Science and Justice 52 (2012) 217–225

Contents lists available at SciVerse ScienceDirect

Science and Justice

j ourna l homepage: www.e lsev ie r .com/ locate /sc i jus

Quality assurance testing of an explosives trace analysis laboratory — Furtherimprovements to include peroxide explosives☆,☆☆

Andrew Crowson a,⁎, Richard Cawthorne b

a Trace Laboratory Manager, Forensic Explosives Laboratory, Defence Science and Technology Laboratory, Fort Halstead, Sevenoaks, Kent, TN14 7BP, UKb Laboratory and Analytical Services, Forensic Explosives Laboratory, Defence Science and Technology Laboratory, Fort Halstead, Sevenoaks, Kent, TN14 7BP, UK

☆ © Crown copyright 2012 published with the permand Technology Laboratory on behalf of the Controller o☆☆ Disclaimer: any views expressed are those of therepresent those of the Department/HM Government.

⁎ Corresponding author. Tel.: +44 1959 892046; faE-mail address: acrowson@dstl.gov.uk (A. Crowso

1355-0306/$ – see front matter. Crown Copyright © 20http://dx.doi.org/10.1016/j.scijus.2012.07.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 2 March 2012Received in revised form 17 July 2012Accepted 24 July 2012

Keywords:Quality assuranceExplosivesTrace analysisGas chromatographyMass spectrometryPeroxide

The Forensic Explosives Laboratory (FEL) operates within the Defence Science and Technology Laboratory(DSTL) which is part of the UK Government Ministry of Defence (MOD). The FEL provides support and adviceto the Home Office and UK police forces on matters relating to the criminal misuse of explosives. During 1989the FEL established a weekly quality assurance testing regime in its explosives trace analysis laboratory. Thepurpose of the regime is to prevent the accumulation of explosives traces within the laboratory at levelsthat could, if other precautions failed, result in the contamination of samples and controls. Designatedareas within the laboratory are swabbed using cotton wool swabs moistened with ethanol:water mixture,in equal amounts. The swabs are then extracted, cleaned up and analysed using Gas Chromatography withThermal Energy Analyser detectors or Liquid Chromatography with triple quadrupole Mass Spectrometry.This paper follows on from two previous published papers which described the regime and summarisedresults from approximately 14 years of tests. This paper presents results from the subsequent 7 years settingthem within the context of previous results. It also discusses further improvements made to the systems andprocedures and the inclusion of quality assurance sampling for the peroxide explosives TATP and HMTD.Monitoring samples taken from surfaces within the trace laboratories and trace vehicle examination bayhave, with few exceptions, revealed only low levels of contamination, predominantly of RDX. Analysis ofthe control swabs, processed alongside the monitoring swabs, has demonstrated that in this environmentthe risk of forensic sample contamination, assuming all the relevant anti-contamination procedures havebeen followed, is so small that it is considered to be negligible. The monitoring regime has also been valuablein assessing the process of continuous improvement, allowing sources of contamination transfer into thetrace areas to be identified and eliminated.Crown Copyright © 2012 Published by Elsevier Ireland Ltd. on behalf of Forensic Science Society. All rights

reserved.

1. Introduction

In 1989 the Forensic Explosives Laboratory (FEL) adopted Gas Chro-matography with a Thermal Energy Analyser (GC/TEA) as its principaltechnique for analysis of organic high explosives traces [1,2]. The origi-nal systemsweremodified by FEL to improve their performance, includ-ing their sensitivity. Where any forensic trace analysis is carried out, arigorous system of contamination prevention procedures is essential.For this reason the instruments are located within a purpose builtsuite of laboratories referred to as the Trace Explosives Laboratory, inwhich forensic casework samples are processed and analysed in orderto detect any explosives present. The GC/TEA systems are routinely

ission of the Defence Sciencef HMSO.authors and do not necessarily

x: +44 1959 892656.n).

12 Published by Elsevier Ireland Ltd

used to analyse for the presence of twelve compounds of explosivessignificance: ethylene glycol dinitrate (EGDN), nitrobenzene (NB), 2-nitrotoluene (2-NT), 3-nitrotoluene (3-NT), 4-nitrotoluene (4-NT),1,2,3-propanetriol trinitrate (nitro-glycerine) (NG), 2,6-dinitrotoluene(2,6-DNT), 2,4-dinitrotoluene (2,4-DNT), 3,4-dinitrotoluene (3,4-DNT),2,4,6-trinitrotoluene (TNT) 2,2‐bis(dihydroxymethyl)-1,3-propanedioltetranitrate (pentaerythritol tetranitrate) (PETN) and 1,3,5-trinitro-1,3,5-triazacyclohexane (cyclotrimethylene trinitramine) (RDX). InDecember 2000 FEL began to use Liquid Chromatography with sin-gle quadrupole Mass Spectrometry to analyse the peroxide explo-sives Hexamethylenetriperoxidediamine (HMTD) and TriacetoneTriperoxide (TATP) [3,4]. Eventually this led to the routine analysisof these compounds using Liquid Chromatography with triple quad-rupole Mass Spectrometry. This paper will outline the contaminationprevention procedures currently in use, present the data from theanalysis of monitoring samples between February 2004 and July2011 and set them in the context of previous results. It also presentsresults produced by extending the quality assurance procedures toinclude the peroxide explosives HMTD and TATP. Since FEL operates

. on behalf of Forensic Science Society. All rights reserved.

218 A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

a policy of continuous improvement, there have been many amend-ments to the procedures and alterations to the facilities over theperiod, some of which are outlined. This paper should be read in con-junction with the previous papers [5,6] that present the results ofsamples taken between November 1989 and January 2004.

2. Main trace laboratory contamination prevention procedures

A detailed discussion of explosives trace contamination controlprinciples and practice has been given by Doyle [7]. Additionally,brief descriptions explaining the main principles of the FEL preven-tion procedures were included in the previous papers [5,6].

Before entering the trace laboratory, there are several “external”precautions. The laboratory is held at positive air pressure to preventthe ingress of particulate contamination via any possible ingresspoints, including the main door. All personnel entering the laboratorymust have showered and had a complete change of clothes since lasthandling bulk explosives or visiting areas where explosives are storedor processed. They must also be trained in an FEL Standard OperatingProcedures and Standard Methods relating to entry procedures andworking practices (SOP205, SOP214 and SM114). Access to the labo-ratory is restricted and is by swipe card only.

Once these precautions have been observed, the laboratorymay beentered. A plan of the Main Trace Laboratory is shown below (Fig. 1).The current entry procedure into the trace laboratory is as follows:

i) The operator enters the trace laboratory corridor, treading on aMicro-pure® clean room ‘sticky’ mat with both feet.

ii) The air handling systemmonitoring panel is observed ensuringthat the air filtration system is in working order, with adequatepositive air pressure, and entry is permitted.

iii) The operator removes outdoor shoes and dons a pair of tracelaboratory shoes, checks the particle monitor, removes allwrist and finger jewellery, and signs the entry logbook.

Examination benches

Examination room

Entrance

Work benches

Sampled floor areas

Swing over bench

Particle monitor

AHU monitoring panel

Ionscan equipment

A Trace Area Corridor B Outer Lobby C Inner Lobby

A

B

C

Fig. 1. Outline map of Main Trace Laboratory (

iv) The operator enters the outer lobby area, stepping on anotherMicro-pure® clean room mat with both feet and immediatelywashes their hands and wrists thoroughly, drying them withdisposable paper towel.

v) The operator dons a pair of disposable gloves and while sittingon the narrow, low, bench dividing the outer and inner lobbies,puts on one disposable overboot, and swings the covered footover the bench onto the inner lobby floor. An overboot isthen placed on the other foot in the same manner. The opera-tor can then stand up in the inner lobby.

vi) The operator dons a disposable oversuit and a disposable hat.vii) An Ion Mobility Spectrometer (Smiths Detection Ionscan 400B)

is used to sample the outer surface of the disposable oversuit.If this indicates the presence of the explosives NG, TNT, PETNor RDX, entry is not permitted. If the Ionscan equipment isnot operational, due to repair or maintenance work, entry isstill permitted provided all the other detailed procedures havebeen carried out. With rigorous monitoring, access controlsand contamination prevention procedures, the possibility of acontaminated user entering the laboratory is remote.

viii) On entering the main trace area, before touching any surface,the operator removes their disposable gloves, washes theirhands and dons a clean pair of disposable gloves.

When all steps described in (i)–(viii) have been followed, a worksurface is prepared. The work surface is cleaned with proprietarycleaner, followed by ethanol, and covered with disposable glazed paper.This ensures that if any traces of explosives were present on the benchthey are either removed or isolated. If casework is to be carried out, con-trol samples would be taken at this stage.

Further protection is affected by controlling all materials and airentering the laboratory. All materials used in the trace laboratoryare carefully sourced so as to optimise their fitness for purpose. Ap-proved suppliers are defined in the laboratory quality management

Instrument benches

Main laboratory

Kit roomAnalytical

bench

not to scale) showing locations sampled.

219A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

system and all materials are covered in extra wrapping at thesupplier's premises, which is removed in stages, during passagethrough the lobby areas, before entry. This ensures that any materialsdestined for the trace laboratory cannot become contaminated ontheir passage through the site. The air supply to the laboratory passesthrough bag filters, followed by large high efficiency particle arrestor(HEPA) filters that remove suspended particles. The filtered air issampled, post‐HEPA, by a particle monitoring system (five ClimetCI-3100 Optical Particle Transducer Instruments linked to a data sys-tem, OptiCal Sciences Ltd, Northampton, UK) via a port mountedabove the fire damper before it enters the laboratory via a diffusergrille. The flow of air is controlled so as to maintain a slightly higherpressure within the laboratory than outside. At the entrance of thetrace laboratory, there is a display panel allowing the air handlingsystem to be monitored before entry. A formalised standard operatingprocedure relates to monitoring, data processing and system mainte-nance (Standard Operating Procedure SOP214).

A formalised laboratory cleaning rota (Standard Operating Proce-dure SOP203), ensures that, in the unlikely event of small explosivestraces being present in the laboratory, they cannot persist or accumu-late. This rota has been revised periodically to take into account variousmodifications to the laboratory, its procedures, and the introduction ofnew equipment.

The intention of the overall system is that contamination of sam-ples could only take place if multiple and simultaneous breaches ofcontamination avoidance procedure were to occur. The routine useand discarding of disposable clothing, gloves and paper along withthe laboratory cleaning regime is designed to ensure that a set ofcasework samples is not put at risk by any previous set. In order tomonitor the effectiveness of the laboratory contamination preventionprocedures, samples are regularly taken from various surfaces withinthe laboratory.

3. The laboratory monitoring regime

3.1. Apparatus, materials and analytical procedure

Laboratorymonitoring samples are taken using solvent-moistenedcotton wool swabs. The solvent used since March 1996 is a mixture ofethanol and water in equal volumes. Sample processing and analysishas evolved slightly since 1989, but has been in essence that describedby Crowson et al. [8]. The moistened swabs are extracted using thesame solvent, followed by solid phase extraction using speciallyprepared clean-up tubes containing Chromosorb 104 (Sigma-Aldrich,Dorset, UK, 60–80mesh) and subsequent desorption of the explosivestraces for GC/TEA analysis using ethyl acetate (Standard OperatingProcedure SOP200). The resulting extracts are concentrated by evapo-ration under dry nitrogen to about 100 μl and their volumes arecarefully estimated by eye, since the desired results are only semi-quantitative in nature. This reduces sample handling and opportuni-ties for cross contamination. March 2009 saw the introduction of asimilar method using commercially available solid phase extractiontubes containing Isolute ENV+ (Biotage UK, 200 mg, 6 ml) to enablethe recovery of the peroxide explosives HMTD and TATP from theChromosorb-104 clean-up eluate (Standard Operating ProcedureSOP220). The peroxide explosives are desorbed from the IsoluteENV+ tube using acetonitrile which is suitable for LC/MS/MS analysis.All monitoring samples are taken and processed using materials fromthe supplies also used in casework; this providesweekly quality assur-ance of materials used in the processing of forensic samples. A singleGC/TEA analysis is made of each ethyl acetate sample and candidateexplosives responses are confirmed by further analysis (see “ActionCriteria” section below) (StandardMethod SM110). A controlled stan-dard, containing known quantities of NG, TNT, PETN and RDX, is usedto measure the limit of detection on the GC/TEA instruments on aformalised weekly basis (Standard Method SM112). The acetonitrile

extract is split into two aliquots which are analysed by LC/MS/MS,one for each of the peroxide explosives (Standard Methods SM171and SM172).

The organic explosives and peroxide explosives chosen for qualityassurance sampling within the FEL trace areas represent those whichhave been, and are perceived to be, the primary threats to UK publicsafety.

3.2. Locations sampled — main trace laboratory

Monitoring samples are taken from all of the laboratory benchsurfaces upon which samples are processed (Standard Operating Pro-cedure SOP202). Fig. 1 is an outline map showing the locations sam-pled in the main trace laboratory. In order to reduce the analyticalburden, samples from three examination/sampling benches are unit-ed as one (referred to below as ‘examination benches’, total swabbedarea 11.25 m2), and the relatively large bench upon which swab ex-tractions and clean-ups are carried out, is not sub-divided (‘analyticalbench’, total swabbed area 4.80 m2). A series of benches upon whichall of the analytical instruments stand are again sampled together(‘instrument benches’, total area 16.20 m2). The three benches inthe room used only for the preparation of swabs and explosivesrecovery kits are also sampled as one (‘kit room’, total swabbed area2.55 m2). It would be very time-consuming to swab the entire labora-tory floor, therefore a series of 14 boxes, that cover a total area ofapproximately 4.25 m2, has been marked in well-trodden sectionsand these are sampled (‘floor’). This is approximately 3.8% of thetotal accessible floor area.

Additionally, monthly monitoring samples are taken from areasthat are not normally covered during weekly sampling, for examplethe telephones, the clampstands, the windowsills, etc. These monthlymonitoring samples are taken according to a rota introduced into thequality assurance standard operating procedure in December 1997.

The weekly monitoring samples have been taken (with few excep-tions) since late 1989. Over the years, progressively more areas havebeen sampled and there was a major change when the laboratorymoved to a new building early in 1992. During the period under dis-cussion the most significant change was the addition of clean-up pro-cedures and analysis for the peroxide explosives in June 2009.

One swab is prepared alongside those used for sampling, but isretained unused as a control. It is then processed and the extractanalysed alongside the monitor samples. Three samples spiked atlow levels with explosives standard are also processed and analysed,the spiked swab sample, the spiked solution sample and the spikedswab–peroxides sample. The percentage of explosives recoveredfrom the spiked samples shows the efficiency of the extraction andclean-up process. Data related to the recovery of the standard organicexplosives from spiked samples has been presented previously [9].The average recoveries of NG, TNT, PETN and RDX from the spikedswab samples during 2010 were 35%, 38%, 53% and 32% respectively.The peroxides spiked swab is spiked with 150 ng of each of the per-oxide explosives. The average recoveries from the swabs during thefirst 7 months of 2011 are 39% for HMTD and 47% for TATP.

3.3. Locations sampled — secondary trace laboratory

The secondary trace laboratory is used as a back up facility for themain trace laboratory. It is used mainly for research but is availablefor use with casework exhibits that are suspected of being heavilycontaminated. On entry to the smaller secondary trace laboratory,similar precautions are taken as when entering the main trace labora-tory, one difference being the lack of any Ionscan® equipment. Mon-itoring samples are taken from all benches in the laboratory uponwhich samples may be processed and from 4 marked floor areas.Fig. 2 is an outline map showing the locations sampled. The instru-ment benches (total swabbed area, 9.0 m2) are sampled together

Instrument bench

Instrument bench

Analytical bench

EntranceSwing over bench

Sampled floor areas

Fig. 2. Outlinemapof Secondary Trace Laboratory (not to scale) showing locations sampled.

Benches

Drainage channel

Swing over bench

1

4

6

2

3

5

Roller shutter door

Sampled floor area

Vehicle entrance

Personnel Entrance

Fig. 3. Outlinemap of Trace Vehicle SamplingBay (not to scale) showing locations sampled.

220 A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

and treated as one area. The analytical bench (total swabbed area,6.7 m2) is also sampled. The 4 floor areas are swabbed and treatedas one sample (approximately 1.8 m2/7.9% of the total accessiblefloor area).

One swab is prepared alongside those used for sampling, but isretained unused as a control sample. Monitoring samples from thesecondary trace laboratory are taken at least monthly. The secondarytrace laboratory was completely refurbished in late 1998. The firstmonthly quality assurance sampling after refurbishment was takenon 27th January 1999.

3.4. Locations sampled — trace vehicle examination bay

The trace vehicle examination bay is used for trace sampling ofvehicles and other exhibits too large to be taken into the other tracelaboratories. Entry procedures for personnel and materials are identi-cal to those described for the secondary trace laboratory. Entry andexit of vehicles and other large exhibits is through a roller shutterdoor into the main trace bay. Monitoring samples are taken from allbenches (total swabbed area, 6.2 m2). Prior to April 2007 the bencheswere swabbed both before and after the cleaning which was carriedout during the QA procedure. The benches were swabbed, cleanedwith proprietary cleaner and ethanol, and swabbed again with afresh swab. In view of the results obtained during this period in April2007 the method was changed and the benches are now swabbedonce, as found. Fig. 3 is an outlinemap showing the locations sampled.Floor area 3 is sampled each time monitoring samples are taken. Oneother floor area is also sampled on a rota basis. Prior to April 2007the 2 sampled floor areas were swabbed separately as 2 differentmonitoring samples but since then they have been sampled usingthe same swab. They cover an area of approximately 2 m2 and makeup approximately 2% of the total accessible floor area.

One swab is prepared alongside those used for sampling, but isretained unused as a control sample. Another swab was used to sam-ple the operator(s) and the glazed paper work surface until April2007 whereupon this sample was discontinued bringing the systemin line with that used in the other trace areas.

Monitoring samples are taken at leastmonthly and/or immediatelybefore a vehicle is placed in the trace bay. Monitoring samples are alsotaken after cleaning has been carried out once forensic samples havebeen removed from the trace bay. The trace vehicle examination baywas completely refurbished in early 1997 and again in March 2010.The first monthly quality assurance samples, after refurbishment,were taken on 6th March 1997 and 30th March 2010 respectively.

3.5. Action criteria

Themain purpose of taking themonitoring samples is to ensure anddemonstrate the continuing cleanliness of the laboratory. Where anyexplosives contamination is detected, actions are taken in accordancewith the graduated response protocols summarised in Tables 1 and 2.The levels of explosives used to define the action criteria are basedupon several factors. These are, the limit of detection of the entire pro-cedure, the levels considered significant during casework, and experi-ence of carrying out such work over a number of years. With regard tothe limit of detection for the entire procedure, the aim is to achieve adetection of 1 ng in 100 μl ethyl acetate extract for the standard organicexplosives. Limits of detection are checked on a weekly basis byanalysing a standard solution containing 100 pg/μl NG, TNT, PETN andRDX. The results obtained are used to calculate the limits of detectionfor these explosives for that week, which are then formulated into thecontrol charts. The control charts are carefully maintained. The levelsused to define the action criteria are kept under review.

3.6. Accreditation

It is a Home Office requirement that, in order for the laboratory tocarry out forensic casework, it shall be externally accredited by theUnited Kingdom Accreditation Service (UKAS) to ISO 17025. ISO17025 is an internationally recognised standard of competence fortesting laboratories and accreditation is based on annual audit.UKAS appointed auditors visit the laboratory and observe demonstra-tions of specified methods and check the results obtained so they mayevaluate whether the work carried out is of the appropriate standard.

4. Summary of test results and discussion

4.1. Monitoring samples — main trace laboratory

Themonitoring sample results have been reviewed and assembledinto a database. The results from February 2004 to July 2011 areexamined here. However, to set these results into the context of pre-vious years, they have been added to the results since 1989 in Fig. 4.This shows that the number of explosives detections in monitoringsamples has decreased markedly over recent years. RDX has beendetected in these samples more frequently than other explosives,since it is the most common nitro-containing explosive detected in

Table 1Action criteria when organic high explosives are detected.

Nominal amount found (ng) Action to be taken

Less than 5 No action required. Levels are acceptable. The result may be confirmed on other TEA systems at the discretion of the Trace Laboratory Manager.Between 5 and 10 1. Confirm result on other TEA systems.

2. Clean the area.Greater than 10 1. Confirm result on other TEA systems.

2. Suspend operations until area is proved to have b5 ng explosives present.3. Re-swab the area and check that the levels of explosives are acceptable before resuming operations.4. The Trace Laboratory Manager or Team Leader/Laboratory and Analytical Services shall review the results as soon as possible and authorise

resumption of work within the laboratory. The actions and authorisation shall be documented on the FEL Trace Areas QA Review Form.Greater than 100 1. As for “Greater than 10” above.

2. A thorough inquiry is to be held. This inquiry must cover the following aspects:

a) what were the possible sources of contamination?b) could such an incident be avoided in future by changing any procedures?c) could the incident have been dealt with more effectively?d) is there any potential effect upon casework processed in the laboratory during the period under investigation?

3. The above inquiry may be carried out, at least in part, by the Trace Laboratory Manager but must have the full approval of the Team Leader/Laboratory and Analytical Services who is ultimately responsible. The progress and results of the inquiry shall be made available to the PrincipalForensic Investigator for consideration in respect to any possible impact on casework.

4. Recommendations from such inquiries must be considered at the following Quality System Review Meeting and those which are acceptedmust be implemented as soon as practicably possible, but no later than six months after the meeting.

Table 2Action criteria when HMTD or TATP are detected.

Action to be taken on detecting HMTD or TATP

1. Suspend operations until area is proved to have no detectable HMTD or TATP present.2. If TATP has been detected using SRM mode, confirm the result using SIM mode3. Clean the area.4. Re-swab the area and check that the levels of explosives are acceptable before resuming operations.5. The Trace Laboratory Manager or Team Leader/Laboratory and Analytical Services shall review the results as soon as possible and authorise resumption of work within thelaboratory. The actions and authorisation shall be documented on the FEL Trace Areas QA Review Form.

6. A thorough inquiry shall be held. This inquiry must cover the following aspects.

– What were the possible sources of contamination?– Could such an incident be avoided in future by changing any procedures?– Could the incident have been dealt with more effectively?– Is there any potential effect upon casework processed in the laboratory during the period under investigation?

7. The above inquiry may be carried out, at least in part, by the Trace Laboratory Manager but must have the full approval of the Team Leader/Laboratory and Analytical Serviceswho is ultimately responsible. The progress and results of the inquiry shall be made available to the Principal Forensic Investigator for consideration in respect to any possibleimpact on casework.

8. Recommendations from such inquiries must be considered at the following Quality System Review Meeting and those which are accepted must be implemented as soon aspracticably possible, but no later than six months after the meeting.

221A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

casework samples. It is the major constituent of Semtex-H, PE4 andC4 plastic explosives.

During the period November 1989 to July 2011 6230 sampleswere taken of which 1099 were control samples, leaving a total of

Fig. 4. Explosives detected in monitoring samples of Main Trace Laborato

5131 from the laboratory benches or the laboratory floor. The datapresented excludes unconfirmed detections less than 5 ng of explo-sives other than RDX prior to 1996, since in view of subsequent datait is unlikely that these would have been confirmed. Since October

ry — Nov 1989–Jul 2011 — number of positive samples versus time.

Fig. 5. Explosives in monitoring samples of Main Trace Laboratory — Nov 1989–Jul 2011 — number of samples versus estimated mass.

222 A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

1996, it has been the norm to confirm all detections however small.During the period November 1989 to July 2011 only 3.5% of the sam-ples taken from the laboratory benches or floor have contained morethan 5 ng of explosives and 87.4% have been negative. Fig. 5 showsthe distribution by amount detected in the main trace laboratoryduring the period November 1989 to July 2011 and Fig. 6 shows thelocations of these detections.

During the period February 2004 to July 2011, there have been8 positive samples, 5 containing RDX and 3 containing other explo-sives. These results are summarised in Table 3. This means that duringthe period February 2004 to July 2011 there were 385 sets of samples.Of the samples from the laboratory benches or the laboratory floor1917 were negative, i.e. 99.6%. No samples contained two or moreexplosives, which demonstrates that low-level multi-componentstandards, used to calibrate the analytical instruments and for spikedsamples, are not a source of laboratory contamination.

Since the introduction of analysis of QA samples for the peroxideexplosives HMTD and TATP in June 2009, there have been no detec-tions in any of the Main Trace Laboratory samples.

By its very nature, casework will always be the most likely sourceof contamination since the levels of explosives are unknown at the

Fig. 6. Explosives in monitoring samples of Main Tra

time of sampling. Thus all casework presents an inherent contamina-tion risk.

Over 160 monthly monitoring samples have been taken duringthe period February 2004 to July 2011. No explosives were detectedin any of these samples.

No explosives were detected in any of the 385 control swabsanalysed during this period.

Unpublished research work, carried out by the FEL, indicates thatthe site background contamination at Fort Halstead consists primarilyof low levels of TNT, PETN and RDX. Significantly, there has not been asingle occasion in the main trace laboratory when RDX has beendetected along with either TNT or PETN, in the same monitoring sam-ple, during the period February 2004 to July 2011. This indicates thatthe contamination prevention procedures work effectively againstthe background contamination. The presence of such backgroundcontamination could be viewed as advantageous in that it providesa constant challenge to the procedures and validates their effective-ness. Furthermore, quality assurance procedures similar to thoseoutlined here would still be required if there were no site backgroundcontamination since casework containing explosives traces wouldstill be examined within the laboratory.

ce Laboratory — Nov 1989–Jul 2011 — locations.

Table 3Explosives detections in trace areas — February 2004–July 2011.

Trace area Date Explosivedetected

Amountdetected(ng)

Area sampled

Main 29/03/04 PETN 146.5 Floor30/08/05 RDX 41.9 Instrument benches03/10/05 RDX 32.5 Floor17/10/05 RDX 12.0 Floor11/02/08 TNT 2.2 Analytical bench

TNT 3.1 Instrument benches25/03/08 RDX 7.3 Examination benches28/07/08 RDX 2.0 Examination benches

Secondary 05/05/05 RDX 11.0 Floor20/01/06 RDX 3.7 Instrument benches12/04/06 RDX 24.7 Floor04/08/06 RDX 9.8 Instrument benches08/05/09 TNT 39.6 Instrument benches

PETN 10.6RDX 1238.6TNT 8.9 FloorRDX 18.5RDX 3.1 Analytical bench

15/05/09 RDX 35.4 Instrument benches20/05/09 RDX 2.2 Analytical bench

Vehicle bay 21/05/04 RDX 32.5 Floor area 309/07/05 RDX 35.1 Floor area 313/01/06 RDX 14.5 Floor area 316/03/07 RDX 1.8 Benches before cleaning

223A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

4.2. Monitoring samples — secondary trace laboratory

Themonitoring samples from January 1999 to July 2011 have beenassembled into a database and reviewed. As in the case of the maintrace laboratory, RDX is the most commonly detected explosive inthe monitoring samples. Along with the RDX detected, TNT andPETN have been found on one occasion in an Instrument Bench sam-ple in May 2009 along with TNT in the Floor sample. These detectionswere a direct result of processing casework. The database contains161 sets of results making a total of 644 samples of which 161 arecontrol samples. Of the 483 laboratory bench and floor samples, 19contained RDX of which only 10 were above 5 ng. Therefore only3.9% of all the samples taken have been confirmed as containing ex-plosives and only 2.1% contain more than 5 ng. Fig. 7 shows the distri-bution by amount detected in the secondary trace laboratory andFig. 8 shows the locations of these detections.

Fig. 7. Explosives in monitoring samples of Secondary Trace Laboratory

Table 3 summarises the explosives detections in the secondarytrace laboratory during the period February 2004 to July 2011.

Since the introduction of analysis of QA samples for the peroxideexplosives HMTD and TATP in June 2009, there have been no detec-tions in any of the Secondary Trace Laboratory samples.

Of the 161 control samples that have been taken between January1999 and July 2011, none have been found to contain any explosives.

4.3. Monitoring samples — trace vehicle examination bay

The monitoring samples taken from the Trace Vehicle Examina-tion Bay between March 1997 and February 2010 have been collatedinto a database and reviewed. There are 209 sets of data with a totalof 1109 samples of which 374 are control samples. In common withthe other trace facilities, RDX is the most commonly detected explo-sive in the monitoring samples. During this period the percentage ofpositive samples, excluding controls, is 4.9%. The percentage of posi-tive samples containing more than 5 ng of explosives, excluding con-trols, is 1.2%. Fig. 9 shows the distribution by amount detected andFig. 10 shows the locations of the detections.

The 16 sets of samples taken after the most recent refurbishment(March 2010 to July 2011) have all been explosives free.

During the period February 2004 to July 2011 only 4 samples havebeen found to contain explosives all of which contained RDX. Theseare summarised in Table 3.

Since the introduction of analysis of QA samples for the peroxideexplosives HMTD and TATP in June 2009, there have been no detec-tions in any of the Trace Vehicle Examination Bay samples.

The fact that some of the Trace Vehicle Examination Bay monitor-ing samples, taken prior to February 2004, contain both RDX and TNTindicates that these explosives are constituents of the site backgroundand highlights the difficulty of keeping an area trace clean when a ve-hicle has to enter via a large roller shutter door. It is important to notethat, when casework exhibits are sampled, further controls are takenonce the exhibit has been brought into the bay.

No explosives were detected in any of the 209 control swabstaken.

4.4. Control samples

During the period February 2004 to July 2011 there have been noexplosives detected in any of the control swabs taken, processed andanalysed alongside monitoring samples during the period under

— Jan 1999–Jul 2011 — number of samples versus estimated mass.

Fig. 8. Explosives in monitoring samples of Secondary Trace Laboratory — Jan 1999–Jan 2011 — locations.

Fig. 9. Explosives in monitoring samples in Trace Vehicle Examination Bay — Mar 1997–Feb 2010 — number of samples versus estimated mass.

Fig. 10. Explosives in monitoring samples in Trace Vehicle Examination Bay — Mar 1997–Feb 2010 — locations.

224 A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

225A. Crowson, R. Cawthorne / Science and Justice 52 (2012) 217–225

question. The control swab results, taken together, are most signifi-cant because they demonstrate that the risk of forensic sample con-tamination, arising either from contaminated sampling materials orfrom contamination during processing, is extremely small, evenwhen one or other of the monitoring samples taken at the same timeshows that some contamination was present in the laboratory. Thereason for this observation is that the prevention procedures have ef-fectively isolated the samples from contamination. These issues havebeen discussed in detail elsewhere [9].

5. Conclusions

A system of contamination prevention procedures incorporatingboth inner and outer protective measures has been implemented,with progressive improvements over the last 22 years. Monitoringsamples taken from surfaces within the trace laboratories and tracevehicle examination bay have, with few exceptions, revealed onlylow levels of contamination, predominantly of RDX. Since the intro-duction of analysis of QA samples for the peroxide explosives inJune 2009, there have been no detections of HMTD or TATP in anyof the samples.

Analysis of the control swabs, processed alongside the monitoringswabs, has demonstrated that in this environment the risk of forensicsample contamination, assuming all the relevant anti-contaminationprocedures have been followed, is so small that it is considered tobe negligible. The monitoring regime has also been valuable inassessing the process of continuous improvement, allowing sourcesof contamination transfer into the trace areas to be identified andeliminated. The main risk of contamination comes from caseworksamples themselves. The complete lack of positive monitoring sam-ples, during the period February 2004 to July 2011, for the main labo-ratory containing RDX along with either TNT or PETN shows that the

contamination prevention procedures in place are effective againstthe background explosives levels at Fort Halstead.

Acknowledgments

The United Kingdom Home Office funded this work. The authoralso acknowledges the contributions of all those who have carriedout the weekly monitoring tests; a routine, but most important, activ-ity. Additionally, the author also acknowledges the sterling efforts ofthe support staff who carry out cleaning within the trace areas.

References

[1] D.H. Fine, W. Yu, U.E. Goff, E.C. Bender, D.J. Reutter, Picogram analysis of explosiveresidues using the thermal energy analyser (TEA), Journal of Forensic Sciences 29(1984) 732–746.

[2] J.M.F. Douse, Trace analysis of explosives at the low nanogram level in hand swabextracts using columns of Amberlite XAD-7 porous polymer beads and silica capillarycolumn gas chromatography with thermal energy analysis and electron capturedetection, Journal of Chromatography 328 (1985) 155–165.

[3] A. Crowson, M.S. Beardah, Development of an LC/MSmethod for the trace analysis ofhexamethylenetriperoxidediamine (HMTD), The Analyst 126 (2001) 1689–1693.

[4] L. Widmer, S. Watson, K. Schlatter, A. Crowson, Development of an LC/MS methodfor the trace analysis of triacetone triperoxide (TATP), The Analyst 127 (2002)1627–1632.

[5] A. Crowson, R.W. Hiley, C.C. Todd, Quality assurance testing of an explosive traceanalysis laboratory, Journal of Forensic Sciences 46 (1) (2001) 53–56.

[6] A. Crowson, S.P. Doyle, C.C. Todd, S. Watson, N. Zolnhofer, Quality assurance testingof an explosive trace analysis laboratory — further improvements, Journal ofForensic Sciences 52 (4) (2007) 830–837.

[7] S. Doyle, Quality and the trace detection and identification of organic high explo-sives, in: A. Beveridge (Ed.), Forensic Investigation of Explosions, Second Edition,Taylor and Francis, London, 2012, pp. 539–583.

[8] A. Crowson, H.E. Cullum, R.W. Hiley, A.M. Lowe, A survey of high explosives tracesin public places, Journal of Forensic Sciences 41 (5) (1996) 980–989.

[9] M.S. Beardah, S.P. Doyle, C.E. Hendey, Effectiveness of contamination preventionprocedures in a Trace Explosives Laboratory, Science & Justice 47 (2007) 120–124.