullscan.Totalioncount104–105................................................................................................................58

Figure26.Positivemode,fullscanhigh-resolutionmassspectrumforTATPanalyzedfromcertifiedreferencestandard.IonscharacteristicofTATPhavebeenboldedandlabelled...63

Figure27.Positivemode,fullscanhigh-resolutionmassspectrumforTATPupondirectanalysisoffragmentfromIED#2.IonscharacteristicofTATPhavebeenboldedandlabelled.........64

Figure28.Positivemode,fullscanhigh-resolutionmassspectrumforTATPanalyzedfromaswabusedtocollectpost-blastresiduesfromIED#2fragments.IonscharacteristicofTATPhavebeenboldedandlabelled.............................................................................................64

Figure29.Fullscanhigh-resolutionmassspectrumforHMTDanalyzedfromcertifiedreference.IonscharacteristicofHMTDhavebeenboldedandlabelled................................................65

Figure30.Fullscanhigh-resolutionmassspectrumdepictingidentificationofHMTDfromdirectanalysisofafragmentcollectedpost-blastfromIED#3.IonscharacteristicofHMTDhavebeenboldedandlabelled......................................................................................................66

Figure31.Fullscanhigh-resolutionmassspectrumdepictingidentificationofHMTDuponanalysisofaswabusedtocollectpost-blastresiduesfromIED#3fragments.IonscharacteristicofHMTDhavebeenboldedandlabelled.......................................................66

Figure32.Positivemode,fullscanhigh-resolutionmassspectrumforMEKPanalyzedfromthecrudesynthesizedproduct.IonscharacteristicofMEKPhavebeenboldedandlabelled...68

Figure33.Positivemode,fullscanhigh-resolutionmassspectrumdepictingidentificationofMEKPfromdirectanalysisoffragmentcollectedpost-blastfromIED#5.IonscharacteristicofMEKPhavebeenboldedandlabelled...............................................................................69

9

Figure34.Positivemode,fullscanhigh-resolutionmassspectrumdepictingidentificationofMEKPuponanalysisofaswabusedtocollectedpost-blastresiduesfromIED#5fragments.IonscharacteristicofMEKPhavebeenboldedandlabelled................................................69

Figure35.Fromlefttoright-cottonswab,paperswab,modifiedpaperswab...........................71

Figure36.Fullscanhigh-resolutionmassspectradepictingidentificationofHMTDuponcollectionofpost-blastresiduesfromIED#5usingdryandsolventdampenedswabs.......73

Figure37.Fullscanhigh-resolutionmassspectrumofglucosedissolvedinwaterasareferencematerial.Collectedinpositivemode.....................................................................................76

Figure38.Fullscanhigh-resolutionmassspectrumofsucrosedissolvedinwater,usedasareferencematerial.Collectedinpositivemode....................................................................76

Figure39.Fullscanhigh-resolutionmassspectrumofTANGdissolvedinwater,usedasareferencematerial.Collectedinpositivemode....................................................................77

Figure40.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11metalsubstratefragment.Collectedinpositivemode...................................................................77

Figure41.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11plasticsubstratefragment.Collectedinpositivemode...................................................................78

Figure42.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11rubbersubstratefragment.Collectedinpositivemode...................................................................78

Figure43.Fullscanhigh-resolutionmassspectrumofdextrinreferencematerialdissolvedinwater.Collectedinpositivemode.........................................................................................79

Figure44.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12plasticsubstratefragment.Collectedinpositivemode...................................................................79

Figure45.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12metalsubstratefragment.Collectedinpositivemode...................................................................80

Figure46.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12rubbersubstratefragment.Collectedinpositivemode...................................................................80

10

Figure47Fullscanhigh-resolutionmassspectrumofanautomotivegrease,usedasareferencematerial.Collectedinpositivemode.....................................................................................83

Figure48.Fullscanhigh-resolutionmassspectrumdepictingidentificationofautomotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10metalsubstratefragment.Collectedinpositivemode....................................................................................................83

Figure49.Fullscanhigh-resolutionmassspectrumdepictingidentificationofautomotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10plasticsubstratefragment.Collectedinpositivemode....................................................................................................84

Figure50.Fullscanhigh-resolutionmassspectrumdepictingidentificationofautomotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10rubbersubstratefragment.Collectedinpositivemode....................................................................................................84

Figure51.Negativemode,fullscanhighresolutionmassspectrumuponoperationoftheDART-MS,depictingtheendogenousions.Totalioncount106......................................................86

Figure52.AnalysisofwaterdepositedontoaQuickStrip,innegativemodeusingfullscan.Totalioncount106.........................................................................................................................87

Figure53.Negativemode,fullscanhigh-resolutionmassspectrumforammoniumnitrateanalyzedasareferencematerial..........................................................................................87

Figure54.Negativemode,fullscanhigh-resolutionmassspectrumforacommerciallyavailablestumpremover(commercialsourceofKNO3),analyzedasareferencematerial................88

Figure55.IdentificationofKNO3innegativemodeviafullscanhighresolutionDART-MSanalysisofpost-blastresiduesextractedfrommetalsubstratefragmentsfromIED#10withwater................................................................................................................................................88

Figure56.IdentificationofKNO3innegativemodeviahigh-resolutionDART-MSanalysisofpost-blastresiduesextractedfromplasticsubstratefragmentsfromIED#10withwater...........89

Figure57.IdentificationofKNO3innegativemodeviafullscanhigh-resolutionDART-MSanalysisofpost-blastresiduesextractedfromrubbersubstratefragmentsfromIED#10withwater................................................................................................................................................89

Figure58.Determinationofionscharacteristicofthethermalbreakdownproducts.................92

Figure59.Relativesolubilityofresiduescontainingthermalbreakdownproducts.....................94

Figure60.Positivemode,fullscan:high-resolutionmassspectrafordiesel,analyzedasareferencematerial...............................................................................................................102

Figure61.Positivemode,fullscan:high-resolutionmassspectraforlampoil,analyzedasareferencematerial...............................................................................................................102

11

Figure62.Positivemode,fullscan:high-resolutionmassspectraforVaseline,analyzedasareferencematerial...............................................................................................................103

Figure63.Positivemode,fullscan:high-resolutionmassspectraforwax,analyzedasareferencematerial...............................................................................................................103

12

ListofSchemesScheme1.Reactionschemefortheelectronicorvibronicproductionofmetastablespecies(M*)

frominertgas(M)occurringinthesourceviaaseriesofelectrodes...................................38

Scheme2.Reactionschemestoproducesecondaryionizingspecies(ionizedwaterclusters)inpositivemodeviareactionofmetastablespeciesproducedbytheDARTsourcewithatmosphericreagents............................................................................................................40

Scheme3.Reactionschemestoproduceionizedanalytespecies(S+•)inpositivemodeviareactionofsecondaryionizedspeciesandanalytemoleculespresentedtothesourceregion....................................................................................................................................40

Scheme4.Reactionschemesforproductionofnegativeionizedanalytespecies(S-)via...........41

13

ListofAbbreviations ACN=Acetonitrile

AN=AmmoniumNitrate

Ar=Argon

DART=DirectAnalysisinRealTime

DART-MS=DARTcoupledMassSpectrometry

DPA=Diphenylamine

EC=EthylCentralite

EIC=ExtractedIonChromatogram

FTIR=Fourier-TransformInfraredSpectroscopy

FLS=ForensicLaboratoryServices

FWHM=Full-WidthHalf-Maximum

GC=Gaschromatography

He=Helium

He*=Excited-stateHelium

HRMS=High-resolutionmassspectrometry

HME=HomemadeExplosives

HMTD=HexamethyleneTriperoxide

HPLC=High-PerformanceLiquidChromatography

IC=Ionchromatography

IED=ImprovisedExplosiveDevices

14

IMS=IonMobilitySpectrometry

LOD=LimitofDetection

LRMS=Low-resolutionmassspectrometry

MEKP=MethylEthylKetonePeroxide

MeOH=Methanol

MS=MassSpectrometry

N2=Dinitrogengas

NC=Nitrocellulose

NG=Nitroglycerin

ONCB=OctanitrateCellobiose

OPBE=Organicperoxide-basedexplosives

PETN=PentaerythritolTetranitrate

RCMP=RoyalCanadianMountedPolice

TATP=TriacetoneTriperoxide

Th=Thomsons

TIC=TotalIonChromatogram

TLC=ThinLayerChromatography

TPOF=TechnicalandProtectiveOperationsFacility

15

Foreword

Forensic science is the application of science to law and is a subject of great

fascinationtothepublicatlargeasoftenportrayedinmainstreamentertainmentmedia

andHollywoodcinematographicmovies.1Entertainmentmediaforpublicconsumption

oftenportraysadistortedorembellishedviewofactivitiesinvolvedwithforensicscience.

The truth about advantages and limitations, policies and regulations are often

misrepresented.InCanada,forensicscienceservicesaredeliveredbyscientistswhowork

atalllevelsofgovernment,includingfederal,provincialandmunicipal.Medicalcoroners

areeithermunicipallyorprovinciallyregulated.Forensicanalysisservicesrequestedby

municipal and provincial police agencies are funded by the provincial government in

OntarioandQuébecandareperformedbyscientistsattheCentreofForensicSciencein

TorontoforOntarioortheLaboratoiredesciencesjudiciairesetdemédecinelégaledu

QuébecinQuébec,respectively.1TheRoyalCanadianMountedPolice(RCMP)Forensic

LaboratoryServices(FLS)deliversforensicanalysisservicestoallpoliceagenciesoutside

ofQuébecandOntario,andforallfederalpolicingactivities inallprovincesacrossthe

country.1

Manystreamsofscience(e.g.biology,psychology,toxicologyandchemistry)are

utilized considerably to assist in criminal investigations. 2 Locard’s exchange principle

states thatwhen twoobjects come intocontact there isalwaysa transferofmaterial

between them, even if only at the microscopic level. The consequence of Locard’s

exchange principle in forensic science is that any action made by an individual in

commissionofacrimeleadstotheproductionofevidence.Thepowerofsuchevidence

16

lies in the ability to detect, identify and understand the information provided for

capturing the individual or re-creating the circumstances of the crime committed. 2

Chemicalanalysesprovidespolicewithanswers to investigationalquestions toensure

valid identification and source attribution of evidence found at the scene of a crime.

Explosive materials represent a serious hazard as they can be illicitly used for mass

destructionandinjuryordeath.Detectionandidentificationofexplosivesthusremain

critical to ensure public safety, infrastructure security, and bolster counter-terrorism

readiness. After an explosion, recovering and analyzing bomb fragments can provide

importantforensiclinksfortheensuinginvestigation,especiallyincaseswherethereare

no biological traces (i.e. DNA) found at the scene. The substrate, size, and degree of

burningofthefragmentscollectedpost-blastcanprovidenecessaryinformationforthe

re-constructionof theexplosivedevice.Post-blast residueswillyieldunconsumedand

combustionproducts3fromtheoriginalenergeticmaterial,whichprovidesthenecessary

informationtoidentifythetypeandthesourceoftheexplosivefillerusedinthedevice.

Variousfieldandlaboratorytechniquesareavailablefordetectionofexplosives,

bothpre-andpost-blast.4Ionmobilityspectrometry(IMS)remainsapopulartechnique

forrapidfielddetectionofawidevarietyofexplosivesandisextensivelyusedforpre-

screeningpeopleandobjectsat securitycheckpointsas theyareeasilyprogrammable

making them user friendly for front line staff. 5 Matrix interference arising from

environmentalcontaminantsorothercongenerscanaffectthediscriminatingpowerof

IMSandtestresultsarelargelyusedforpresumptivepurposesonly.6,7Furthermore,the

complexity of post-blast samples necessitates a multiplexed analytical scheme to

17

uniquelycharacterizethevariousorganicandinorganiccomponentsofexplosiveresidues

apart from matrix interference. 8 Therefore, much attention is devoted to

chromatography and mass spectrometry 9, 10 because the requisite sensitivity and

selectivityiswellestablishedandaccepted8incourtsoflaw.However,thesemethods

remainstationary,laboratorybased,andwaittimesforresultscanbelengthyiflaborious

sampleprocessing is required;allofwhichcanresult in frustratingdelaysat theearly

stagesofaninvestigation.

Recent research on explosive detection has begun to focus on ambient mass

spectrometry(MS)asitprovidesmechanismsforrapiddetectionandidentificationthat

doesnotrequirecomplexsamplepreparation.11Primarytechnologiesprovidingambient

MScapabilitiesincludedesorptionelectrosprayionization(DESI)anddirectanalysisinreal

time(DART);bothambientionizationsourcesweredevelopedintheearly2000s.11By

eliminatingtheconstraintssufferedbycommonionizationsourcessuchaselectrospray

ionization (ESI) and matrix assisted laser desorption/ionization (MALDI), ambient

ionizationsourcesprovidecapabilitiestoanalyzesamplesurfacesdirectlyinstitutingrapid

andhigh-throughputsamplingregimes.Inorganicandorganicexplosivematerialshave

both been identified and quantified using ambient ionization techniques. 12, 13, 14

However, literature lacks inprovingtheapplicabilityof thesetechniquestobeableto

identifyexplosivematerialsfromgenuinepost-blastfragments.

18

1.Introduction

1.1.ExplosivesExplosivesareenergeticmaterials thatupon ignitionundergorapidexothermic

decomposition to instantaneously release high pressure gas, heat and light. 15

Decomposition is predominately driven towards the production of more

thermodynamicallystableproducts,namelyCO2,N2andH2O.Thekineticstabilityofan

explosive is affected by conventional reactivity trends based on structure and bond

strength.Theweakoxygen-oxygenbondinperoxideexplosivesandthenitratespecific

carbon-oxygen bond in nitrated organic explosives increases reactivity leading to

productionof kinetically stableproducts. 16 In the caseofperoxides, instabilityof the

peroxidebond (-O-O-) isattributed inpart to theelectron repulsionbetween the two

electron-richoxygenatoms.16Thedegreeofkineticstabilityofanyexplosivewilllargely

determineignitionsensitivity.

Themechanismofanexplosionisinfactarapidcombustionreaction,whereinthe

oxidizerdecomposestosupplyoxygentosupportcombustionofthefuel.Ifthemixture

issufficientlysensitivetoshock,however,itwilldetonateinsteadofsimplyburn.17High

order explosives detonate creating a supersonic explosive shock front that travels at

velocitiesgreaterthan1000m•s-1.18Loworderexplosivesdeflagratebysurfaceburning

thatoccursatspeedslessthan1000m•s-1.18Explosivesextremelysensitivetoenergetic

stimuli are classified as primary explosives; often used as an initiator for larger less

sensitivemaincharges.Secondaryexplosivesarerelativelylesssensitivetoshock;often

used as themain charge as they are safer to handle, transport and store. 18 Tertiary

19

explosives are relatively insensitive requiring initiation by larger amounts of primary

and/orsecondaryexplosives,suchasso-calledboosters.18

To date, the number and varieties of different explosive materials has grown

immenselyalongwithmanydifferentapplications.Differenttypesofexplosivescanbe

categorized using a variety of classification schemes, depending on the property or

measurement of interest, including use, chemical composition, energetics, or blast

properties.Explosivesofsignificantconcernare thoseaccessible foruse incriminalor

terroristactivity.InaCanadiancontext,commercially-availablefirearmpropellantsand

consumer fireworks are commonly used as explosive fillers for IEDs, but homemade

explosives(HME)areanever-increasingthreatduetoeaseoffabricationusingreagents

sourcedfromcommonhouseholdchemicals.HMEincludeanyexplosivematerialthathas

beenalteredbeyonditsintendeduse,hasbeencreatedbycombiningproductstogether

orhasbeensynthesizedfromreadilyavailablereagents.19

Afterabombing, rapid sourceattributionof theexplosive fillerusedashaving

beeneitheracommercialorHMEproductcanprovidekeyforensiclinksfortheensuing

investigation.Equallyas important,early identificationofexplosive tracescanprovide

pivotal investigative forensic intelligence to help prevent a tragedy from occurring.

Therefore, delivering quality assured answers to front-line personnel with faster

turnaround times motivates improvement of the methods used for explosive trace

detection.Manydifferentanalyticalschemeshavebeendevelopedtoachievethisgoal.

20

1.1.1.Homemadeexplosives

1.1.1.1.PeroxideExplosivesOrganic peroxides are highly reactive compounds containing oxygen-oxygen

bonds.Theelectronrepulsionexperiencedbythelonepairofbothoxygenatomsinthe

peroxidebonddecreases theenergy required tobreak thebond. 20Organicperoxide-

basedexplosives(OPBE)requirenoconfinementtodetonateandproducehigh-pressure

shockwavestravelingatspeedsbetween4500-5500m•s-1,classifyingthemashigh-order

explosives.21Withrespecttosensitivity,OPBEareclassifiedasprimaryexplosivesdueto

extremesensitivitytoanyenergeticstimuli.21Thesematerialsposesignificantconcern

asthesynthesisusesreadilyavailableandcommonhouseholdproductsrequiringvery

basicknowledgeortraining.22UponsynthesisanduseasexplosivefillerinanIED,OPBEs

cause considerable damage and harm, as unfortunately demonstrated by several

domestic and international terrorist attacks in recent years. 23, 24 The detection and

identificationofOPBEsremainsacriticaloperationinensuringpublicsafety.

Triacetone triperoxide (TATP), hexamethylene triperoxide (HMTD) and methyl

ethyl ketone peroxide (MEKP) OPBE are the most common clandestine OPBEs

encounteredincriminal investigationsandforensic laboratories(Figure1-3).TATPand

HMTD are very sensitive to impact, heat and friction and find no legitimate use as

commercialorindustrialmanufacturedexplosives.25MEKPisslightlylesssensitiveand

hasfounduseindilutesolutionsasapolymerizationcatalystincommercialmanufacture

ofpolyesterandacrylicresins.26

21

The reagents required to synthesize OPBEs are commercially available and

syntheticmethodsareavailablefrommanyillicitinternetsources,suchaschatroomsor

other forms of social media. 22 Synthesis of OPBE occurs via step-wise insertion

mechanisticsteps. 26Ifan insertionstepproducesastableproduct the finalyieldmay

containmixturesof linearorcyclicdimer, trimerand tetramer forms;observed in the

synthesisofTATPandMEKP.26Thetrimeracetoneperoxide(i.e.TATP)isproducedasthe

mostabundantproduct,comparedtoamixtureofoligomersproducedinthesynthesis

of MEKP. 26 Purification of these synthetic products requires difficult and resource-

intensivemethodsresultingintheuseofimpureproductsinIEDs.Thehomemadenature,

limited solubility, lack of UV absorbance or fluorescence moieties and sensitivity to

mechanicalstresscreatesmanyanalyticalchallengesforthedetectionandidentification

ofOBPE.

Figure1.Triacetonetriperoxide(TATP).

22

Figure2.Hexamethylenetriperoxide(HMTD).

Figure3.Methylethylketone(MEKP).

23

1.1.1.2.BinaryExplosives Binaryexplosivesaremixturesconsistingoftwocomponentsblendedtogether,

namely: a combustible fuel and strong oxidizer. Separately, neither component is

explosive,butuponmixingtogetherinthecorrectratiotheresultingblendwillbehaveas

anexplosiveuponshockwithsufficient force.Upon ignitionofbinaryexplosiveshigh-

pressure shock waves traveling at speeds between 2500-4500 m•s-1 are produced

classifying them as high-order explosives. 27 With respect to sensitivity, most binary

explosives are classified as tertiary explosives as they require significant energy for

initiationandaresafetohandle,transportandstore.Mostcommerciallymanufactured

explosivesintendedforuseinmining,quarryingandblastingareabaseduponabinary

explosive formulation, albeit other different additives are regularly included (e.g.

emulsifiers,plasticizers,binders,etc).27Stringentregulatoryrequirementsstipulatethat

commercial explosives must be secured in licensed magazines to limit unauthorized

access and prevent theft. Consequently, clandestine fabrication of homemade binary

explosivemixtureshasbecomeanattractivealternativeforcriminalactivity,becausethe

oxidizer and fuel components can be sourced from commercial household products.

Many petroleum-based products can be used as the fuel source. Sugar-based food

products,suchasstarchorconfectionarysugar,canalsobeusedasasuitablecombustible

fuel source. Commercial fertilizers, compression-type instant cold packs and stump

removersareall sourcesofsuitablestrongoxidizers.Bysimplymixingorblendingthe

correctcombinationofcombustiblefuelwithastrongoxidizerabinaryexplosivecanbe

prepared.

24

Table1.Examplesofproductsthatsourcecomponentsforbinaryexplosives.

FuelPrecursors OxidizerPrecursorsComponent Source Component Source

Petroleumbased

DieselAutomotiveGrease

ParaffinWaxLampOilVaseline

PotassiumChlorate Textiles,matches,pyrotechnics

PotassiumPerchlorate

Airbaginitiator,pyrotechnics

Carbohydratebased

StarchSucroseFlour

AmmoniumNitrateFertilizers,Coldpacks,

ExplodingtargetsPotassiumNitrate StumpRemover

CurrentaccreditedmethodsutilizedbytheRCMPfordetectionandidentification

of binary explosives from recovered from post-blast residues include gas

chromatography-mass spectrometry (GCMS), ion chromatography-mass spectrometry

(ICMS) and Fourier-transfer infrared spectroscopy (FTIR). Chromatography methods

combinedwithmassspectrometryarewellsuitedtoseparatethedifferentcomponents

of binary explosives and identify each component in isolation. The specific choice of

whichtechniqueisusedwilldependonthechemicalcompositionandphaseoftheactual

binaryexplosiveexamined(e.g.ICMSforwater-solublesalts,GCMSforvolatileorganics).

With the identification of both components, the overall energetic mixture can be

exposed.

25

1.1.1.3.SmokelessPowderProducts Modernsmokelesspowdersarethepropellantsusedinsmallfirearmammunition.

28 Ignition of the propellant the energy released, via formation of gaseous products,

resultsintheejectionofthebulletfromthechamberofagun.28Smokelesspowdersare

largely produced and used in the assembly of self-packed ammunition. However,

smokelesspowdercanalsobeprocuredforillicituseinIEDs.28,29Theseproductsproduce

negligible smoke when ignited and burned as they are largely composed of organic

explosives that produce only CO2 and H2O gaseous products upon combustion. In

contrast, other propellants, such as black powder, mainly produce solid, non-volatile

productsuponcombustionthusproducingairborneblacksootvisibleassmoke.

Nitrocellulose(NC)isthebaseorganicexplosiveusedtomanufacturesmokeless

powders (Figure4).However,nitroglycerine (NG) canalsobeused in certain typesof

smokeless powders to increase the output energy (Figure 5). Single-base smokeless

powdercontainsNConlyanddouble-baseproductscontainingNCandNG.Adouble-base

smokelesspowdermaycontainbetween10-50%NGcontentbyweightdependingonthe

productused.BecauseNGcandetonateithasthepotentialtoshatterafirearm.NCis

obtained from nitration of cellulose; nitration is an exothermic esterification reaction

wherebyvariouspendanthydroxylgroupsarenitratedbuttheb(1-4)linkagesbetween

monomerunitsinthecellulosechainarenotbroken.30Eachglucosemonomercontains

three potential hydroxyl groups that can be nitrated. The degree of total nitration

dependson thecellulosesourceand the reactionconditions. 30NCcanbenitrated to

different,varyingdegreesandthefinalextentofnitrationcanaffectthecommercialuse

oftheNCprepared.Highly-nitratedNCisconsideredtohaveanitrogencontentofatleast

26

14%byweightandistheformusedinsmokelesspowders.30LessernitratedNCisused

to fabricate cigarette paper and party streamers. Stabilizers, plasticizers and surface

coatings are used in different smokeless powders tomodify or improve performance

characteristicsandprolongshelflifeofthefinalpropellantpowder.Acommonexample

isdiphenylamine(DPA)whichisaweakbaseaddedtosmokelesspowderstoneutralize

theslightly,naturallyacidicNCandpreventspontaneousdecompositionovertime(Figure

6).28,29Anothercommonexampleisethylcentralite(EC)whichisabothaplasticizerand

flameretardanttoraiseignitiontemperatureandslowtheburningrateofthepropellant

powder(Figure7).28,29

Figure4.Fullynitratednitrocellulose.

27

Figure5.Nitroglycerin.

Figure6.Diphenylamine.

Figure7.Ethylcentralite

28

Methods to characterize NG, NC and the various additives from smokeless

powdershavebeendeveloped.31-35However,currentmethodstoidentifynitrocellulose

relyuponrecoveryofanintactpropellantgrainfromwhichthevariousconstituents(NC,

NG, additives) can be extracted and characterized. No method has been previously

reportedtoidentifynitrocellulosepostblastbasedonitsthermal-degradationproducts

in the absenceof a recoverable intact grain for analysis. Identificationof the thermal

breakdown products of nitrocellulose remains challenging due to absence of a

characteristicreferencematerial.

29

1.2.ExplosiveAnalysis

1.2.1.IonMobilitySpectrometry(IMS) Ion mobility spectrometry (IMS) is a commonly used technique to screen for

contraband at border security checkpoints, including concealed drugs and explosives.

Detectionofillicitmaterialsoccursbyobservingcharacteristicmobilityofionsconverted

fromsamplevaporsinaweakelectricfield.36MostIMSinstrumentationisengineeredto

includefourmainsub-components:anionsource,aniongate,adrifttubeandadetector

(Figure8).36Uponionizationofsamplevaporsinthesource,theiongateelectronically

ejectsionsintothedrifttubewherebyanelectricfieldisapplied.12Asionsexperience

theelectricfieldtheymovetowardsthedetector,whichinmostdevicesisaFaradaycup.

36Uponcollisionalneutralizationatthedetector,currentflowiscollectedasameasurable

signal.36Amobilityspectrumisproduced,plottingioncurrentagainstdrifttime.Based

solely on a specimen’s drift time, detection and identification of illicit materials is

achievable.

Figure8.Schematicdiagramofanionmobilityspectrometer(IMS).

30

Ionmobilityisonepreferredmethodusedtoscreenforexplosiveresiduesasitis

relativelyinexpensive,easytouseandprovidescapabilityforrapidanalysisthatisfield

deployable,allwhilemaintaininghighsensitivity.36However,duetosinglemechanism

discrimination(i.e.drifttime)lossofsensitivityandselectivityduetomatrixinterference

arising from environmental contaminants and other congeners remains a significant

critiqueofIMStechnology.Therefore,withrespecttoidentificationanddetectionthese

methodslargelyremainpresumptivetests.Duetothecomplexityofpost-blastsamples

multiplexedanalyticalschemesarerequiredtouniquelycharacterizethevariousorganic

andinorganiccomponentsofexplosiveresiduesapartfrommatrixinterference.37

1.2.2.ChromatographyMethods Many forensic laboratories are equipped with severeal chromatographic

instrumentation as they remain the gold-standard techniques for separation,

identification and quantification of compounds in a mixture. 38 All chromatography

methodsincludeamobileandstationaryphase.Physicalseparationofamixtureisbased

ondifferentpartitioningfactorsofcomponentsinthemixturebetweenthemobileand

stationary phases. 38 Factors such as adsorption, affinity, polarity and size affect

separation processes. 38 Many different detectors are coupled to chromatography

instrumentationprovidingidentificationmechanismsbasedonstructure,mass,charge,

volatility and polarity. Separation techniques are included in many forensic practices

because forensically-relevant samples often contain many unknown compounds in a

complexmixture.Inclusionofchromatographymethodsprovidesensitiveandselective

detection,identificationandquantificationofforensicallyrelevantcompounds,suchas

31

explosives. Therefore, much attention has been invested in using chromatography

methodsforexplosiveanalysis.9,10

High-performanceliquidchromatography(HPLC),gas-chromatography(GC)and

ion chromatography (IC) are themost common chromatographymethods utilized for

explosiveanalysis.39 HPLC employs a closed, pressurized column containing a solid

phase(Figure9).Athigh-pressuresthemobilephaseispassedthroughacolumncarrying

components to be separated. 38 Separation occurs via differences in analyte relative

affinityforthemobileandstationaryphases.Thepolarityoftheanalytesdictatesrelative

affinityformobileandstationaryphasesprovidingamechanismforphysicalseparation

ofthecomponents.38Samplemixturesareinitiallyloadedontothecolumnviaaffinityfor

thestationaryphase.38Uponaswitchinpolarityofthemobilephase,analytesaredriven

backtothemobilephaseandelutefromthecolumn.38HPLCisadesirableseparation

and identification technique for explosive analysis due to itswell-respected accuracy,

efficiency and reproducibility. Methods can be set up in a highly automated fashion

allowingforhigh-throughputanalysesofawidevarietyofsamples.

Figure9.Schematicdiagramoftheinstrumentationusedtoseparateanalytesofa

mixtureviahigh-performanceliquidchromatography(HPLC).

32

Methods involving GC are used in explosive analysis for detection and

identification of volatile compounds. GC columns contain a liquid stationary phase

adsorbedontoaninertsolid.Themobilephaseisusuallycomposedofinertcarriergases

(e.g.heliumornitrogengas).Volatileanalytesenter thegaseousmobilephaseandas

they pass through the column, depending on relative affinity for the mobile phase,

separationoccurs (Figure10). 14As a simple,multi-faceted, rapid andhighly sensitive

method, GC has proven its ruggedness and robustness while providing appreciable

sensitivityandselectivityrequiredforexplosiveanalysis.33-35

Figure10.Schematicdiagramoftheinstrumentationusedtoseparatevolatilecomponentsinamixtureviagaschromatography(GC).

33

Due to thewidevarietyofexplosivematerialsposingsignificant threats toour

safetyandsecurity, capabilitiesofouranalyticalmethods todetectand identify them

continuetobechallenged.11Whilemanyexplosivesareorganiccompoundsamenableto

HPLC and GC methods, inorganic explosive classes are not compatible. Ion

chromatographyisthepreferredseparationtechniqueformanyexplosive-relatedionic

species such as ionic salt oxidizers used in binary explosivemixtures (Figure 11). 40, 41

Basedonelectrostaticinteractionsbetweenmobileandstationaryphases,separationof

ionicspeciesoccurs.ChangesinpH,concentrationofionsaltsandionicstrengthofthe

bufferedmobilephaseareusedtoeluteionsfromthecolumn.38Bothanion-exchange

and cation-exchange columns are available. IC is a commonly used, quality assured

methodforanalysisofinorganicexplosives.

Figure11.Schematicdiagramoftheinstrumentationusedforsimultaneousseparation

ofanionsandcationsinsolutionviaionchromatography(IC).

34

Chromatographymethodsremainprimarydetectionandidentificationmethods

forexplosiveanalysis.However,thestationaryandlaboratorybaseddesignandrequired

laborious sample preparation creates lengthy wait times for results. Associated

frustratingdelayscausedatearlystagesofaninvestigation,duetolimitationsassociated

withchromatographymethods,motivatesadaptation,innovationandvalidationofnew

methodsandtechniques.Ambient-ionizationmassspectrometrymethods,suchasDART-

MS,havebecomelucrativemethodsfordetectionandidentificationofexplosivesasthey

are simple, facile and rapid methods that still maintain robustness, reproducibility,

sensitivityand selectivity required for court. 11, 42-47 Furthermore,due to compatibility

withlow-resolutionmobilemassspectrometersthereispotentialformobilityfromthe

labtothecrimescene.11TheapplicabilityofDART-MSfordetectionofmanydifferent

nitro, nitrosamine and nitroaromatic explosives has been studied excessively due to

prolific use of these explosives in IED. 48 However, adaptability to combat the more

contemporaryforensicchallengesassociatedwithdetectionofhomemadeexplosivehas

yettobeexplored.

35

1.2.3.MassSpectrometry(MS) Since invention intheearly1900s,massspectrometry(MS)remainsoneofthe

mostpowerfulanalyticaltoolsavailable.Predominantlyusedforstudyandrecognitionof

matter by filtering substances based on mass-to-charge ratio (m/z). 49 Versatile

applications of mass spectrometers results in inclusion of these instruments in

laboratories of many scientific disciplines all around the world. 50 With continuous

innovation, adaptation, andmodificationsMSmethods continue to prove robustness,

sensitivityandselectivity.

Thefundamentalconceptsandengineeringofmassspectrometershasremained

constant since invention. As described, characterization of an analyte via MS is

accomplishedbyionization,filtrationanddetectionofgaseousanalytespeciesbasedon

the m/z. 50 Instrumental configurations of mass spectrometers include four main

components: vacuumsystems,an ionization source,amass filterandadetector. Ions

producedinthesourceareacceleratedintoanelectricfieldwherebyseparationbased

onm/ztakesplace(i.e.massfilter).Compoundswithslightlydifferentmassesresultin

variable m/z and unique trajectories through the mass filter providing a robust

mechanism for differentiation. As ions reach the detector the electrical response is

plotted against m/z to create a mass spectrum exposing identity and abundance of

speciespresentinthesample.17Vacuumsystemsareincorporatedtoreducelikelihood

andfrequencyofioncollisionsresultinginpotentialchargetransfersultimatelyhindering

thepathofionsfromsourcetodetector.50

36

CompatibilityofananalytewithMSmethodsdependsonitscapabilitytotransfer

to the gaseous phase and become ionized. To facilitate this required transformation,

manyionizationsourceshavebeendeveloped.Innovationhasledtomanymodifications

andadaptationsproducingalonglistofionizationsourceswithcompatibilitytoawide

variety of analytes such as: small molecules, inorganic compounds, large organic

compounds and biomolecules. Ionization sources can produce ions with negative or

positive charges. Historically, ionization sources have been located inside the mass

spectrometerundervacuumbuttodatemanyambientionizationsourceshavebecome

available.51Ionizationsourcesarepredominantlyclassifiedbyvacuumrequirements,but

also are classified by the strength of the ionization. Hard-ionization sources describe

ionizationofamoleculebyproducingfragmentionsfromparentanalyte.Soft-ionization

mechanismspredominantlyproduceionizedparentionswithlittletonoproductionof

fragmentions.

Productionofionsinthesourcemigratethroughthemassfiltertothedetector.

Manyfiltersareavailableandingeneral,manipulateanestablishedelectromagneticfield

tocontrolthesuccessfultrajectoryofanionfromthesourcetothedetector.Massfilters

aredifferentiatedbyshape,size,andmaterialofthecomponents;ultimatelyleadingto

differences in resolution andmass limits. Low-resolutionmass spectrometers (LRMS)

measurem/zbywholenumbermassesofatoms;high-resolution(HRMS)instrumentation

provides superior mass accuracy by measuring the exact mass of each atom to the

thousandth decimal place. 51 HRMS instrumentation becomes extremely useful when

37

masses of many analytes are similar as it has the power to resolve and uniquely

characterizesimilarmasscompounds.51

Detectionofionsoccursbyconvertingtheelectricalresponse,createdwhenan

ion reaches the detector, into representable and readable signals. Many different

detectors are on the market today. Examples of commonly used detectors include

photoplates, photomultipler tubes, Faradaydetectors, electron-multipliers and image-

currentdetectors.

In most cases, post-blast explosive analysis relies solely on detection and

identificationofresiduesfromanyunconsumedexplosivesfoundpost-blast.However,

thedestructivenatureofanexplosion(heat,pressure,oxidationandpyrolysis)creates

considerablechallengesasanyundetonatedmaterialtypicallyoccursinsmallamounts

andisspreadovermanyfragmentsacrosslargeareasofland.Therefore,qualityassured

methods and techniques are required for success. MS remains one of the most

predominant methods used for explosive analysis from trace quantities of residues

collectedpost-blast.

38

1.2.3.1.DirectAnalysisinRealTime(DART)IonizationDART, an ambient soft-ionization technique, was designed and engineered by

Codyetal.in2005.50Afteradecadefrominitialrelease,DARTsourceshavefoundtheir

wayintomanyfood,environmental,healthandindustryrelatedlaboratories.DARThas

foundspecificapplicationinfoodanalysis,chemicalidentificationandcharacterization,

pesticidedevelopmentanddetection,drugdevelopmentandscreeningandforensics.42-

47, 54-55 Particularly with respect to forensic analysis, DART-MS has proven to be a

powerfullyreliabletechniqueforexplosiveanalysis.42-47

Operating in ambientenvironmentswithout samplepreparation requirements,

DARTprovidescapabilitytoionizeliquid,solidandgaseoussamplesintheirnativeform.

49 Ionizationofanalytemolecules thermally-desorbedfromsamplesurfacesoccursvia

production and resulting reactions with metastable species produced by the source.

Heatedinertgasessuchashelium(He),argon(Ar)ordinitrogengas(N2)enterthesource

andpass througha seriesof electrodesproducingmetastable species suchas excited

statehelium(He*)(Scheme1).Uponexitingthesource,adrystreamofexcitedgaspasses

throughafinalelectrodedirectingionstotheMS,removinganychargedmoleculesto

preventundesiredionrecombinationandcontrollingthepolaritymode(i.e.positivevs.

negativemode)(Figure12).53

M+energy=M*

Scheme1.Reactionschemefortheelectronicorvibronicproductionofmetastablespecies(M*)frominertgas(M)occurringinthesourceviaaseriesofelectrodes.

39

Figure12.SchematicdiagramofDARTsource.

Mechanismsresponsibleforproducingionizedanalytemoleculesaredictatedby

theinertcarriergasused.HeandN2arethemostcommonwiththeformerreportedto

be most effective. 53 Metastable species exiting the source react with atmospheric

moleculestoproducereagentions.UponreactionbetweenHe*andatmosphericspecies,

ionizedreagentmoleculesareproducedviaPenningionization.19Ionizedwaterclusters,

the primary reagent ion produced, are responsible for consecutive ionization of

thermally-desorbedanalytemolecules.53Applicabilityofthisionizationsourceislimited

tomoleculeswithmassrangesfromm/z50-1200asmanycompoundsoverm/z1200lack

requiredvolatility. 53Exactmechanismsforproductionof thesereactivespecies isnot

clear however further investigation has led to a few proposals. 49 Protonation,

deprotonation, direct charge transfer and adduct ion formation are key mechanisms

responsible for production of positively charged analyte molecules (Scheme 2, 3).

Negativelychargedanalytemoleculesareproducedbyflippingthepotentialontheend

gridelectrodetonegativepotentials.ElectronsproducedbyPenningionizationundergo

electroncapturewithatmosphericmolecules inthereactivezonebetweenthesource

andtheMS(Scheme4).49

40

(a) M*+N2àM+N2+•+e-

(b) M*+H2OàM+H2O+•+e-

(c) N2

+•+N2+N2(3rdbody)àN4+•+N2(3rdbody)

N4+•+H2Oà2N2+H2O+•

(d) H2O+•+H2OàH3O++OH•

H3O++nH2Oà[nH2O+H]+

Scheme2.Reactionschemestoproducesecondaryionizingspecies(ionizedwaterclusters)inpositivemodeviareactionofmetastablespeciesproducedbytheDART

sourcewithatmosphericreagents.

S+[nH2O+H]+à[S+H]++nH2O

S+N4+•àS+•+2N2

S+O2+•àS+•+O2

S+NO+àS+•+NO

S+[NH4]+à[S+NH4]+

Scheme3.Reactionschemestoproduceionizedanalytespecies(S+•)inpositivemodeviareactionofsecondaryionizedspeciesandanalytemoleculespresentedtothesource

region.

41

O2+e-àO2-•

S+O2-•àS-•+O2

S+e-àS-•

SX+e-àS-+X•

SHà[S]-+H+

Scheme4.Reactionschemesforproductionofnegativeionizedanalytespecies(S-)via

In summary, operation in positive mode predominantly produces protonated

analytemoleculesandinnegativemodeproducesdeprotonatedmolecules.49Depending

on theanalytemolecule,other ionized speciesmaybe favorable (i.e. ammoniumand

chloride adducts). Coupling to amass spectrometer, analysis in eithermode provides

relativelysimplemassspectra.

Tofacilitatethetransitionofionsfromtheambientionizationreactionzonetothe

massspectrometerunderhighvacuum,aninterfacehousingskimmerorificewithslight

potentialdifferencesbetweenthemisinstalledonthefrontend.53Thesecomponentsof

theinterfaceareresponsibleforremovingneutralcontaminationsanddirectingionized

species into theMS.A roughpumpconnectedto the interfaceand isusedto remove

neutralcontamination.

42

Operation of DART-MS can occur in surface desorptionmode or transmission

modesimplybymanipulatingorientationofthesourcewithrespecttotheMS-interface.

Set at 45° (Figure 13 (a)) surface desorption of analytemolecules occurs providing a

simple, easy to usemethod for direct analysis. Set at 0°, (Figure 13 (b)) transmission

analysisofsamples ispossibleprovidingmechanismsforanalysisof liquidsorsamples

foundonporousmaterials.Commerciallyavailableconsumablesupportmechanismsare

availableforanalysisofsolidandliquidsamples.Bysimplyplacingthesamplesontoretro

fitted supports, which are set into a mechanical rail (Figure 16. (b)), a reproducible

mechanismtomovethesamplesintoandoutofthereactionzonebetweentheDARTand

theMSisprovided.

Figure13.ImagesofthedifferentorientationsoftheDARTsourcewithrespecttotheMSinterface:(a)surfacedesorptionmodeand(b)transmissionmode.

(a) (b)

43

1.2.3.2.Q-ExactiveMassSpectrometer As previously described, the mass analyzer is a key component of a mass

spectrometerasitprovidescapabilitytofilterandselectivelydetectspecieswithspecific

m/z.LRMS includemassanalyzerssuchasquadrupolesand linear iontrapswhichcan

detectionsbasedonnominalm/z.HRMSdemonstratesitspowerofofferingaccuracy,

sensitivity and selectivity simply by providing m/z measurements to the thousandth

decimal place. 52 Instrumental platforms available include Fourier-transform (FT) ion

cyclotronresonance(ICR),time-of-flight(TOF)andOrbitrapmassanalyzers(Figure14).52

UsingHRMSforexplosiveanalysisprovidescapabilityfordetectionandidentificationof

explosives from complex sample matrix with the sensitivity, selectivity, and accuracy

requiredbythecourtsoflaw.

By combining the Orbitrap and quadrupole instrumentation, the hybrid

technology coined the Q-Exactive was released in 2011. 56 Since original design and

engineering theQ-Exactiveencompassesaquadrupole,C-trap,High-energyCollisional

Dissociation (HCD)celland theOrbitrapmassanalyzer. 56Selectivityofdesired ions is

offeredbymanipulatingtheelectromagneticfieldestablishedbetweenthequadruples.

57Ionsthatdonothavethespecificm/z,migrateirregularlythroughthefield,crashing

intotherodsorthesidesofthemassanalyzeranddonotreachthedetector.57TheC-

trapcollectsionsintopacketspriortoinjectionintotheHCDortheOrbitrap.Withinthe

HCDcell,fragmentationofionsoccurswhenanincreaseinthekineticenergyoftheions

results in collisions with neutral molecules. Due to a conversion of kinetic energy to

internalenergyuponcollision,bondsbreakandfragmentsareproduced.58FromtheHCD

cell, newly produced fragments are sent back into the C-trap for re-focusing and

44

subsequentlyinjectedintotheOrbitrapforthefinalmassanalysisandproductionofthe

massspectrum.58TheOrbitrapmassanalyzerisanelectrostaticdevicethatconsistsofa

central,spindle-shapedelectrodethationsoscillatearound.59Bydetectingaxialmotion

around the inner electrode the signal produced is Fourier-transformed yielding high

resolutionmassspectra.24-27OperationoftheQ-Exactiveinfullscanmodedoesnotutilize

theHCDcell.Simplyswitchingoperationmodestotheallionfragmentation(AIF)mode

providesthecapabilityforMS/MSanalysis.

Theouterelectrodeissplitupsymmetricallyservingasasensorsurroundingthe

centralelectrode.56Detectionoftheionsoccursviaimagecurrentdetectionwherebythe

encapsulatingouterelectrodesmapsthecurrentinducedbyaxialmotionofionsaround

thecentralcylindricalelectrode.56Thedatacollectedbythesensorsisconvertedtom/z

byFouriertransformation.

Figure14.SchematicdiagramofthedesignandengineeringoftheQ-Exactivehybridmassspectrometer.

45

Duetothecomplexsamplematrixcreatedwhenablastoccurs,itisimportantto

demonstrate and verify the capability of ambientMS techniques to identify explosive

residuesfrompost-blastsampleswithappropriatesensitivityandselectivityrequiredby

courtsoflaw.Thescopeofthisthesiscomprisesanexplorationintotheapplicabilityof

DART-MS to characterize typical homemade explosives (HME) used in simulated

improvisedexplosivedevices(IEDs).

46

2.MaterialsandMethods

2.1.Consumables,ReagentsandStandardReferenceMaterialsAll consumables, reagents and solvents purchasedwere used as received.ACS

gradeorbetteracetone(>99.7%,CaledonLaboratoriesLtd.,Georgetown,ON),methanol

(>99.8%,CaledonLaboratoriesLtd.,Georgetown,ON)andacetonitrile(≥99.9%,Sigma-

Aldrich,Oakville,ON)wereusedassolvents;hydrogenperoxide(50wt.%,Sigma-Aldrich,

Oakville, ON), sulfuric acid (95.0 -98.0%, Caledon Laboratories Ltd, Georgetown, ON),

citricacid(≥99.5%,Sigma-Aldrich,Oakville,ON),nitricacid,anhydrideaceticacid,D(+)-

cellobiose,hexamethylenetetramine(≥99.9%,Sigma-Aldrich,Oakville,ON),andmethyl

ethyl ketone (ACP Chemicals Inc., Montreal, Quebec) were used as HME synthesis

reagents; diesel, Vaseline, lamp oil, wax, grease, sugar, dextrin, and stump remover

(collectively purchased locally), ammonium nitrate, potassium perchlorate, potassium

chlorate were all used as either the fuel or oxidizer source for homemade binary

explosives;BlueDotsinglebasesmokelesspowderandGreenDotdoublebasesmokeless

powder commercial explosive products were used as purchased; certified explosive

standardswerepurchasedassolutions(0.1mg·mL-1 ineithermethanoloracetonitrile)

fromChromatographicSpecialtiesInc.(Brockville,ON).

47

2.2.IEDConstruction,DetonationandFragmentCollectionMultiple IEDs were assembled to yield a variety of post-blast fragments

characteristicofmaterialscommonlyrecoveredpost-blastandreceivedatourlaboratory

analysis,suchas:cellphones,wires,batteries,nuts,bolts,metalswitches,andmechanical

timers(Figure15).Forsafetyreasons,thecontainerusedforeachIEDwasathin-walled

aluminumcan.EachIEDwasconfiguredwithamaincharge(Table2)andfiredusinga

commercialelectricblastingcap(approximately1gPETN).TATP,HMTD,MEKPandall

binaryexplosiveswerepreparedbyaqualifiedchemist,inaccordancewithstandardbest

practicesand incompliancewithCanadianExplosivesRegulations(SOR/2013-211)and

characterized before use to demonstrate fit for purpose. The MEKP prepared was

confirmedbyLCMSanalysistocontainthelineardimer,trimerandtetramerasthemajor

constituents, and the cyclic trimer as aminor constituent. 27 Bluedot single-base and

GreenDotdouble-basesmokelesspowderwereplacedinindividualvials,aspurchased.

Figure15.Fragmentscollectedpost-blastfromthedetonationofdevicesutilizinghomemadeexplosivesasthemaincharge(IED#1-14).

48

Tolimitthespreadofdebrisandergonomicallyfacilitatefragmentcollection,each

IEDwasenclosedbycinderblocksbeforedetonation.EachindividualIEDwasremotely

detonatedusingacommandwireatRoyalCanadianMountedPolice(RCMP)Technical

andProtectiveOperationsFacility(TPOF)inOttawa,ON,byaqualifiedRCMPExplosive

Disposal technician. Conventional contamination-prevention protocols were followed

consistent-withactualcrimescenepractices,andallfragmentscollectedfromeachIED

sealedintonylon-linedevidencebagspriortotransportandduringstorage.

Table2.TypeandamountofexplosiveusedasmainchargeforeachIED.

IED# MainCharge Amount(g)1 TATP 402 TATP 53 HMTD 34 HMTD 5

5 MEKP 56 Perchlorate+Vaseline 107 AN(prills)+Diesel 10

8 AN(ground)+Wax 109 Chlorate+lampoil 1010 KNO3+grease 10

11 Chlorate+sugar 10

12 KNO3+dextrin 1013 Singlebase 20

14 Doublebase 20

49

2.3.ReferenceMaterialSamplePreparation

2.3.1.FuelandOxidizer The commercially available petroleum-based fuels used in the binarymixtures

were obtained and analyzed unmodified. Carbohydrate-based fuel products used to

prepare the binarymixtures were obtained and reference samples were prepared in

deionizedwater.Referencesamplesoftheoxidizersusedinthedeviceswereprepared

indeionizedwater.

2.3.2.OctanitrateCellobioseSynthesisandSamplePreparationOctanitrate cellobiose was synthesized and used as a reference material for

nitrocellulose.b-cellobiosewaschosenasthestartingmaterialbecauseitcontainsthe

same1-4-b-linkageasobserved innitrocellulose.Nitrationofb-cellobiosevia reaction

with anhydride acetic acid and fuming nitric acid produced the desired b-cellobiose

octanitrate(ONCB).Thefinalproductwasfiltered,usingawateraspirator,andwashed

multipletimeswithsodiumbicarbonatetoneutralizeanyresidualacid.Characterization

ofthefinalproductwasdonebyFTIRandmatchedliteraturevalues.

Approximately5mgONCBwasdissolvedin5mLacetone,methanol,acetonitrile,

water and dichloromethane. Additional samples of approximately the same mass of

ONCBwereplacedinglassPetridishesandexposedtoabutaneflamefromabarbeque

lighter.DuetotheenergeticpropertiesofONCB,thewhitepowder-productburnedtoa

stickysyruplikeresidue.TheresiduesweresubsequentlycollectedfromthePetridishes

using 5 mL acetone, methanol, acetonitrile, water and dichloromethane. No further

purification,filtrationorpre-concentrationwasconducted.

50

2.3.3.SmokelessPowderSamplesofapproximately10mgsingle-baseanddouble-basesmokelesspowders

were prepared similarly to the preparation methods for ONCB. Samples of both

smokelesspowdersweredissolvedindividuallyin5mLacetone,methanol,acetonitrile,

water and dichloromethane. In addition, samples were ignited and the remaining

residueswerecollectedfromthePetridishesusing5mLacetone,methanol,acetonitrile,

wateranddichloromethane.Nofurtherpurification,filtrationorpre-concentrationwas

conducted.

2.4.Post-BlastExtractPreparation Solventextractsofresiduesfrompost-blastfragmentsisanidealsamplingmethod

forobjectstoolargeorirregularlyshapedfromblastdamage;whichmaybeunsuitable

to be shipped to the laboratory and/or analyzed directly in the DART sample region.

Residuesfromthepost-blastfragmentsforthebinaryexplosiveswerecollectedbyrinsing

thefragmentswithhexane(petroleumbasedfuels)orwater(sugarbasedfuels).Residues

were collected from fragments from the smokeless powder device using acetone.

Nitrogengaswasusedasamechanismtopre-concentratetheextractswhenrequired.

Nopre-concentrationoffiltrationoftheextractswasincluded.

51

2.5.DART-MSAnalysisDatacollectedinpositiveandnegativemodewasdonewithanIonSense®DART

SVP 100 source coupled to a ThermoFisher Scientific Q Exactive™ (Orbitrap) mass

spectrometer(ThermofisherScientific,Waltham,MA).TheDARTsourcewaskeptat250°

Cand350Vgridvoltage.MSscan rangeofm/z50–700,witha resolutionof70000

FWHM. Thermo Scientific XCalibur software was used for data collection and

QualBrowsersoftwareforqualitativedatainterpretation.Fullscanacquisitionwasused

andnodopantwas added to promote adduct formation. Sampleswereprobedusing

three different DART configurations; direct analysis, analysis of swabs and extracts.

Fragmentswereanalyzeddirectlybyplacingtheminthesamplingregionforanalysisat

45°,whereasswabsandextractswereanalyzedat0°.Rectangularcottonswabs(4x4cm,

SmithsDetection,Mississauga,ON)wereused towipe fragments, andwere collected

eitherdryormoistenedwithasinglesolvent:acetone,methanoloracetonitrile.Swabs

with residue sub-sampled from fragmentswere thenanalyzedbyplacing them in the

samplingregion.ExtractsweredepositedontoanIonSenseQuickstrip(consumablecard

withstainlesssteelmeshwells),placed intoacardholdersittingonamechanical rail,

responsibleformovingthesamplestoandfromthesamplingregion.

52

3.ResultsandDiscussion

3.1DART-MSParameterOptimizationPrior tochallengingtheapplicationofDART-MSto identifyHMEresiduespost-

blast, a bench level method was validated in the laboratory. Target explosives were

analyzedinreplicateusingaQuickStripcardwhilesystematicallyvaryingDARTandMS

settings to ensure reproducibility and robustness of detection. Parameters optimized

included temperature of the DART probe, distance between the DART and MS and

resolution of the Orbitrap. Limit of detection (LOD) was evaluated to determine the

lowestquantityofexplosivesresultinginobservationofthreeormorecharacteristicions

with ≥ 3:1 signal-to-noise ratio (3σ). The LODsmeasured (Table 3) were found to be

comparabletoreportedvalues.44OncesatisfiedwiththeDART-MSparametersandthe

methoddesign,theapplicabilitytodetectexplosiveresiduesfromthepost-blastbomb

fragmentsandidentifythevarietyHMEusedinthedeviceswasinvestigated.

Table3.LODsobservedforexplosivesofinterestmeasuredusingin-housemethodandQuickStripcomparedtoliteraturevalues.

Explosives MeasuredLOD(ng) Reported45LOD(ng)TNT 0.01 0.25HMX 0.10 10RDX 0.01 0.50PETN 0.10 5Tetryl 0.10 1NG 1 5

2,4-DNT 10 0.502,6-DNT 10 0.50EGDN 100 100HMTD 1 -TATP 100 -MEKP 100 -

53

3.2.AnalysisofFragmentsDue to the design and capability of the DART-MS method, multiple sampling

regimes (e.g. direct, in-direct and extracts) were evaluated to determine efficacy of

residue recovery. Each sampling method revealed the HME used, however different

recoveryefficiencieswereexposed.Asidefromminordifferencesinrelativeabundance

ofions,sample-to-sample,themassspectraandcorrespondingfragmentationpatterns

forallexplosiveswereinhighagreementwiththereferencematerials.

As expected without pre-concentration or sample clean-up prior to analysis,

background ions were commonly observed in the DART-MS spectra. Spectra were

collected prior, during and after each sample analysis to monitor cleanliness of the

instrument and to identifybackground ionsendogenous to theDARTor the sampling

technique(i.e.swabbingoruseofQuickstrip)(Figure16-19).Analysisofbothpolarand

organic solvents deposited on theQuickStrip produced spectrawith total ions counts

ranging from 104 – 106 (Figure 20 and 25). To facilitate elimination of the solvent

contribution,solventblankswereincludedintheanalysisalongwithextractedsamples.

54

Figure16.IonsobservedviaoperationoftheQExactiveinpositivefullscanmode,

withouttheDARTsourceturnedon.Totalioncount103.

Figure17.EndogenousDART-MSions.Totalioncount104-105.

55

Figure18.Analysisofanunusedcottonswabinpositivemodeusingfullscan.

Totalioncount104

Figure19.AnunusedQuickStripanalyzedinpositivemodeusingfullscan.

Totalioncount103-105.

56

Figure20.WaterdepositedontoaQuickStrip,analyzedinpositivemodeusingfullscan.Totalioncount104–105.

Figure21.MethanoldepositedontoaQuickStrip,analyzedinpositivemodeusingfull

scan.Totalioncount104–106.

57

Figure22.AcetonedepositedontoaQuickStrip,analyzedinpositivemodeusingfull

scan.Totalioncount104–106.

Figure23.AcetonitriledepositedontoaQuickStrip,analzyedinpositivemodeusingfull

scan.Totalioncount104–106.

58

Figure24.AnalysisofdichloromethanedepositedontoaQuickStrip,inpositivemodeusingfullscan,tobeusedforsolventassociatedionsubtraction.Totalioncount104–

106.

Figure25.AnalysisofhexanedepositedontoaQuickStrip,inpositivemodeusingfullscan.Totalioncount104–105.

59

DetectionandidentificationoftheHMEusedforeachIEDwasobtainedbydirect

analysis of at least one fragment recovered from each device. In real casework it is

commontonotdetectexplosiveresiduefromeverybombfragmentduetotheunequal

geospatialdistributionofresidueswhichisadirectresultoftheunpredictablenatureof

anexplosiveblast.62Thephysicalsizeofthefragmentsfoundvariedfromsmall(e.g.SIM

card~0.25 cm2) to large (e.g.D-cell battery ~17.25 cm2),whichmeans fragments can

retaindifferentquantitiesof residue.The interactionareaof theDARTprobewill not

instantly desorb all residue from surfaces, which will limit the amount of material

transmittedtotheMSinletperscan.Moreover,manyofthefragmentsfromeachdevice

wereinstorageforatleastfour(4+)monthspriortoanalysis,whichisnotanuncommon

time-frame with actual investigations. Thus, some extent of residue loss from the

fragment surface(s) via degradation can be expected. 59-61 Recovery efficiencies with

longerexposuretimesofpost-blastIEDfragmentstoairorweatherwasnotquantified,

butwillbegivenfurtherconsiderationinfuturework.

In-directanalysiswasdonetoevaluatecompatibilityoftheDART-MSmethodwith

using swabs. Swabs are ideal to sub-sample objects too large to be shipped to the

laboratory or irregularly shaped fragments unsuitable for direct analysis. Analyzing a

fragmentdirectlymayalsocontributetohigherbackgroundintheMSifthesubstrateof

thefragmentitselfcanbedesorbedandionizedbytheDART,makingswabsanattractive

alternative.Inaddition,swabbingissuitedtoaccumulatemoreresiduefordesorptionby

theDARTprobeandtransmissiontotheMSinletperscan,whichwillenhancerecovery

and detection. Efforts tomaximize residue collection using swabs included evaluating

60

commercially available swabs manufactured with different substrates as well as

comparingcollectionefficiencyandeffectivenesswhenusingdryorsolvent-dampened

swabs.

Solventextractionwasconvenienttorecoverresiduefromafragmentwhendirect

analysiswasnotsuitable.Liquidextractscanbefilteredand/orpre-concentrationasa

mechanism to improve detection. Common best practices for forensic sampling of

explosives residues recommend either acetone or methanol be used as the solvent.

Acetone, methanol (MeOH), water, acetonitrile (ACN), dichloromethane (DCM), and

hexanewere evaluated as solvents to extract post-blast residueswhileminimally co-

extractingenvironmentalcongeners.Onedrawbackofusingpolarorganicsolventsisthe

potential for co-extraction of other substances and contaminants from the

substrate/bombfragment.Ingeneral,organicsolventsunselectivelyextractanyorganic

substancespresentfromaspecimen.

Substrate ionization canoccurwith plastics 64 and textiles 65, 66whichwill also

contributeunwanted spectral background. Fragments representingdifferentmaterials

recoveredpost-blast(e.g.plastic,metal,rubber,etc.)werethussurveyedtodetermine

possible spectral contribution from the different substrates themselves. No explicit

variationwasobservedbetweendifferentfragmentstoindicateanyonesubstratewas

preferentially ionizedoveranother, indicatingthebackgroundionsobservedcouldnot

bedistinguishedfromsubstrateorenvironmentalcontaminationwhenrecoveredfrom

the ground. Fortunately, this is probative for our verification purposes, because real

specimens from post-blast debris are expected to be dirty from environmental

61

contaminationandunwanted substances areoftenobserved in thematrixof forensic

samples.TovisuallyenhancewherethetargetHMEsweredetectedinthetotalDART-MS

desorptionprofileobservedforeachfragment,extractedionchromatograms(EIC)were

found particularly useful to visually isolate regions of interest. High mass accuracy

permittedassuredreliableidentificationofallHME.

3.2.1.OrganicPeroxideExplosivesTATP and HMTD are the most common OPBE encountered in case work at

Canadianforensiclaboratories.MEKPislesscommon,butiseasilypreparedanalogousto

TATPorHMTDanditremainsaprioritytoensurestandardmethodscandetectMEKP

post-blast aswell. Via both direct and in-direct samplingmethods, DART-MS analysis

revealedtheOPBEusedfromatleastonefragmentrecoveredfromeachdevice(Table

4).Analysisofswabswasgenerallyfoundtobemoreeffectivethanbydirectanalysisof

fragments, except for MEKP, which proved the most challenging peroxide HME to

sufficientlydetectpost-blast(Table4).

62

Table4.DepictionoftheOPBEidentifiedviaDART-MSdirectanalysisofamultitudeofdifferentpost-blastfragmentscomparedtoin-direct.

DIRECT SWAB TATP HMTD MEKP TATP HMTD MEKPPlasticsheathedelectricalwire

Hardplasticsubstrate Cellphonebody

9Vbattery Dcellbattery

Cellphonebattery Washer

IEDcontainer Nail

Substratenotfoundpost-blast 3-6characteristicions 2characteristicions 0-1characteristicions

TATPwaspositivelyidentifiedastheexplosiveusedinIED#1and2viadirectand

in-directanalysisofthepost-blastfragments(Figure27and28).Identificationwasoffered

byobservingthefollowingionsidentifiedascharacteristicviaanalysisofTATPcertified

reference material (Figure 26); ammoniated TATP molecular ion (m/z 240.1436

[M+NH4]+),mono-acetoneperoxide(m/z74.0364[M/3]+)andprotonatedfragmentsfrom

theparenttrimerm/z75.4406[C3H6O2+H]+,89.0597[C4H9O2]+and91.0390[C3H6O3+

H]+. The base peak observed for TATP from analysis-to-analysis was either the m/z

91.0390 fragment or m/z 240.1436 [M+NH4]+. Characteristic ions for TATP were

consistently detected with an abundance of 105 – 107 counts and with spectral

backgroundionstypicallybetween103–104counts.TheamountofTATPusedinIED#1

versus#2wasdeliberatelychosenas it replicatedpastcaseworkscenariosexamined.

63

QuantitiesofTATPrangingfrom101–103gramscale(i.e.smallandlargescale)havebeen

encounteredincriminalinvestigations,wherelargeramountsweremanufacturedtobe

usedasdemolitioncharge(s)21andsmalleramountsforconcealmentinvictim-operated

IEDs.22ThesimilarabundanceofionsobservedforTATPdetectedfrombothIED#1and

#2isindicativethatsufficientresiduewasretainedoncertainfragmentsfromIED#2to

be detected in similar quantity to that recovered from IED #1. The maximum signal

responsefortheMSis108counts,whichindicatestheionsdetectedforTATPfromeither

device did not saturate the detector. Retaining sufficient TATP for detectionwas not

unexpected because all fragments were collected soon after detonation and well-

preserved in air-tight evidence collection bags to optimize the opportunity for HME

residuerecovery.

Figure26.Positivemode,fullscanhigh-resolutionmassspectrumforTATPanalyzedfromcertifiedreferencestandard.IonscharacteristicofTATPhavebeenboldedand

labelled.

64

Figure27.Positivemode,fullscanhigh-resolutionmassspectrumforTATPupondirectanalysisoffragmentfromIED#2.IonscharacteristicofTATPhavebeenboldedand

labelled.

Figure28.Positivemode,fullscanhigh-resolutionmassspectrumforTATPanalyzedfromaswabusedtocollectpost-blastresiduesfromIED#2fragments.Ions

characteristicofTATPhavebeenboldedandlabelled.

65

HMTD was identified as the main charge for IED #3 via observation of

characteristicionsidentifiedinanalysisofHMTDcertifiedreferencematerial(Figures29-

31);protonatedmolecularion(m/z209.0768[M+H]+)andcommonlyobservedfragments

9, 42m/z 88.0394 [M-C3H6NO4]+, 145.06081 [M-CH2O3 + H]+, 179.0663 [M-CH2O + H]+,

191.0664 [M-H-O]+ and224.0878 [M–H2+NH4]+. Themajor ionobserved forHMTD

analysis-to-analysiswaseitherthem/z145.06077or224.0878fragments.Characteristic

ionsforHMTDwereconsistentlydetectedwithanabsoluteabundancerangingfrom105

–106andwithbackground ions remainingbetween103–104 counts.Thequantityof

HMTDselectedforIED#3agreeswithpastcaseworkexamples,whereHMTDhasbeen

encounteredasfillerusedforhomemadedetonators.Anionwithm/z207.0979wasnot

reproduciblyobserved,consistentwithpreviousreports.9,63

Figure29.Fullscanhigh-resolutionmassspectrumforHMTDanalyzedfromcertifiedreference.IonscharacteristicofHMTDhavebeenboldedandlabelled.

66

Figure30.Fullscanhigh-resolutionmassspectrumdepictingidentificationofHMTDfromdirectanalysisofafragmentcollectedpost-blastfromIED#3.Ionscharacteristicof

HMTDhavebeenboldedandlabelled.

Figure31.Fullscanhigh-resolutionmassspectrumdepictingidentificationofHMTDuponanalysisofaswabusedtocollectpost-blastresiduesfromIED#3fragments.Ions

characteristicofHMTDhavebeenboldedandlabelled.

67

Whensynthesizedandleftunseparatedbychromatography,MEKPisamixtureof

monomeric, dimeric, trimeric (linear and cyclic) and higher oligomeric species. 39 For

realism,wepreparedMEKPandonlyisolatedfromunreactedreagents,butdidnotpurify

further before use. DART-MS analysis of our MEKP reference material yielded four

characteristic ions (Figure 32); ammoniated linear trimer (m/z 316.1965 [C12H26O8 +

NH4]+),ammoniatedlineartetramer(m/z404.2488[C16H34O10+NH4]+),andmonomeric

fragments m/z 77.0233 [C2H5O3]+ and 89.0597 [C4H9O2]+. Characteristic ions were

detectedwithanabsoluteabundanceof105–106andspectralbackgroundionsremained

between 103 – 104 counts. The ion observed at m/z 316.1965 is best calculated as

corresponding to a linear trimer adductwith ammonium [C12H26O8 +NH4]+ and not a

monomer adduct as previously reported. 43 Parent ions attributable to either the

monomerordimerspecieswerenotobserved.ThedesorptionprofileobservedforMEKP

byDART-MSwasbroadandtherelativeratiosofionsobservedchangedfrombeginning

toend. Initially, the lowermass fragmentsm/z77.0233and89.0597weredominantly

favouredatthebeginningofthescan,buttransitionedtopredominantlythehighermass

ions316.1965andm/z404.2488towardstheend.Thefragmentionsm/z77.0233and

89.0597arecommontoboththemonomeranddimer,suggestingapseudo-distillation

profileisoccurring,wherethelighterMEKPmonomeranddimerspeciesaredesorbed

and ionizedbefore theheavieroligomers, but theparent ionswere simplynot stable

enough to be observed.When the averageMS is plotted over all scans in the broad

desorptionprofile, themajor parent ionsobserved fromanalysis-to-analysiswere the

ammoniumadductofeitherthetrimer(m/z316.1965)ortetramer(m/z404.2488).We

68

didnotelucidatefurtherifthemonomerordimerparentionsmaybepresent,because

detectionofthefourionsobservedissufficientforidentificationofMEKPinapost-blast

residue.MEKPwasidentifiedasthemainchargeforIED#5asthesamecharacteristicions

identifiedinanalysisofthecrudeMEKPproductwereobservedinthespectraobtained

fromanalysisofdirectandin-directsamplingtechniques(Figure33and34).MEKPwas

themostchallengingperoxideHMEtodetectpost-blast(Table3).

Figure32.Positivemode,fullscanhigh-resolutionmassspectrumforMEKPanalyzedfromthecrudesynthesizedproduct.IonscharacteristicofMEKPhavebeenboldedand

labelled.

69

Figure33.Positivemode,fullscanhigh-resolutionmassspectrumdepicting

identificationofMEKPfromdirectanalysisoffragmentcollectedpost-blastfromIED#5.IonscharacteristicofMEKPhavebeenboldedandlabelled.

Figure34.Positivemode,fullscanhigh-resolutionmassspectrumdepicting

identificationofMEKPuponanalysisofaswabusedtocollectedpost-blastresiduesfromIED#5fragments.IonscharacteristicofMEKPhavebeenboldedandlabelled.

70

Both direct and in-direct analysis of the post-blast fragments permitted

identificationofeachOPBEused.In-directanalysisusingswabbingprovidedamechanism

to sample large or irregular objects not amenable to direct analysis. Swabbing also

enhancesrecoverybyaccumulatingmoreresidueonasmallersurfacearea.Therefore,

furtheroptimizationwascompletedtodeterminethemostsuitableswabsubstrateand

comparerecoverywithdryandsolvent-dampenedswabs.

Post-blast fragments from IED#3wereused for thisoptimizationwork. Swabs

madeofcotton,Teflonandpapermodifiedwithathinadhesivecoatingweretestedto

determinetheirrespectiverecoveryefficiencies(Figure35).Cottonswabsarereported

tobeaneffectiveresourceforcollectingsamplesof forensic interests. 65Collectionof

residues using cotton swabs resulted in positive identification of the explosive used

(HMTD) (Table 5) and cotton was therefore chosen as the preferred swab material.

Detectionofexplosivespresentonaswabviaanalysisintransmissionmoderequiresthe

active ionizinggas streamtopass through thematerialand interactwith the residues

present.Comparedtothemodifiedpaperswabs,cottonswabsaresufficientlyporousto

facilitatedesorptionandionizationforMSdetection.Usingcottonswabsisinagreement

with best practice for forensic sample collection as cotton swabs are regularly

recommendedtofront-linestaffmembersandinvestigatorsforsamplecollection

71

Table5.CharacteristicionsofHMTDpresentuponanalysisofresiduescollectedviaswabsdifferentiatedbysubstrate.

M–H2+NH4]

+224.0878

[M+H]+

209.07681[M-C3H6NO4]

+88.0394

[M-CH2O3+H]+

145.0608[M-CH2O+H]

+179.0663

Cotton Teflon

Modifiedpaper

Ionpresent Ionabsent

Figure35.Fromlefttoright-cottonswab,paperswab,modifiedpaperswab.

Analysis of dry versuswet swabswas conducted to evaluate any difference in

effectiverecoveryofresidueswhenusing(ornotusing)solvents.Analysisofdryswabs

collected from, afforded identificationofHMTD (top, Figure36).DART-MSanalysis of

swabswettedwithacetone,acetonitrileormethanolresultedinpositiveidentificationof

HMTDfromIED#5container(Figure36).Wettingaswabwithasolventpriortoswabbing

can enhance explosive residue recovery 9, but solvents will also co-dissolve other

substancespresentfromenvironmentalcontaminationorthesubstrateitself.Clean-up

proceduresforswabextractspriortoanalysiscanbeused.9Dryswabbing,however,will

72

not dissolve any substances or the substrate, reducing unwanted sample matrix and

fewer background ions to clutter the MS spectrum (Figure 36). Dry swabbing can

eliminateextractionorclean-up,andwasfoundeffectiveforexplosiveresiduerecovery.

Whenanalyzingdryandwetswabscollectedfromthepost-blastcontainerofIED#3,we

observed an additional predominant ionm/z 207.0614 (Figure 35) not apparent from

directanalysis(Figure29).Thisdaughterionisbestcalculatedas[C6H11N2O6]+whichisin

agreementwithstructuralassignmentasthedialdehydederivativeprotonatedinsteadof

ammoniated,i.e.[M–H2+H]+.Theappearanceofionm/z207.0614post-blastbutnot

apparent from the referencematerial can be associatedwith degradation from blast

effects(heat,pressure).Tothebestofourknowledge,ionm/z207.0614hasnotbeen

previouslyreportedfrompreviousLCMSstudies.10,62

73

Figure36.Fullscanhigh-resolutionmassspectradepictingidentificationofHMTDuponcollectionofpost-blastresiduesfromIED#5usingdryandsolventdampenedswabs.

74

3.2.2.BinaryExplosives Homemadebinaryexplosiveswereincludedinthisstudybecausetheindividual

components (i.e. fuel and oxidizer) can be easily sourced from common commercial

productsandposeasignificantthreatforfabricatingIEDs.Detectionofbinaryexplosives

products post-blast poses challenges because the different chemical nature between

fuels and oxidizers necessitates different analytical tools. Many of the commercial

products containing the precursor chemicals have complexmatrices arising from the

additionofdifferentstabilizers, fragrancesandpreservatives.UsingDART-MS,thefuel

and oxidizer can be detected separately by switching between positive or negative

polarity.ExamplesofcommoncommercialsourcesofHMEprecursorfuelsandoxidizers

arelistedinTable1.

Currently,theRCMPdoesnothaveavalidmethodforcharacterizationofsugar-

basedfuelspost-blast.TheuseofDART-MStodetectsugar-basedfuelswasthusexplored

asapotentialsolutiontothiscapabilitygap.Avarietyofdifferentcommerciallyavailable

carbohydrate-based or sugar-containing products are suitable as fuel sources,making

identificationofanysourcechallenging.ThemainchargeofIEDs#11and12wasbinary

explosive prepared using a carbohydrate-based fuels (TANG ® juicemix and dextrin).

TANG® isa food-gradesourceof sucrose,which is themain ingredient.Sucrose is the

dimerglucoseandfructosemonomerslinkedviaa(1,2)glycosidiclinkages.Dextrinisa

complexcarbohydratepolymercomposedofhighlycross-linkedglucoseunits.DART-MS

analysisofsucrose,TANGanddextrinindilutesolutionsverifiedeachcarbohydratefuel

was identifiable (Figures38,39and43).Becauseglucose isabasemonomer ineither

carbohydrate,analysisofeithersucroseordextrinresultedinasimilarMSspectrumand

75

fragmentation pattern, including common ions that can be used to identify each fuel

whencomparedtoaglucosereferencematerial.61DART-MSanalysisofglucoseyields

the following characteristic ions (Figure 37); ammoniated glucose ion (m/z 198.0973

[C6H12O6+NH4]+)andfragmentionsviathelossofwater(m/z180.0867[C6H12O6+NH4–

H2O]+,m/z163.0601[C6H12O6+NH4–NH3–H2O]+,m/z145.0496[C6H12O6+NH4–NH3–

2H2O]+andm/z 123.0390 [C6H12O6+ NH4 – NH3 – 3H2O]+). 60 Characteristic ionswere

detectedwithanabsoluteabundanceof105–106andspectralbackgroundionsremained

between103–104counts.

Ablendedproductofasugar-derivedfuelandastrongoxidizerisanexampleof

high-brisanceexplosive.Consequently,manyoftheIEDfragmentsrecoveredpost-blast

weresmall(e.g.SIMcard~0.25cm2)renderingrecoveryofresiduefordetectionbyDART-

MSdifficult.Co-extractionofmultiplefragmentsofsimilarmaterialstogetherafforded

detectionofacarbohydratebasedfuelinIED#11and12viaDART-MSanalysis(Figures

40-42and44-46).Thisisacommonsamplingmethodusedintheforensicanalysisofpost-

blast explosives. Similar DART-MS spectrum of both sucrose and dextrin prevents

identificationofthecommercialproductsourcepost-blast.Becausemostcarbohydrate-

basedfuelsarenon-volatileresiduelossbyevaporationisnotaconcern.Sugar-derived

fuelresiduesareexpectedtopersistandbedetectableifanalyzedatalatertime,which

isanalogouswiththeexpectationfortheinorganicresiduesoftheoxidizersaltsusedin

thesamebinaryexplosivemixtures.

76

Figure37.Fullscanhigh-resolutionmassspectrumofglucosedissolvedinwaterasareferencematerial.Collectedinpositivemode.

Figure38.Fullscanhigh-resolutionmassspectrumofsucrosedissolvedinwater,usedasareferencematerial.Collectedinpositivemode.

77

Figure39.Fullscanhigh-resolutionmassspectrumofTANGdissolvedinwater,usedasareferencematerial.Collectedinpositivemode.

Figure40.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11metal

substratefragment.Collectedinpositivemode.

78

Figure41.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11plastic

substratefragment.Collectedinpositivemode.

Figure42.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#11rubber

substratefragment.Collectedinpositivemode.

79

Figure43.Fullscanhigh-resolutionmassspectrumofdextrinreferencematerialdissolvedinwater.Collectedinpositivemode.

Figure44.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12plastic

substratefragment.Collectedinpositivemode.

80

Figure45.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12metal

substratefragment.Collectedinpositivemode.

Figure46.Fullscanhigh-resolutionmassspectrumdepictingidentificationofaglucosecontainingproductfrompost-blastresiduesextractedwithwaterfromIED#12rubber

substratefragment.Collectedinpositivemode.

81

Currently,GC-MSisthequalityassuredmethodusedbytheRCMPfortheanalysis

ofpetroleum-derivedfuelsinbinaryexplosives.However,DART-MSprovidesasuitable

andrapidalternative.Avarietyofcommonpetroleum-derivedproductsencounteredin

binaryexplosivesweresurveyed(Table1)andeachfuelexhibitsadifferent,distinctive

profile(Figures47,60-63).Itwasbeyondthescopeofthisworktoanalyzeallthedifferent

typesofpetroleum-derivedfuelsthatcouldpotentiallybeusedtomakeabinaryHME.

Eachfuelanalyzedwasdilutedinhexane,whichisasolventcommonlyusedtoextract

petroleum-derived fuels from explosives. Analysis of neat hexane assisted in the

background subtraction of any solvent associated ions (Figure 25). Identification of

automotivegreasepost-blastwasinvestigatedbyanalyzingfragmentsfromIED#10.

Automotive greases are manufactured by blending many different substances

together (e.g. lubricants, thickener, additives,preservatives),which togetheraffordsa

complexmixture.Analysisofatypicalautomotivelubricatinggrease(Figure47)exhibited

ahigh-abundancepatternof lowmass ions in them/z rangeof100-250anda lower-

abundancehydrocarbonprofileinthem/zrangeof300-450.Analysisofhexaneextracts

from the post-blast fragments (e.g. metal, plastic and rubber substrate fragments)

resulted in observation of three predominant ions (m/z 149.0237, 279.1589, and

391.2844) (Figure 48-50). EICs depicted observation of these ions in the DART-MS

spectrumoftheunusedautomotivegrease,whichwasanalyzedasareferencematerial.

However,theionsobservedpost-blastwerenotasabundantlyobservedfortheunused

grease.Thisisnotunexpected,astheblasteffects(heat,pressure)areknowntochange

thecompositionofpetroleum-derivedfuels,asiscommonlyobservedinignitableliquid

82

analysis.63Thebestpracticetoidentifythecombustionresidueofapetroleum-derived

fuelistocomparewithasuitablyweatheredorignitedreferencematerial;thescopeof

whichwasbeyondthispresentstudy.Regardless,thedetectionofionspost-blastthatare

identifiablycommonwiththeunusedgreaseindicatesthatapetroleum-derivedgrease

fuelcanbeidentifiedpost-blast.

Combustion is expected to consume some (or all) of the components of the

grease. Pyrolysis and other irreversible oxidation side-reactions are also expected to

occurduringtheexplosion.66Consequently,aweatheredsampleofgreaseisexpectedto

affordamorerepresentativematerialforcomparisonandidentificationofagreasepost-

blast,whichisthecommontechniqueusedinfire-debrisinterpretationofignitableliquids

after an arson fire. 66 Weathering grease samples is not a straightforward task. The

preparation of suitably weathered petroleum-derived fuels as reference samples for

comparisonwithresidueanalysisoffuel-productsusedindevices#6-10byDART-MSwill

bethesubjectoffuturework.

83

Figure47Fullscanhigh-resolutionmassspectrumofanautomotivegrease,usedasareferencematerial.Collectedinpositivemode.

Figure48.Fullscanhigh-resolutionmassspectrumdepictingidentificationofautomotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10metal

substratefragment.Collectedinpositivemode.

84

Figure49.Fullscanhigh-resolutionmassspectrumdepictingidentificationof

automotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10plasticsubstratefragment.Collectedinpositivemode.

Figure50.Fullscanhigh-resolutionmassspectrumdepictingidentificationof

automotivegreasefrompost-blastresiduesextractedwithhexanefromIED#10rubbersubstratefragment.Collectedinpositivemode.

85

Identificationoftheoxidizercomponentofthebinaryexplosivesmixturesusing

DART-MSwas also studied. Table 1 lists commonexamples of inorganic salts used as

oxidizersintheproductionofhomemadebinaryexplosives.DART-MSanalysisofthese

exemplarycommercialsourcesofoxidizerswasperformedinnegativemode.Instrument

andsolventblankswereincludedtomonitorcleanlinessoftheinstrumentandtoidentify

background ions endogenous to the DART itself (Figure 51 and 52). Analysis of the

commercialsourcesofoxidizerswasincludedtodeterminedetectioncapabilityforthe

realsubstancesthatareusedtomakeHME.Ammoniumnitratesourcedfrominstantcold

compressionpacks,wasidentifiedbytheobservingthenitrateion(m/z61.9867[NO3]-)

anditsadductwithnitricacid(m/z124.9819[HNO3+NO3]-).Potassiumnitratesourced

from commercial stump remover, was identified by observing the nitrate ion (m/z

61.9867[NO3]-).Anotherionwasalsoreproduciblyobservedatm/z121.9819([C7H5O2]-)

whichisattributabletoanorganicadditivepresentintheproduct.Detectionofnitrate

inorganic salts was in agreement with expectation because of their sufficient vapour

pressure at room temperature. 12 Upon analysis of aqueous extracts collected from

fragmentsrecoveredforIED#10post-blast,thesamecharacteristicionsobservedwere

observedidentifyingtheoxidizercomponentofthebinaryHME.Analysisofthefragments

fromtheremainingdevices(#7,8and12)willbecompletedinfuturework.

86

AnalysisofperchloratesandchloratesviaDART-MSprovedtobechallenging,in

agreement with other reported studies. 67 The low vapour pressure of chlorates and

perchlorates has been cited as the cause for poor ability to detect, which poses a

challenge to thermal desorption required for the DART ionization process. 67 Upon

analysisofchlorateandperchloratesaltswith theDARTprobe temperatureof250°C,

whichisstandardforanalysisofexplosives,detectionofperchloratesandchlorateswas

notobserved.NoionscharacteristicofperchlorateorchloratewereobservedinanyEICs

generatedfromanalysisofthebinaryexplosiveresidues.Increasingthetemperatureof

the DART ionizing gas has been reported as ineffective to recover either chlorate or

perchlorate.67FurthermethodoptimizationtouseDART-MStodetecteitherchlorateor

perchloratesaltsfrompost-blastbinaryexplosiveresidueswasnotperformedwithinthe

scopeofthisstudypresented,butwillbeexploredwithfuturework.

Figure51.Negativemode,fullscanhighresolutionmassspectrumuponoperationof

theDART-MS,depictingtheendogenousions.Totalioncount106.

87

Figure52.AnalysisofwaterdepositedontoaQuickStrip,innegativemodeusingfull

scan.Totalioncount106.

Figure53.Negativemode,fullscanhigh-resolutionmassspectrumforammonium

nitrateanalyzedasareferencematerial.

88

Figure54.Negativemode,fullscanhigh-resolutionmassspectrumforacommercially

availablestumpremover(commercialsourceofKNO3),analyzedasareferencematerial.

Figure55.IdentificationofKNO3innegativemodeviafullscanhighresolutionDART-MSanalysisofpost-blastresiduesextractedfrommetalsubstratefragmentsfromIED#10

withwater.

89

Figure56.IdentificationofKNO3innegativemodeviahigh-resolutionDART-MSanalysisofpost-blastresiduesextractedfromplasticsubstratefragmentsfromIED#10with

water.

Figure57.IdentificationofKNO3innegativemodeviafullscanhigh-resolutionDART-MSanalysisofpost-blastresiduesextractedfromrubbersubstratefragmentsfromIED#10

withwater.

90

3.2.3.SmokelessPowders

Asblasteffects(heatandpressure)canchangethecompositionofacompound,

ignitionofNCisexpectedtoresultinthermalbreakdowntooligomersofvaryinglength

and degree of nitration. Characterization of these thermal breakdown products is

challengingasreferencematerialsdonotexist.Commercialsourcesofmonomerordimer

nitratedsugarproductsarenotavailablemotivatingthesynthesisofareferencematerial

in-house.Successfulsynthesisofb-cellobioseoctanitrate(ONCB)affordedareference

material structurally similar to potential breakdown products from the thermal

degradationofNC.

Initial analysis of ONCB and unconsumed grains of single and double based

smokelesspowderresultedindissimilarspectra.Ionscharacteristicoftheadditivesand

preservatives included in smokelesspowder (e.g.DPAandEC)dominated the spectra

obtained upon analysis of both single and double base products unconsumed.

PredominantionsobserveduponanalysisofONCBwereabsentinthespectraobtained

viaanalysisofthesmokelesspowder.ComparisonofONCBandresiduescollectedfrom

the ignited smokeless powder also did not warrant similar spectra. New ions were

observed in the spectra from analysis of the ignited smokeless powders but were

dissimilarfromthepredominantionsintheONCB.TheseresultssuggestedthattheONCB

inpureformmaynotbeacomparablematerialforthecharacterizationofNCthermal

breakdownproducts.

91

ComparisonofignitedONCBandsmokelesspowderwascompletedtodetermine

if the ignited product was a more suitable reference material than unburned ONCB.

Analysis of extracted ONCB and smokeless powder residues with acetonitrile was

completed. After subtracting the spectral contributions from acetonitrile (Figure 23),

similar ions in the m/z range of 300-500 were observed upon analysis of ONCB and

smokelesspowderresidues.Proposedcharacteristicionsareasfollows;m/z323.0718,

341.0822,491.0744,508.0643,and536.0597.Ionabundancewashighlyvariableacross

analyses of all samples. The highly complex sample matrix created by burning the

productsresultedinmassspectrawithionsobservedfromm/z50–500atcountsof106

-107.Subtractionofspectralcontributionfromthesolventreducedioncountsbyonlya

signalmagnitude (107 to 106) further emphasizing complexity of the ignited samples.

Figure58depictshowEICswereusedtoconfirmthattheproposedcharacteristicionsare

notattributedtoanysamplecomponentotherthannitratedsugarthermalbreakdown

products. The ions used to produce the EICs are indicated on the left of the figure.

DescriptorsofthesamplesplacedineachwellonaQuickstriparelabelledalongthetop.

ThelongblackboxwithgreycirclesoverlaidontheEICdepictstheQuickstripcardthat

blank, control and extract sampleswere deposited onto. The firstwell contained the

extraction solvent; included to facilitate background subtractions. Blank wells were

analyzedbetweeneachsampletofacilitatesubtractionofionsendogenoustotheDART-

MSandmonitorinstrumentcleanliness.AnalysisofTNTservedasanegativecontrolas

the thermal breakdown products would not be present in the TNT sample and their

structurewouldnotbesimilartothatofTNT.Theignitedresiduesweredepositedonto

92

wellsfive,sevenandnine.ObservationofthecharacteristicionsintheEICforallthree

samples confirms detection and identification of similar thermal breakdown products

betweenignitedONCB,SBandDBsmokelesspowders.Absenceofthecharacteristicions

in the wells associated with the instrument background, solvent, and TNT further

confirmsassociationtothenitratedsugarthermalbreakdownproducts.Characterization

of the thermal breakdown products is supported by identification of common

characteristic ions in both the reference sample (ignited ONCB) and the consumed

smokelesspowders.

Figure58.Determinationofionscharacteristicofthethermalbreakdownproducts.

93

Mass analysis of each ion was used for structure determination based on

acceptable mass shift (± 0.003 amu) with respect to high-resolution accurate mass

measurements (Table 6). MS/MS analysis may provide further clarification on the

structureofthethermalbreakdownproductsbyproducingfragmentsthatcanbeused

toelucidatefunctionalgroupsandstereochemistry,andshallbeexploredinfuturework.

Table6.Listofmassformulaefortheionscharacteristicofnitratedsugarthermalbreakdownproducts,withassociatedmassshift(amu).

m/z MassFormulae Massshift(amu)

323.0718C9H9O5N9 -0.486

C10H15O10N2 -0.491

341.0822C9H11O6N9 -0.611

C10H17O11N2 -0.616

491.0744C13H15O13N8 -1.129

C26H11O7N4 -0.616

508.0643C12H14O14N9 -1.112

C26H12O8N4 -0.605

536.0597C13H14O15N4 -0.907

C27H12O9N4 -0.399

Solubility of the thermal breakdown products was explored using a series of

solvents commonly used in accepted forensic practices (acetone, MeOH, water,

acetonitrile,DCMandhexane).ChangesinobservationofionsintheEICswasusedasa

measurementofsolubility.Thethermalbreakdownproductsareassumedtoberelatively

polar if they retain thepolarnitrategroups.Solventsofvaryingpolaritywereused to

confirm this hypothesis. Figure 59 depicts a combination of the EICs obtained from

analysis of ONCB, SB and DB residues extracted with six different solvents. Wells

highlighted by the red box indicates samples extracted with hexane. Absence of all

94

characteristic ions upon analysis of hexane extract confirms the thermal breakdown

products are not soluble in a non-polar solvent. Wells highlighted in the purple box

indicatesamplesextractedusingDCM.Characteristicionswereobserveduponanalysis

of ignited ONCB but not smokeless powders confirming many species with slightly

differentpolaritycouldbepresent.Yellow,greenandorangeboxesindicateextraction

withacetone,ACNandMeOH,respectively.Asthesearerelativelypolarsolvents,with

retentionofthenitrategroupsuponthermaldegradation,itisnotsurprisingthatresidues

were soluble in these polar solvents. The blue box depicts extractionwithwater and

absenceofcharacteristicionsconfirmstheyarenotsolubleinwater.

Figure59.Relativesolubilityofresiduescontainingthermalbreakdownproducts.

95

SynthesisofareferencematerialsuchasONCBwasessentialfordetectionand

identificationofNCthermalbreakdownproducts.IgnitionofONCBaffordedproduction

of residuesalike thoseproducedupon the ignitionof a smokelesspowder. Successful

detection and identification of similar ions upon analysis of ignited products provides

mechanisms for characterization of thermal breakdown products. Post-blast

identification of NC would remain impossible without detection and identification of

characteristicthermalbreakdownproducts.

FuturegoalsincludeusingtheignitedONCBasareferencematerialandDART-MS

to fully characterize smokeless powders from post-blast residues when used in IED.

Therefore, it is imperative to test applicability of this novelworkwith real post-blast

fragments.Thesetestswillbecarriedoutinfutureworktoensurethatcharacteristicions

observeduponburningtheproductsareidentifieduponanalysisofpost-blastresidues.

Toensureefficientextractionpost-blastresiduesshouldbeextractedwithACN.

96

4.ConclusionDARTcoupledwithhigh-resolutionMSwasdemonstratedasasuitableanalytical

techniqueapplicableforforensicidentificationofhomemadeexplosivefrompost-blast

residuesrecoveredfromIEDfragmentsrepresentativeofitemsandsubstratescommonly

recoveredfromgenuinebombinginvestigations.Explosivesofintereststudiedincluded

organicperoxides (triacetonetriperoxide,hexamethylenetriperoxideandmethylethyl

ketone peroxide), binary explosives (fuel and oxidizer mixtures) and commercial

smokelesspowders(bothsingle-anddouble-base).Eachexplosivewascharacterizedby

comparing questioned mass spectra with known reference materials to identify ions

characteristic of the target explosive. DART-MS was also successfully used to

characterization the thermal breakdownproducts of nitrocellulose (i.e. nitrated sugar

derivatives)whichcanbeusedtoidentifyasmokelesspowderpost-blastintheabsence

ofrecoveringanintactgrain.

DART-MSwasverifiedtoidentifyHMEresiduesbydirectanalysisofthepost-blast

IEDfragmentsthemselvesandbyindirectanalysisofsub-samplescollectedusingdryor

wetswabswithsolvents.Drycottonswabswere foundeffective forexplosive residue

recovery and yielded the least co-extracted environmental background. Since dry

swabbing is also a commonplace sampling techniqueused at security check-points to

detect concealed explosives or residues,we suggestDART-MS is compatible for rapid

forensicdetectionoftraceexplosivesinscreeningapplications.

97

5.FutureWork The residues studied in this project were not subjected to weathering or

degradationeffectswhich are commonlyobserved in forensic samples collected from

crime scenes as this was beyond the scope of the study. Post-blast fragments were

collected andpackaged in sealable nylon-evidence bags immediately after detonation

unlike realistic scenarios where many safety precautions are taken before front-line

memberscanenterablastscenetocollectevidence.Exposuretoenvironmentalfactors,

suchasprecipitation,temperatureschangesorsunlight(e.g.UV)canaffectthelifetime

and recoverability of explosive residues post-blast. The potential for sample loss or

contaminationofevidence(e.g.sourcedbyexposuretoweather)couldbestudiedinthe

futuretofurtherreplicaterealisticscenarios.Delayedcollectionofthefragmentsintime

increments representative of realistic scenarios is oneway to study exposure effects.

Simulatingchangesintemperaturebyexposingthefragmentstoheat(e.g.placinginan

oven with temperature set to represent realistic Canadian summer heat) is another

mechanisminwhichmeasuringtheeffectsofweatheringcouldbestudied.

Furthermore,duetothehomemadenatureofOPBEandbinarymixturesdifferent

precursor sources and synthesis methods can contribute to changes in the residues

collectedpost-blast.Additionalanalysisofpost-blastresiduesfromdetonationofdevices

that containproducts synthesizedwithdifferentprecursors (i.e. lowerwt%hydrogen

peroxide,differentacidcatalysts,andadditionalcommercialfuelandoxidizersources)

wouldfurtherdepicttherobustcapabilityofDART-MSasan identificationmethodfor

homemadeexplosives.

98

References[1] TheGlobalPracticeofForensicScience,editedbyDouglasH.Ubelaker,John

Wiley&Sons,Incorporated,2014.

[2] R.W.Byard,H.James,J.Berketa,K.Heath.JForensicSci(2016)61:545-547.

[3] N.Abdul-Karim,C.S.Blackman,P.P.Gill,E.M.M.Wingstedt,B.A.P.Reif.RSCAdv.

4(2014)54354–54371.

[4] J.S.Caygill,F.Davis,SP.J.Higson.Talanta(2012)88:14-29.

[5] EicemanGA,KarpasZ.Ionmobilityspectrometry,2ndrev.edn.BocaRotan,FL:

TaylorandFrancis,2005.

[6] G.W.Cook,P.T.LaPuma,G.L.Hook,B.A.Eckenrode.JournalofForensicScience

(2010)55:1582-1591.

[7] C.L.Crawford,H.H.Hill,Jr.AnalyticaChimicaActa(2013)795:36-43.

[8] Beveridge,ForensicInvestigationsofExplosions,CRCPressTaylor&Francis

Group,India,2012.

[9] D.DeTata,P.Collins,A.McKinley.ForensicScienceInternational(2013)233:63-

74.

[10] X.Xu,M.Koeberg,C.-J.Kuijpers,E.Kok.ScienceandJustice(2014)54:3-21.

[11] T.P.ForbesandE.Sisco.Analyst(2018)143:1948-1969.

[12] J.Pavlov,A.B.Attygalle.Anal.Chem.(2013)85:278−282.

[13] E.Sisco,J.Dake,C.Bridge.For.Sci.Int.(2013)232:160–168.

[14] T.P.Forbes,E.Sisco,M.Staymates,G.Gillen.Anal.Methods(2017)9:4988-

4996.

[15] Kubota,PropellantsandExplosives:ThermochemicalAspectsofCombustion,2nd

ed.,Wiley–VCH,Germany,2007.

[16] J.C.Oxley,J.L.Smith,J.Huang,W.Luo.J.ForensicSci(2009)54:1029–1033.

[17] L.R.RothsteinandR.Petersen.PropellantsandExplosives(1979)4:50-60.

[18] M.VaullerinandA.Espagnacq.Proppelants,ExplosivesandPyrotechnics.(1998)

23:237-239.

[19] N.Hagan,I.Goldberg,A.Graichen,A.St.Jean,C.Wu,D.Lawrence,P.Demirev.J.

Am.Soc.MassSpectrom.(2017)28:1531-1539.

99

[20] R.D.Bach,P.Y.Ayala,H.B.Schlegel.J.Am.Chem.Soc.(1996)118:12758-12765.

[21] Agrawal,J.P.;Hodgson,R.D.OrganicChemistryofExplosives(2007).

[22] J.Pachman,R.Matyas.ForensicScienceInternational(2011)207:212–214.

[23] R.vs.BRIDGES,Dane.OntarioCourtofJustice;Ottawa:2012-2013.

[24] R.vs.AMSEL,Guido.ProvincialCourtofManitoba;Winnipeg:2017–2018.

[25] F.SRomolo,L.Cassioli,S.Grossi,G.Cinelli,M.V.Russo.ForensicSci.Int.(2013)

224:96–100.

[26] N.A.Milas,A.Golubovic.J.A.C.S.(1959)81:5824-5826.

[27] A.Miyake,K.Takahara,T.Ogawa,Y.Ogata,Y.WadaandH.Arai.Jour.ofLoss

PreventioninthePro.Ind.(2001)14:533-538.

[28] R.M.Heramb,B.R.McCord.For.Sci.Comms.(2002)4.

[29] R.Gonzalez-Mendez,C.A.Mayhew.J.Am.Soc.MassSpectrom.(2017)30:1531-

1539.

[30] C.Christodoulatos,T.Su,A.Koutsospyros.WaterEnvironRes.(2001)27:185-

191

[31] E.Goudsmits,G.Sharples,J.W.Birkett.TrACTrendsAnal.Chem.(2015)74:46–

57

[32] R.V.Taudte,A.Beavis,L.Blanes,N.Cole,P.Doble,C.Roux.Biomed.Res.Int.

(2014)16:16.

[33] M.Joshi,K.Rigsby,J.R.Almirall.ForensicSci.Int.(2011)208:29–36

[34] M.Mach,A.Pallos,P.Jones.J.ForensicSci.(1978)23:433–445

[35] A.Zeichner,B.Eldar,B.Glattstein,A.Koffman,T.Tamiri,D.MullerJ.Forensic

Sci.(2003)48:961–972

[36] R.Ewing.Talanta54(2001)515–529.

[37] Beveridge,ForensicInvestigationsofExplosions,CRCPressTaylor&Francis

Group,India,2012.

[38] O.Coskun,NorthClinIstanb.(2016)3:156–160.

[39] D.S.Moore,J.V.Goodpaster.AnalBioanalChem(2009)395:245–246

[40] BarronandE.Gilchrist,AnalyticaChimicaActa(2014)806:27-54.

[41] E.Gilchrist,N.Smith,L.Barron,Analyst(2012)137.

100

[42] J.M.Nilles,T.R.Connell,S.T.Stokes.PropellantsExplos.Pyrotech.(2010)35:

446-451.

[43] K.Clemons,J.Dake,E.Sisco,G.F.VerbecIV.ForensicSci.Int.(2013)231:98-101.

[44] L.Frederick,T.Joseph,B.D.Muselman,A.B.ScienceandJustice(2016)56:321-

328.

[45] F.Rowell,J.Seviour,A.Y.Lim,C.G.Elumbaring-Salazar,J.Loke,J.Ma.Forensic

SciInt(2012)221:84-91.

[46] J.R.Swider.JournalofForensicSciences(2013)58:1601-1606.

[47] E.Sisco,J.Dake,C.Bridge.ForensicSci.Int.(2013)232:160-168.

[48] F.Rowell,J.Seviour,A.Y.Lim,C.G.Elumbaring-Salazar,J.Loke,J.Ma.Forensic

Sci.Int.(2012)221:84-91.

[49] Dong,DirectAnalysisinRealTimeMassSpectrometry,Wiley–VCH,Germany,

2018.

[50] D.C.Harrisetal,in“QuantitativeChemicalAnalysis.”7thedition.W.H.Freeman

andCompany.NY.USA.2007.

[51] M.Huang,C.Yuan,S.Cheng,Y.Cho,J.Shiea.Annu.Rev.Anal.Chem.(2010)

3:43–65.

[52] A.G.MarshallandC.L.Hendrickson.Annu.Rev.Anal.Chem.(2008)1:579–99.

[53] Cody,R.B.;Laramée,J.;Nilles,J.M.Anal.Chem.(2005)77:2297-2302.

[54] E.S.Chernetsova,G.E.Morlock.MassSpectrometryReviews(2011)30:875-

883.

[55] J.M.Nilles,T.R.Connell,H.D.Durst.AnalChem(2009)81:6744-6749.

[56] Michalski,A.etal.Mol.Cell.Proteomics(2011)9.

[57] Brunnee,C.Int.J.MassSpectrom.IonProc(1987)76:125-237.

[58] J.V.Olsen,B.Macek,O.Lange,A.Makarov,S,Horning,M,Mann.Nature

Methods(2007)4:709-712.

[59] Hardman,M.,andMakarov,A.A.Anal.Chem.(2003)75:1699–1705.

[60] J.C.Oxley,J.L.Smith,K.Shinde,J.Moran.Propellants,Explos.,Pyrotech.(2005)

30:127-130.

[61] L.E.DeGreet,M.M.Cerreta,C.J.Katille.ForensicChem.(2017)4:41-50.

101

[62] T.A.Ong,T.Mendum,G.Geurtsen,J.Kelley,A.Ostrinskaya,R.Kunz.Anal.Chem.

(2017)89:6482-6490.

[63] N.Abdul-Karim,R.Morgan,R.Binions,T.Temple,K.Harrison.J.ForensicSci.

(2013)58:365-371.

[64] D.S.Saang’onyo,D.L.Smith.RapidCommun.MassSpectrom.(2012)26:385–

391.

[65] R.J.Brownlow,K.E.Dagnall,C.E.Ames.J.ForensicSci.(2012)57:713-717.

[66] E.Stauffer,J.DolanandR.Newman.FireDebrisAnalysis,Elsevier,2008.

[67] T.P.Forbes,E.Sisco,M.Staymates.Anal.Chem.(2018)90:6419-6425.

102

Appendix1:SupplementaryInformation

Figure60.Positivemode,fullscan:high-resolutionmassspectrafordiesel,analyzedasareferencematerial

Figure61.Positivemode,fullscan:high-resolutionmassspectraforlampoil,analyzed

asareferencematerial

103

Figure62.Positivemode,fullscan:high-resolutionmassspectraforVaseline,analyzedasareferencematerial

Figure63.Positivemode,fullscan:high-resolutionmassspectraforwax,analyzedasareferencematerial