Flow injection analysis of organic peroxide explosives using acid degradation and chemiluminescent...
8
Dear Author, Please, note that changes made to the HTML content will be added to the article before publication, but are not reflected in this PDF. Note also that this file should not be used for submitting corrections.
Transcript of Flow injection analysis of organic peroxide explosives using acid degradation and chemiluminescent...
Dear Author Please note that changes made to the HTML content will be added to the article before publication but are not reflected in this PDF Note also that this file should not be used for submitting corrections
Flow injection analysis of organic peroxide explosives using aciddegradation and chemiluminescent detection of released hydrogenperoxide
Parvez Mahbub a Philip Zakaria a Rosanne Guijt b Mirek Macka a Greg Dicinoski aMichael Breadmore a Pavel N Nesterenko anQ1
a Australian Centre for Research on Separation Science School of Physical Sciences University of Tasmania Australiab Pharmacy School of Medicine Australian Centre for Research on Separation Science University of Tasmania Australia
a r t i c l e i n f o
Article historyReceived 8 April 2015Received in revised form12 May 2015Accepted 22 May 2015
KeywordsOrganic peroxidesExplosivesFlow injection analysisAcid degradationChemiluminescence
a b s t r a c t
The applicability of acid degradation of organic peroxides into hydrogen peroxide in a pneumaticallydriven flow injection system with chemiluminescence reaction with luminol and Cu2thorn as a catalyst (FIA-CL) was investigated for the fast and sensitive detection of organic peroxide explosives (OPEs) The targetOPEs included hexamethylene triperoxide diamine (HMTD) triacetone triperoxide (TATP) and methy-lethyl ketone peroxide (MEKP) Under optimised conditions maximum degradations of 70 and 54 forTATP and HMTD respectively were achieved at 162 mL min1 and 9 degradation for MEKP at180 mL min1 Flow rates were precisely controlled in this single source pneumatic pressure drivenmulti-channel FIA system by model experiments on mixing of easily detectable component solutionsThe linear range for detection of TATP HMTD and H2O2 was 1ndash200 mM (r2frac14098ndash099) at both flow rateswhile that for MEKP was 20ndash200 mM (r2frac14097) at 180 mL min1 The detection limits (LODs) obtainedwere 05 mM for TATP HMTD and H2O2 and 10 mM for MEKP The detection times varied from 15 to 3 minin this FIA-CL system Whilst the LOD for H2O2 was comparable with those reported by other in-vestigators the LODs and analysis times for TATP and HMTD were superior and significantly this is thefirst time the detection of MEKP has been reported by FIA-CL
amp 2015 Published by Elsevier BV
1 Introduction
The simplicity of in-house preparation of the organic peroxideexplosives (OPEs) such as hexamethylene triperoxide diamine(HMTD) triacetone triperoxide (TATP) and methylethyl ketoneperoxide (MEKP) from readily available materials was the mainreason of their use in some recent terrorists attacks [1] So there isa strong demand for the development of fast simple and sensitivemethods of their identification and quantitative determination [2]This task is not trivial because of high volatility and absence ofchromophoric groups in the molecules of OPEs For these reasonsthe use of common analytical methods such as GC or HPLC withUVvisible detection is not readily suitable for their direct de-termination so the application of more complex hyphenatedtechniques typically involving mass spectrometry (MS) is re-quired Schulte-Ladbeck et al [3] proposed RP HPLC with on-line
Fourier transfom infrared (FTIR) detection for direct determinationof TATP and HMTD De Tata et al [4] reported the application of RPHPLC with quadrupole time-of-flight mass spectrometry (HPLC-QToF-MS) for direct determination of various OPEs However mi-cellar electrokinetic chromatography with UV detection was em-ployed by Johns et al [5] recently for separation of OPEs includingHMTD and TATP in post blast scenario without any marked im-provement compared to the hyphenated HPLC methods in termsof sensitivity
Alternatively the determination of OPEs can be based onelectrochemical [6] fluorescent [7] and chemiluminescent [8]detection of hydrogen peroxide as the main degradation productof OPEs It should be noted that decomposition of one OPE mole-cule can result in more than one molecule of H2O2 so in case of100 degradation a magnified analytical response can be expectedBecause of its robustness sensitivity and simplicity of integrationwith FIA chemiluminescent detection (FIA-CL) is one of the mostpopular techniques for the on-line detection of hydrogen peroxideThe application of FIA-CL has been reported for the determinationof H2O2 in natural waters [9] rainwater [10] and seawater [11] Theuse of chemiluminescent detection of H2O2 after decomposition of
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
6768697071727374757677787980818283848586878889909192
Contents lists available at ScienceDirect
journal homepage wwwelseviercomlocatetalanta
Talanta
httpdxdoiorg101016jtalanta2015050520039-9140amp 2015 Published by Elsevier BV
n Correspondence to Australian Centre for Research on Separation ScienceSchool of Chemistry University of Tasmania Private Bag 75 Hobart 7001 AustraliaFax thorn61 03 6226 2858
E-mail address PavelNesterenkoutaseduau (PN Nesterenko)
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎
TATP and HMTD was reported for determination of these ex-plosives by a number of research groups [3612] The analysistimes reported varied from 5 to 12 min with the sensitivity ran-ging from 25 mM for TATP [6] to 500 mM for HMTD [3]
Obviously the crucial parameters in the flow-through methodsof analysis of explosives via detection of hydrogen peroxide are thevelocity and conversion degree of OPEs into H2O2 According to theliterature data the degradation of peroxide explosives can be ac-complished enzymatically [1213] photochemically [14] or byusing mineral acids [15] The latter option appears more simpleand robust but it takes almost 12 h for complete decomposition ofTATP and HMTD which is not suitable for rapid screening Ad-ditionally there has been no report of the decomposition of me-thylethyl ketone peroxide (MEKP)-listed as a priority explosive byseveral government agencies including US National Counter-terrorism Centre and the Australian Army [16] It should be notedthat 100 degradation of OPEs does not mean the maximumpossible concentration of hydrogen peroxide suitable for detectionas hydrolysis of H2O2 occurs together with organic peroxides Onthis reason acid degradation conditions of OPEs should be care-fully attuned to provide fast and sensitive response in FIA system
This study was a part of a larger project directed on construc-tion of portable automated system for the fast screening of sam-ples on presence of traces of homemade explosives including or-ganic peroxides and inorganic explosives According to the pro-posed design the whole system composed of three separate unitsnamely sample extraction unit FIA-CL detection unit for organicperoxides and capillary zone electrophoresis unit for parallelprofiling of inorganic anions contents The high degree of auto-mation minimal consumption of reagents and robust operationfor extended period of time were considered as important re-quirements for this system The system should provide eitherpositive or negative answer in a short time on the presence ofexplosives at low concentration level
Typically multicomponent FIA analysers use multi-channelperistaltic pumps for delivery of sample and reagents solutionsAccording to Dasgupta to get very low flow rates at mLmin levelas required in this study peristaltic pumps are of little help [17]Piezoelectric pumps pneumatic pressure or gravity driven systemsare often preferred choice in this case [18] Pressure driven FIAinstruments have advantages over systems using peristaltic pumpsin terms of lower signalnoise ratio better reproducibility of tim-ing and expanded possibilities for the use of aggressive solventsand carriers [18] According to Valcarcel and De Castro the maindrawback of pneumatic driven instruments is connected withdifficulty in accurate control of the flow rates in multiple channelFIA systems due to complex changes in hydraulic resistances re-sulting from different channel geometry and reagent viscosity [19]In this study a special attention was paid to the development of acontrol mechanism of flow rates in separate lines of multichannelFIA system driven by pneumatic pressure from a single source
The objective of this study is to optimise conditions for aciddegradation of organic peroxides and subsequent chemilumines-cent luminol based detection of hydrogen peroxide within pres-sure driven FIA-CL analytical unit for the purpose of fast qualita-tive determination of HMTD TATP and MEKP The target totalanalysis time was less than 2 min for rapid screening and high-throughput analysis imposing a significant challenge to the opti-misation of the flow conditions
2 Experimental section
21 Reagents and chemical standards
Hydrated copper sulphate (CuSO45H2O) 32 concentrated
hydrochloric acid (HCl) 30 (ww) reagent grade H2O2 iso-propanol and sodium hydroxide (NaOH) pellets were purchasedfrom Sigma-Aldrich (Sydney Australia) Luminol was purchasedfrom Fluka (Sydney Australia) Element free deionised water wasused to prepare all stock and working solutions The TATP standard(10000 mg L1 999 single component) and HMTD standard(5000 mg L1 984 single component) were procured from Ac-custandard USA The MEKP standard (10000 mg L1) was sup-plied by the Australian Defence Science and Technology Organi-sation (DSTO)
22 Preparation of precise assay of the standard solution for H2O2
A stock solution of approximately 1000 mg L1 of the 30 H2O2
was prepared through serial dilution Then 50 mL of stock solutionwas transferred into a 500 mL conical flask diluted with 200 mL ofdeionised water and then 30 mL of 25 sulphuric acid was addedThe solution was titrated with a standard 002 M potassium per-manganate solution until the colour changed to pink The workingsolutions were prepared by further dilution of stock solution in DIW
23 Instrumentation
A FIA-Cl system consisting of a low pressure Cheminert 6 port2 position injector valve (C22-3186EH-FL VICI Houston USA) fiveSMC precision pressure regulators (IR 1000-01 SMC Japan) and aHamamatsu photomultiplier (10493-001 Hamamatsu Japan) wereused A schematic of the instrument used for acid degradationstudy of TATP HMTD and MEKP is illustrated in Fig 1
The system comprised SMC pressure regulators connectedthrough a manifold to an external compressed air supply Eachregulator was used to control the reagent solution flow from a500 mL glass bottle (Schott AG Sigma-Aldrich Australia) by ap-plying pressures ranging from 001 to 02 MPa (15ndash29 psi) Thepneumatic lines consisted of polyurethane tubing (25 mm IDSMC Japan) and the hydraulic lines consisted of FEP tubing(0203 mm ID Upchurch USA)
The 100 mL sample plug was carried into the Cheminert mixer(CM1XKF VICI Houston USA) in 5050 vv deionised waterndashiso-propanol where it was mixed with 32 HCl Isopropanol was usedto ensure complete dissolution of OPEs from the collected andextracted samples (the exact procedure is not included in thispaper) The OPE containing samples were degraded in acidicmixture in a 1 m PTFE knitted tubing coil reactor (025 mm ID491 mL internal volume Biotech AB Onsala Sweden) resulting inthe release of H2O2 molecules The excess HCl was neutralised bythe addition of 18 NaOH and the resulting solution was thenmixed with the luminol ndash Cu2thorn reagent in a tee mixer (P-71229 mL swept volume Upchurch Oak Harbour USA) The sub-sequent chemiluminescence reaction was detected with the pho-tomultiplier tube The chemiluminescence flow cell was fabricatedin-house using FEP tubing (0508 mm ID 1548 L Upchurch USA)with the total volume of the flow cell being 8532 mL The chemi-luminescence signal was acquired by a Powerchrom data acquisi-tion system (ER280 EDaq Sydney Australia) with proprietarysoftware version 81 The pH of the effluent was monitored by anin-line pH monitoring flowcell (Cole-Parmer Australia verticalflow glass electrode 50 mL internal volume) as shown in Fig 1
3 Results and discussions
31 Optimisation of reagent concentrations
The crucial part of FIA-CL under development is acid degrada-tion of OPEs in isopropanolndashwater (5050) extracts of swabs used
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎2
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
for sample collection from different surfaces It was shown thatsuch mixture provides the best extraction of various types of in-organic and organic explosives [20] However as a result of ele-vated viscosity of this mixture the FIA system in this study need tooperate at pressures higher than those normally provided byperistaltic pumps Also initial experiments using peristaltic pumpsdemonstrated poor stability of the silicone tubing when in contactwith solutions of isopropanol hydrochloric acid and sodium hy-droxide requiring the frequent replacement of the tubing andleading to poor reproducibility of data For this reason a pneumaticFIA-CL system consisting of five pressure lines equipped withprecision low-pressure regulators (maximum operating pressure02 MPa) was constructed as shown in Fig 1
The acid fumes and the heat generated from the reaction be-tween concentrated HCl and NaOH was a practical challenge inselecting the acid and base concentrations in this study Ad-ditionally the solubility of NaCl formed as a product of the neu-tralisation reaction of HCl and NaOH in isopropanol is very low(0013 g g1 at 235 degC) To avoid frequent regulator malfunctionand system blockage due to NaCl precipitation 5050 vv deio-nised waterndashisopropanol mixture was used to carry the 100 mLsample plug into the cheminert mixing chamber The system op-erated optimally without blockage from NaCl precipitation andsystem malfunction from acid fumes and excessive heat when 32HCl (concentrated) and 18 NaOH (wv) were used
32 Optimisation of the operating pressures and calculation of in-dividual reagent flow rates
The optimisation of reagent concentrations and optimum flowrates influencing degradation degree of OPEs and maximum re-sponse of chemiluminescent detection is not trivial task in mul-tichannel FIA system with pressure driven flows This is due todifficulties in control of flow rate in separate lines of pressuresystem when changes in backpressureflow rate in one linemay effect on flow rates in other lines Therefore the exact
concentration of used reagents should be measured in a separateexperiment
Initially the system was investigated to establish the optimuminput pressure at the regulators to provide the required flow rateswith the least variance The mass flow rates of deionised water(DIW) in all five lines were determined at 003 007 01 015 and018 MPa by precise measuring the weight losses in the containersafter 8 min operation The experiments were repeated for 5 con-secutive days The relative standard deviations (RSD) of the flowrates in five lines ranged from 25 to 51 for 04 MPa supplypressure whilst RSDs ranged from 15 to 30 for 06 MPa supplypressure As 06 MPa supply pressure resulted in the lower rangeof RSDs of flow rates for a working pressure range of 003ndash018 MPa in the FIA system 06 MPa supply pressure was used inall further experiments To understand the precision of reagentdelivery in the FIA system the relative standard deviations of themass flow rates of IPADIW 32 HCl 18 NaOH 05 mM Cu2thorn and088 mM luminol in five lines were calculated and presented inTable 1
In-depth investigations of flow rates in each line and workingpressures revealed that mass flow rates were linearly varied withthe working pressures in copper and luminol lines of Fig 1 The linemass flow rates remained almost constant in the NaOH line andvaried non-linearly with working pressures in acid and IPADIWline The variations of mass flow rates and working pressures areplotted in Figs S1ndashS5 in supplementary information As the FIAsystem was designed to work at very low Reynolds number (o20)and the temperature and all geometric characteristics of the tubingsremained constant throughout the experiment only viscosities of32 HCl (173 mPa s) 18 NaOH (278 mPa s) DIW water (1 mPa s)and isopropanol (237 mPa s) at 20deg C may have caused the non-linear nature of the mass flow rates To further confirm the non-linear nature of flow rates in the individual lines of the FIA systemshown in Fig 1 an experiment with an ion-chromatographic (IC)column was undertaken to separate and measure concentrations offive different transition metals added in the effluents of the FIA
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 1 FIA-CL network for acid degradation of HMTD TATP and MEKP
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 3
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
system by varying the pressure in one individual line linearly Theexperimental conditions of the IC column FIA system as well asvariations of flow rates applied pressure and corresponding metalconcentrations are detailed in the Supplementary information(Section 6 in SI Tables S5ndashS6 and Figs S6-S9)
Regarding selection of optimal flow rates in individual streamsthree separate reactions including decomposition of OPEs to pro-vide maximum yield of H2O2 neutralisation of acidic hydrolysateand chemiluminescent reaction detection of H2O2 with luminolwere considered in this work
321 Acid degradation reactionThe concentrated HCl was used to provide maximum H2O2 yield
in a short time The flow rates in IPADIW line and 32 HCl linewere varied to optimise the contact time between acid and samplein the reaction coil to achieve this goal The reaction times of aciddegradation were calculated between 215 and 138 min (Table 1) A215 min reaction time corresponded to maximum possible H2O2
yield from acid degradation of TATP and HMTD with HClIPA flowratio of 188 Hence line flow rates FLIPAfrac140043 g min1 and FLHClfrac14081 g min1 were chosen from Table 1 for the optimum acid de-gradation of HMTD and TATP However we observed that a contacttime 201 min with corresponding HCLIPA flow ratio of 32 wererequired for maximum possible H2O2 yield from acid degradation ofMEKP Hence line flow rates FLIPAfrac140046 g min1 andFLHClfrac14149 g min1 were chosen from Table 1 for the optimum aciddegradation of MEKP
322 Neutralisation of acidic hydrolysateAs it was found in this study that the flow rate in NaOH line
(ie FLNaOH) did not change with working pressure the flow ratein this line was kept the same as shown in Table 1 Additionallythe HCl flow rate could not exceed 149 g min1 as it causes the pHof the effluent at the outlet of FIA system to fall below 9 whichresulted poor chemiluminescent signal from the detector
323 Chemiluminescent reactiondetection of H2O2
For optimal chemiluminescent detection of H2O2 the flow ratesin copper and luminol lines (ie FLCu and FLLum) were chosen tocompromise between the reagent usage and Ruzickarsquos DispersionCoefficient (Dmax) As the chemiluminescent reaction betweenH2O2 and luminol is pH dependent FLCu and FLLum could not beindependently chosen to produce a minimum Dmax Table S2 insupplementary information illustrated that the chemiluminescentsignal was significantly reduced at pH 8 and was absent at pH 25However Dmax only varied between 7 and 875 at pH 102 and108 respectively As the corresponding FLCufrac14008 g min1 andFLLumfrac14 012 g min1 at pH 108 represented minimal reagentusage these two line flow rates were chosen for the optimumchemiluminescent reactiondetection of H2O2
The calculation of Dmax [21] for total flow rates at the outlet ofthe FIA system is described in Supplementary information (Table
S2) The calculations of optimum concentrations of Nathorn Cl Cu2thorn
and luminol during the chemiluminescent reaction are also pre-sented in Section 7 in Supplementary material
33 Acid degradation of HMTD TATP and MEKP
Samples containing different concentrations of TATP MEKP orHMTD were tested in the FIA-CL system The flow rates in theindividual lines were adjusted within the experimental range asdescribed in the previous section The measured line mass flowrates were then converted to volumetric flow rates by dividing bythe corresponding densities of the reagents at 20deg C Four sets ofexperimental line flow rates in Table 1 resulted total flow rates of162 mL min1 180 mL min1 200 mL min1 and 220 mL min1 atthe outlet The HClIPA flow ratios at these set ups were 188 3235 and 438 (Table 1) The corresponding contact times betweenthe sample plug and concentrated acid were calculated (see TableS3 in Supplementary material) Fig 2 shows detector responses for175 mM injection of MEKP 14 mM injection of TATP and 175 mMinjection of HMTD at contact times of 165 min 182 min 201 and215 min
From Fig 2 it can be seen that the detector responses for 14 mMTATP and 175 mM HMTD increased sharply at acid contact timesover 201 min To the contrary the detector response for 175 mMMEKP actually decreased with a contact time greater than201 min providing the maximum response at 201 min
In this study injections of equimolar concentrations of H2O2
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Table 1Experimental working pressures and corresponding measured mass flow rates in the FIA-CL system the mass of the reagents dispensed in 8 min from each bottle weremeasured and the experiments repeated for five days working pressures in the system were represented by PIPA PHCl PNaOH PCu and PLum flow rates in the system wererepresented by FLIPA FLHCl FLNaOH FLCu and FLLum
Operating pressure MPa Flow rate g min1 RSD Total FlowRateμL min1
Contact time be-tween acid andsample min
HClNaOHflow ratio
HClIPAflowratio
HCl concentrations inreactor coil ()
pH
PIPA PHCl PNaOH PCu PLum FLIPA FLHCl FLNaOH FLCu FLLum
002 007 002 004 002 0043 081 003 008 012 12ndash30 162 215 27 188 304 108003 009 003 006 003 0046 149 0031 01 023 23ndash35 180 201 48 32 311 102004 01 004 008 004 0051 179 003 013 044 09ndash15 200 182 60 35 3107 81005 012 005 01 004 0056 245 0032 015 006 16ndash28 220 165 77 438 3128 35006 015 006 012 007 0067 357 003 018 07 12ndash22 290 138 119 53 3145 25
Fig 2 Detector responses for 175 mM MEKP 175 mM HMTD and 14 mM TATP atdifferent contact times between OPEs and 32 HCl in the 1 m knitted reactor coil attotal flow rates 162 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac14188)180 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac1432) 200 μL min1 (cor-responding line flow ratio FLHCl FLIPAfrac1435) and 220 μL min1 (corresponding lineflow ratio FLHCl FLIPAfrac1443) Concentrations of HCl acid in the reactor coil corre-sponding to the flow rates are also shown
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎4
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Flow injection analysis of organic peroxide explosives using aciddegradation and chemiluminescent detection of released hydrogenperoxide
Parvez Mahbub a Philip Zakaria a Rosanne Guijt b Mirek Macka a Greg Dicinoski aMichael Breadmore a Pavel N Nesterenko anQ1
a Australian Centre for Research on Separation Science School of Physical Sciences University of Tasmania Australiab Pharmacy School of Medicine Australian Centre for Research on Separation Science University of Tasmania Australia
a r t i c l e i n f o
Article historyReceived 8 April 2015Received in revised form12 May 2015Accepted 22 May 2015
KeywordsOrganic peroxidesExplosivesFlow injection analysisAcid degradationChemiluminescence
a b s t r a c t
The applicability of acid degradation of organic peroxides into hydrogen peroxide in a pneumaticallydriven flow injection system with chemiluminescence reaction with luminol and Cu2thorn as a catalyst (FIA-CL) was investigated for the fast and sensitive detection of organic peroxide explosives (OPEs) The targetOPEs included hexamethylene triperoxide diamine (HMTD) triacetone triperoxide (TATP) and methy-lethyl ketone peroxide (MEKP) Under optimised conditions maximum degradations of 70 and 54 forTATP and HMTD respectively were achieved at 162 mL min1 and 9 degradation for MEKP at180 mL min1 Flow rates were precisely controlled in this single source pneumatic pressure drivenmulti-channel FIA system by model experiments on mixing of easily detectable component solutionsThe linear range for detection of TATP HMTD and H2O2 was 1ndash200 mM (r2frac14098ndash099) at both flow rateswhile that for MEKP was 20ndash200 mM (r2frac14097) at 180 mL min1 The detection limits (LODs) obtainedwere 05 mM for TATP HMTD and H2O2 and 10 mM for MEKP The detection times varied from 15 to 3 minin this FIA-CL system Whilst the LOD for H2O2 was comparable with those reported by other in-vestigators the LODs and analysis times for TATP and HMTD were superior and significantly this is thefirst time the detection of MEKP has been reported by FIA-CL
amp 2015 Published by Elsevier BV
1 Introduction
The simplicity of in-house preparation of the organic peroxideexplosives (OPEs) such as hexamethylene triperoxide diamine(HMTD) triacetone triperoxide (TATP) and methylethyl ketoneperoxide (MEKP) from readily available materials was the mainreason of their use in some recent terrorists attacks [1] So there isa strong demand for the development of fast simple and sensitivemethods of their identification and quantitative determination [2]This task is not trivial because of high volatility and absence ofchromophoric groups in the molecules of OPEs For these reasonsthe use of common analytical methods such as GC or HPLC withUVvisible detection is not readily suitable for their direct de-termination so the application of more complex hyphenatedtechniques typically involving mass spectrometry (MS) is re-quired Schulte-Ladbeck et al [3] proposed RP HPLC with on-line
Fourier transfom infrared (FTIR) detection for direct determinationof TATP and HMTD De Tata et al [4] reported the application of RPHPLC with quadrupole time-of-flight mass spectrometry (HPLC-QToF-MS) for direct determination of various OPEs However mi-cellar electrokinetic chromatography with UV detection was em-ployed by Johns et al [5] recently for separation of OPEs includingHMTD and TATP in post blast scenario without any marked im-provement compared to the hyphenated HPLC methods in termsof sensitivity
Alternatively the determination of OPEs can be based onelectrochemical [6] fluorescent [7] and chemiluminescent [8]detection of hydrogen peroxide as the main degradation productof OPEs It should be noted that decomposition of one OPE mole-cule can result in more than one molecule of H2O2 so in case of100 degradation a magnified analytical response can be expectedBecause of its robustness sensitivity and simplicity of integrationwith FIA chemiluminescent detection (FIA-CL) is one of the mostpopular techniques for the on-line detection of hydrogen peroxideThe application of FIA-CL has been reported for the determinationof H2O2 in natural waters [9] rainwater [10] and seawater [11] Theuse of chemiluminescent detection of H2O2 after decomposition of
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
6768697071727374757677787980818283848586878889909192
Contents lists available at ScienceDirect
journal homepage wwwelseviercomlocatetalanta
Talanta
httpdxdoiorg101016jtalanta2015050520039-9140amp 2015 Published by Elsevier BV
n Correspondence to Australian Centre for Research on Separation ScienceSchool of Chemistry University of Tasmania Private Bag 75 Hobart 7001 AustraliaFax thorn61 03 6226 2858
E-mail address PavelNesterenkoutaseduau (PN Nesterenko)
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎
TATP and HMTD was reported for determination of these ex-plosives by a number of research groups [3612] The analysistimes reported varied from 5 to 12 min with the sensitivity ran-ging from 25 mM for TATP [6] to 500 mM for HMTD [3]
Obviously the crucial parameters in the flow-through methodsof analysis of explosives via detection of hydrogen peroxide are thevelocity and conversion degree of OPEs into H2O2 According to theliterature data the degradation of peroxide explosives can be ac-complished enzymatically [1213] photochemically [14] or byusing mineral acids [15] The latter option appears more simpleand robust but it takes almost 12 h for complete decomposition ofTATP and HMTD which is not suitable for rapid screening Ad-ditionally there has been no report of the decomposition of me-thylethyl ketone peroxide (MEKP)-listed as a priority explosive byseveral government agencies including US National Counter-terrorism Centre and the Australian Army [16] It should be notedthat 100 degradation of OPEs does not mean the maximumpossible concentration of hydrogen peroxide suitable for detectionas hydrolysis of H2O2 occurs together with organic peroxides Onthis reason acid degradation conditions of OPEs should be care-fully attuned to provide fast and sensitive response in FIA system
This study was a part of a larger project directed on construc-tion of portable automated system for the fast screening of sam-ples on presence of traces of homemade explosives including or-ganic peroxides and inorganic explosives According to the pro-posed design the whole system composed of three separate unitsnamely sample extraction unit FIA-CL detection unit for organicperoxides and capillary zone electrophoresis unit for parallelprofiling of inorganic anions contents The high degree of auto-mation minimal consumption of reagents and robust operationfor extended period of time were considered as important re-quirements for this system The system should provide eitherpositive or negative answer in a short time on the presence ofexplosives at low concentration level
Typically multicomponent FIA analysers use multi-channelperistaltic pumps for delivery of sample and reagents solutionsAccording to Dasgupta to get very low flow rates at mLmin levelas required in this study peristaltic pumps are of little help [17]Piezoelectric pumps pneumatic pressure or gravity driven systemsare often preferred choice in this case [18] Pressure driven FIAinstruments have advantages over systems using peristaltic pumpsin terms of lower signalnoise ratio better reproducibility of tim-ing and expanded possibilities for the use of aggressive solventsand carriers [18] According to Valcarcel and De Castro the maindrawback of pneumatic driven instruments is connected withdifficulty in accurate control of the flow rates in multiple channelFIA systems due to complex changes in hydraulic resistances re-sulting from different channel geometry and reagent viscosity [19]In this study a special attention was paid to the development of acontrol mechanism of flow rates in separate lines of multichannelFIA system driven by pneumatic pressure from a single source
The objective of this study is to optimise conditions for aciddegradation of organic peroxides and subsequent chemilumines-cent luminol based detection of hydrogen peroxide within pres-sure driven FIA-CL analytical unit for the purpose of fast qualita-tive determination of HMTD TATP and MEKP The target totalanalysis time was less than 2 min for rapid screening and high-throughput analysis imposing a significant challenge to the opti-misation of the flow conditions
2 Experimental section
21 Reagents and chemical standards
Hydrated copper sulphate (CuSO45H2O) 32 concentrated
hydrochloric acid (HCl) 30 (ww) reagent grade H2O2 iso-propanol and sodium hydroxide (NaOH) pellets were purchasedfrom Sigma-Aldrich (Sydney Australia) Luminol was purchasedfrom Fluka (Sydney Australia) Element free deionised water wasused to prepare all stock and working solutions The TATP standard(10000 mg L1 999 single component) and HMTD standard(5000 mg L1 984 single component) were procured from Ac-custandard USA The MEKP standard (10000 mg L1) was sup-plied by the Australian Defence Science and Technology Organi-sation (DSTO)
22 Preparation of precise assay of the standard solution for H2O2
A stock solution of approximately 1000 mg L1 of the 30 H2O2
was prepared through serial dilution Then 50 mL of stock solutionwas transferred into a 500 mL conical flask diluted with 200 mL ofdeionised water and then 30 mL of 25 sulphuric acid was addedThe solution was titrated with a standard 002 M potassium per-manganate solution until the colour changed to pink The workingsolutions were prepared by further dilution of stock solution in DIW
23 Instrumentation
A FIA-Cl system consisting of a low pressure Cheminert 6 port2 position injector valve (C22-3186EH-FL VICI Houston USA) fiveSMC precision pressure regulators (IR 1000-01 SMC Japan) and aHamamatsu photomultiplier (10493-001 Hamamatsu Japan) wereused A schematic of the instrument used for acid degradationstudy of TATP HMTD and MEKP is illustrated in Fig 1
The system comprised SMC pressure regulators connectedthrough a manifold to an external compressed air supply Eachregulator was used to control the reagent solution flow from a500 mL glass bottle (Schott AG Sigma-Aldrich Australia) by ap-plying pressures ranging from 001 to 02 MPa (15ndash29 psi) Thepneumatic lines consisted of polyurethane tubing (25 mm IDSMC Japan) and the hydraulic lines consisted of FEP tubing(0203 mm ID Upchurch USA)
The 100 mL sample plug was carried into the Cheminert mixer(CM1XKF VICI Houston USA) in 5050 vv deionised waterndashiso-propanol where it was mixed with 32 HCl Isopropanol was usedto ensure complete dissolution of OPEs from the collected andextracted samples (the exact procedure is not included in thispaper) The OPE containing samples were degraded in acidicmixture in a 1 m PTFE knitted tubing coil reactor (025 mm ID491 mL internal volume Biotech AB Onsala Sweden) resulting inthe release of H2O2 molecules The excess HCl was neutralised bythe addition of 18 NaOH and the resulting solution was thenmixed with the luminol ndash Cu2thorn reagent in a tee mixer (P-71229 mL swept volume Upchurch Oak Harbour USA) The sub-sequent chemiluminescence reaction was detected with the pho-tomultiplier tube The chemiluminescence flow cell was fabricatedin-house using FEP tubing (0508 mm ID 1548 L Upchurch USA)with the total volume of the flow cell being 8532 mL The chemi-luminescence signal was acquired by a Powerchrom data acquisi-tion system (ER280 EDaq Sydney Australia) with proprietarysoftware version 81 The pH of the effluent was monitored by anin-line pH monitoring flowcell (Cole-Parmer Australia verticalflow glass electrode 50 mL internal volume) as shown in Fig 1
3 Results and discussions
31 Optimisation of reagent concentrations
The crucial part of FIA-CL under development is acid degrada-tion of OPEs in isopropanolndashwater (5050) extracts of swabs used
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎2
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
for sample collection from different surfaces It was shown thatsuch mixture provides the best extraction of various types of in-organic and organic explosives [20] However as a result of ele-vated viscosity of this mixture the FIA system in this study need tooperate at pressures higher than those normally provided byperistaltic pumps Also initial experiments using peristaltic pumpsdemonstrated poor stability of the silicone tubing when in contactwith solutions of isopropanol hydrochloric acid and sodium hy-droxide requiring the frequent replacement of the tubing andleading to poor reproducibility of data For this reason a pneumaticFIA-CL system consisting of five pressure lines equipped withprecision low-pressure regulators (maximum operating pressure02 MPa) was constructed as shown in Fig 1
The acid fumes and the heat generated from the reaction be-tween concentrated HCl and NaOH was a practical challenge inselecting the acid and base concentrations in this study Ad-ditionally the solubility of NaCl formed as a product of the neu-tralisation reaction of HCl and NaOH in isopropanol is very low(0013 g g1 at 235 degC) To avoid frequent regulator malfunctionand system blockage due to NaCl precipitation 5050 vv deio-nised waterndashisopropanol mixture was used to carry the 100 mLsample plug into the cheminert mixing chamber The system op-erated optimally without blockage from NaCl precipitation andsystem malfunction from acid fumes and excessive heat when 32HCl (concentrated) and 18 NaOH (wv) were used
32 Optimisation of the operating pressures and calculation of in-dividual reagent flow rates
The optimisation of reagent concentrations and optimum flowrates influencing degradation degree of OPEs and maximum re-sponse of chemiluminescent detection is not trivial task in mul-tichannel FIA system with pressure driven flows This is due todifficulties in control of flow rate in separate lines of pressuresystem when changes in backpressureflow rate in one linemay effect on flow rates in other lines Therefore the exact
concentration of used reagents should be measured in a separateexperiment
Initially the system was investigated to establish the optimuminput pressure at the regulators to provide the required flow rateswith the least variance The mass flow rates of deionised water(DIW) in all five lines were determined at 003 007 01 015 and018 MPa by precise measuring the weight losses in the containersafter 8 min operation The experiments were repeated for 5 con-secutive days The relative standard deviations (RSD) of the flowrates in five lines ranged from 25 to 51 for 04 MPa supplypressure whilst RSDs ranged from 15 to 30 for 06 MPa supplypressure As 06 MPa supply pressure resulted in the lower rangeof RSDs of flow rates for a working pressure range of 003ndash018 MPa in the FIA system 06 MPa supply pressure was used inall further experiments To understand the precision of reagentdelivery in the FIA system the relative standard deviations of themass flow rates of IPADIW 32 HCl 18 NaOH 05 mM Cu2thorn and088 mM luminol in five lines were calculated and presented inTable 1
In-depth investigations of flow rates in each line and workingpressures revealed that mass flow rates were linearly varied withthe working pressures in copper and luminol lines of Fig 1 The linemass flow rates remained almost constant in the NaOH line andvaried non-linearly with working pressures in acid and IPADIWline The variations of mass flow rates and working pressures areplotted in Figs S1ndashS5 in supplementary information As the FIAsystem was designed to work at very low Reynolds number (o20)and the temperature and all geometric characteristics of the tubingsremained constant throughout the experiment only viscosities of32 HCl (173 mPa s) 18 NaOH (278 mPa s) DIW water (1 mPa s)and isopropanol (237 mPa s) at 20deg C may have caused the non-linear nature of the mass flow rates To further confirm the non-linear nature of flow rates in the individual lines of the FIA systemshown in Fig 1 an experiment with an ion-chromatographic (IC)column was undertaken to separate and measure concentrations offive different transition metals added in the effluents of the FIA
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 1 FIA-CL network for acid degradation of HMTD TATP and MEKP
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 3
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
system by varying the pressure in one individual line linearly Theexperimental conditions of the IC column FIA system as well asvariations of flow rates applied pressure and corresponding metalconcentrations are detailed in the Supplementary information(Section 6 in SI Tables S5ndashS6 and Figs S6-S9)
Regarding selection of optimal flow rates in individual streamsthree separate reactions including decomposition of OPEs to pro-vide maximum yield of H2O2 neutralisation of acidic hydrolysateand chemiluminescent reaction detection of H2O2 with luminolwere considered in this work
321 Acid degradation reactionThe concentrated HCl was used to provide maximum H2O2 yield
in a short time The flow rates in IPADIW line and 32 HCl linewere varied to optimise the contact time between acid and samplein the reaction coil to achieve this goal The reaction times of aciddegradation were calculated between 215 and 138 min (Table 1) A215 min reaction time corresponded to maximum possible H2O2
yield from acid degradation of TATP and HMTD with HClIPA flowratio of 188 Hence line flow rates FLIPAfrac140043 g min1 and FLHClfrac14081 g min1 were chosen from Table 1 for the optimum acid de-gradation of HMTD and TATP However we observed that a contacttime 201 min with corresponding HCLIPA flow ratio of 32 wererequired for maximum possible H2O2 yield from acid degradation ofMEKP Hence line flow rates FLIPAfrac140046 g min1 andFLHClfrac14149 g min1 were chosen from Table 1 for the optimum aciddegradation of MEKP
322 Neutralisation of acidic hydrolysateAs it was found in this study that the flow rate in NaOH line
(ie FLNaOH) did not change with working pressure the flow ratein this line was kept the same as shown in Table 1 Additionallythe HCl flow rate could not exceed 149 g min1 as it causes the pHof the effluent at the outlet of FIA system to fall below 9 whichresulted poor chemiluminescent signal from the detector
323 Chemiluminescent reactiondetection of H2O2
For optimal chemiluminescent detection of H2O2 the flow ratesin copper and luminol lines (ie FLCu and FLLum) were chosen tocompromise between the reagent usage and Ruzickarsquos DispersionCoefficient (Dmax) As the chemiluminescent reaction betweenH2O2 and luminol is pH dependent FLCu and FLLum could not beindependently chosen to produce a minimum Dmax Table S2 insupplementary information illustrated that the chemiluminescentsignal was significantly reduced at pH 8 and was absent at pH 25However Dmax only varied between 7 and 875 at pH 102 and108 respectively As the corresponding FLCufrac14008 g min1 andFLLumfrac14 012 g min1 at pH 108 represented minimal reagentusage these two line flow rates were chosen for the optimumchemiluminescent reactiondetection of H2O2
The calculation of Dmax [21] for total flow rates at the outlet ofthe FIA system is described in Supplementary information (Table
S2) The calculations of optimum concentrations of Nathorn Cl Cu2thorn
and luminol during the chemiluminescent reaction are also pre-sented in Section 7 in Supplementary material
33 Acid degradation of HMTD TATP and MEKP
Samples containing different concentrations of TATP MEKP orHMTD were tested in the FIA-CL system The flow rates in theindividual lines were adjusted within the experimental range asdescribed in the previous section The measured line mass flowrates were then converted to volumetric flow rates by dividing bythe corresponding densities of the reagents at 20deg C Four sets ofexperimental line flow rates in Table 1 resulted total flow rates of162 mL min1 180 mL min1 200 mL min1 and 220 mL min1 atthe outlet The HClIPA flow ratios at these set ups were 188 3235 and 438 (Table 1) The corresponding contact times betweenthe sample plug and concentrated acid were calculated (see TableS3 in Supplementary material) Fig 2 shows detector responses for175 mM injection of MEKP 14 mM injection of TATP and 175 mMinjection of HMTD at contact times of 165 min 182 min 201 and215 min
From Fig 2 it can be seen that the detector responses for 14 mMTATP and 175 mM HMTD increased sharply at acid contact timesover 201 min To the contrary the detector response for 175 mMMEKP actually decreased with a contact time greater than201 min providing the maximum response at 201 min
In this study injections of equimolar concentrations of H2O2
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Table 1Experimental working pressures and corresponding measured mass flow rates in the FIA-CL system the mass of the reagents dispensed in 8 min from each bottle weremeasured and the experiments repeated for five days working pressures in the system were represented by PIPA PHCl PNaOH PCu and PLum flow rates in the system wererepresented by FLIPA FLHCl FLNaOH FLCu and FLLum
Operating pressure MPa Flow rate g min1 RSD Total FlowRateμL min1
Contact time be-tween acid andsample min
HClNaOHflow ratio
HClIPAflowratio
HCl concentrations inreactor coil ()
pH
PIPA PHCl PNaOH PCu PLum FLIPA FLHCl FLNaOH FLCu FLLum
002 007 002 004 002 0043 081 003 008 012 12ndash30 162 215 27 188 304 108003 009 003 006 003 0046 149 0031 01 023 23ndash35 180 201 48 32 311 102004 01 004 008 004 0051 179 003 013 044 09ndash15 200 182 60 35 3107 81005 012 005 01 004 0056 245 0032 015 006 16ndash28 220 165 77 438 3128 35006 015 006 012 007 0067 357 003 018 07 12ndash22 290 138 119 53 3145 25
Fig 2 Detector responses for 175 mM MEKP 175 mM HMTD and 14 mM TATP atdifferent contact times between OPEs and 32 HCl in the 1 m knitted reactor coil attotal flow rates 162 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac14188)180 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac1432) 200 μL min1 (cor-responding line flow ratio FLHCl FLIPAfrac1435) and 220 μL min1 (corresponding lineflow ratio FLHCl FLIPAfrac1443) Concentrations of HCl acid in the reactor coil corre-sponding to the flow rates are also shown
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎4
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
TATP and HMTD was reported for determination of these ex-plosives by a number of research groups [3612] The analysistimes reported varied from 5 to 12 min with the sensitivity ran-ging from 25 mM for TATP [6] to 500 mM for HMTD [3]
Obviously the crucial parameters in the flow-through methodsof analysis of explosives via detection of hydrogen peroxide are thevelocity and conversion degree of OPEs into H2O2 According to theliterature data the degradation of peroxide explosives can be ac-complished enzymatically [1213] photochemically [14] or byusing mineral acids [15] The latter option appears more simpleand robust but it takes almost 12 h for complete decomposition ofTATP and HMTD which is not suitable for rapid screening Ad-ditionally there has been no report of the decomposition of me-thylethyl ketone peroxide (MEKP)-listed as a priority explosive byseveral government agencies including US National Counter-terrorism Centre and the Australian Army [16] It should be notedthat 100 degradation of OPEs does not mean the maximumpossible concentration of hydrogen peroxide suitable for detectionas hydrolysis of H2O2 occurs together with organic peroxides Onthis reason acid degradation conditions of OPEs should be care-fully attuned to provide fast and sensitive response in FIA system
This study was a part of a larger project directed on construc-tion of portable automated system for the fast screening of sam-ples on presence of traces of homemade explosives including or-ganic peroxides and inorganic explosives According to the pro-posed design the whole system composed of three separate unitsnamely sample extraction unit FIA-CL detection unit for organicperoxides and capillary zone electrophoresis unit for parallelprofiling of inorganic anions contents The high degree of auto-mation minimal consumption of reagents and robust operationfor extended period of time were considered as important re-quirements for this system The system should provide eitherpositive or negative answer in a short time on the presence ofexplosives at low concentration level
Typically multicomponent FIA analysers use multi-channelperistaltic pumps for delivery of sample and reagents solutionsAccording to Dasgupta to get very low flow rates at mLmin levelas required in this study peristaltic pumps are of little help [17]Piezoelectric pumps pneumatic pressure or gravity driven systemsare often preferred choice in this case [18] Pressure driven FIAinstruments have advantages over systems using peristaltic pumpsin terms of lower signalnoise ratio better reproducibility of tim-ing and expanded possibilities for the use of aggressive solventsand carriers [18] According to Valcarcel and De Castro the maindrawback of pneumatic driven instruments is connected withdifficulty in accurate control of the flow rates in multiple channelFIA systems due to complex changes in hydraulic resistances re-sulting from different channel geometry and reagent viscosity [19]In this study a special attention was paid to the development of acontrol mechanism of flow rates in separate lines of multichannelFIA system driven by pneumatic pressure from a single source
The objective of this study is to optimise conditions for aciddegradation of organic peroxides and subsequent chemilumines-cent luminol based detection of hydrogen peroxide within pres-sure driven FIA-CL analytical unit for the purpose of fast qualita-tive determination of HMTD TATP and MEKP The target totalanalysis time was less than 2 min for rapid screening and high-throughput analysis imposing a significant challenge to the opti-misation of the flow conditions
2 Experimental section
21 Reagents and chemical standards
Hydrated copper sulphate (CuSO45H2O) 32 concentrated
hydrochloric acid (HCl) 30 (ww) reagent grade H2O2 iso-propanol and sodium hydroxide (NaOH) pellets were purchasedfrom Sigma-Aldrich (Sydney Australia) Luminol was purchasedfrom Fluka (Sydney Australia) Element free deionised water wasused to prepare all stock and working solutions The TATP standard(10000 mg L1 999 single component) and HMTD standard(5000 mg L1 984 single component) were procured from Ac-custandard USA The MEKP standard (10000 mg L1) was sup-plied by the Australian Defence Science and Technology Organi-sation (DSTO)
22 Preparation of precise assay of the standard solution for H2O2
A stock solution of approximately 1000 mg L1 of the 30 H2O2
was prepared through serial dilution Then 50 mL of stock solutionwas transferred into a 500 mL conical flask diluted with 200 mL ofdeionised water and then 30 mL of 25 sulphuric acid was addedThe solution was titrated with a standard 002 M potassium per-manganate solution until the colour changed to pink The workingsolutions were prepared by further dilution of stock solution in DIW
23 Instrumentation
A FIA-Cl system consisting of a low pressure Cheminert 6 port2 position injector valve (C22-3186EH-FL VICI Houston USA) fiveSMC precision pressure regulators (IR 1000-01 SMC Japan) and aHamamatsu photomultiplier (10493-001 Hamamatsu Japan) wereused A schematic of the instrument used for acid degradationstudy of TATP HMTD and MEKP is illustrated in Fig 1
The system comprised SMC pressure regulators connectedthrough a manifold to an external compressed air supply Eachregulator was used to control the reagent solution flow from a500 mL glass bottle (Schott AG Sigma-Aldrich Australia) by ap-plying pressures ranging from 001 to 02 MPa (15ndash29 psi) Thepneumatic lines consisted of polyurethane tubing (25 mm IDSMC Japan) and the hydraulic lines consisted of FEP tubing(0203 mm ID Upchurch USA)
The 100 mL sample plug was carried into the Cheminert mixer(CM1XKF VICI Houston USA) in 5050 vv deionised waterndashiso-propanol where it was mixed with 32 HCl Isopropanol was usedto ensure complete dissolution of OPEs from the collected andextracted samples (the exact procedure is not included in thispaper) The OPE containing samples were degraded in acidicmixture in a 1 m PTFE knitted tubing coil reactor (025 mm ID491 mL internal volume Biotech AB Onsala Sweden) resulting inthe release of H2O2 molecules The excess HCl was neutralised bythe addition of 18 NaOH and the resulting solution was thenmixed with the luminol ndash Cu2thorn reagent in a tee mixer (P-71229 mL swept volume Upchurch Oak Harbour USA) The sub-sequent chemiluminescence reaction was detected with the pho-tomultiplier tube The chemiluminescence flow cell was fabricatedin-house using FEP tubing (0508 mm ID 1548 L Upchurch USA)with the total volume of the flow cell being 8532 mL The chemi-luminescence signal was acquired by a Powerchrom data acquisi-tion system (ER280 EDaq Sydney Australia) with proprietarysoftware version 81 The pH of the effluent was monitored by anin-line pH monitoring flowcell (Cole-Parmer Australia verticalflow glass electrode 50 mL internal volume) as shown in Fig 1
3 Results and discussions
31 Optimisation of reagent concentrations
The crucial part of FIA-CL under development is acid degrada-tion of OPEs in isopropanolndashwater (5050) extracts of swabs used
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎2
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
for sample collection from different surfaces It was shown thatsuch mixture provides the best extraction of various types of in-organic and organic explosives [20] However as a result of ele-vated viscosity of this mixture the FIA system in this study need tooperate at pressures higher than those normally provided byperistaltic pumps Also initial experiments using peristaltic pumpsdemonstrated poor stability of the silicone tubing when in contactwith solutions of isopropanol hydrochloric acid and sodium hy-droxide requiring the frequent replacement of the tubing andleading to poor reproducibility of data For this reason a pneumaticFIA-CL system consisting of five pressure lines equipped withprecision low-pressure regulators (maximum operating pressure02 MPa) was constructed as shown in Fig 1
The acid fumes and the heat generated from the reaction be-tween concentrated HCl and NaOH was a practical challenge inselecting the acid and base concentrations in this study Ad-ditionally the solubility of NaCl formed as a product of the neu-tralisation reaction of HCl and NaOH in isopropanol is very low(0013 g g1 at 235 degC) To avoid frequent regulator malfunctionand system blockage due to NaCl precipitation 5050 vv deio-nised waterndashisopropanol mixture was used to carry the 100 mLsample plug into the cheminert mixing chamber The system op-erated optimally without blockage from NaCl precipitation andsystem malfunction from acid fumes and excessive heat when 32HCl (concentrated) and 18 NaOH (wv) were used
32 Optimisation of the operating pressures and calculation of in-dividual reagent flow rates
The optimisation of reagent concentrations and optimum flowrates influencing degradation degree of OPEs and maximum re-sponse of chemiluminescent detection is not trivial task in mul-tichannel FIA system with pressure driven flows This is due todifficulties in control of flow rate in separate lines of pressuresystem when changes in backpressureflow rate in one linemay effect on flow rates in other lines Therefore the exact
concentration of used reagents should be measured in a separateexperiment
Initially the system was investigated to establish the optimuminput pressure at the regulators to provide the required flow rateswith the least variance The mass flow rates of deionised water(DIW) in all five lines were determined at 003 007 01 015 and018 MPa by precise measuring the weight losses in the containersafter 8 min operation The experiments were repeated for 5 con-secutive days The relative standard deviations (RSD) of the flowrates in five lines ranged from 25 to 51 for 04 MPa supplypressure whilst RSDs ranged from 15 to 30 for 06 MPa supplypressure As 06 MPa supply pressure resulted in the lower rangeof RSDs of flow rates for a working pressure range of 003ndash018 MPa in the FIA system 06 MPa supply pressure was used inall further experiments To understand the precision of reagentdelivery in the FIA system the relative standard deviations of themass flow rates of IPADIW 32 HCl 18 NaOH 05 mM Cu2thorn and088 mM luminol in five lines were calculated and presented inTable 1
In-depth investigations of flow rates in each line and workingpressures revealed that mass flow rates were linearly varied withthe working pressures in copper and luminol lines of Fig 1 The linemass flow rates remained almost constant in the NaOH line andvaried non-linearly with working pressures in acid and IPADIWline The variations of mass flow rates and working pressures areplotted in Figs S1ndashS5 in supplementary information As the FIAsystem was designed to work at very low Reynolds number (o20)and the temperature and all geometric characteristics of the tubingsremained constant throughout the experiment only viscosities of32 HCl (173 mPa s) 18 NaOH (278 mPa s) DIW water (1 mPa s)and isopropanol (237 mPa s) at 20deg C may have caused the non-linear nature of the mass flow rates To further confirm the non-linear nature of flow rates in the individual lines of the FIA systemshown in Fig 1 an experiment with an ion-chromatographic (IC)column was undertaken to separate and measure concentrations offive different transition metals added in the effluents of the FIA
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 1 FIA-CL network for acid degradation of HMTD TATP and MEKP
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 3
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
system by varying the pressure in one individual line linearly Theexperimental conditions of the IC column FIA system as well asvariations of flow rates applied pressure and corresponding metalconcentrations are detailed in the Supplementary information(Section 6 in SI Tables S5ndashS6 and Figs S6-S9)
Regarding selection of optimal flow rates in individual streamsthree separate reactions including decomposition of OPEs to pro-vide maximum yield of H2O2 neutralisation of acidic hydrolysateand chemiluminescent reaction detection of H2O2 with luminolwere considered in this work
321 Acid degradation reactionThe concentrated HCl was used to provide maximum H2O2 yield
in a short time The flow rates in IPADIW line and 32 HCl linewere varied to optimise the contact time between acid and samplein the reaction coil to achieve this goal The reaction times of aciddegradation were calculated between 215 and 138 min (Table 1) A215 min reaction time corresponded to maximum possible H2O2
yield from acid degradation of TATP and HMTD with HClIPA flowratio of 188 Hence line flow rates FLIPAfrac140043 g min1 and FLHClfrac14081 g min1 were chosen from Table 1 for the optimum acid de-gradation of HMTD and TATP However we observed that a contacttime 201 min with corresponding HCLIPA flow ratio of 32 wererequired for maximum possible H2O2 yield from acid degradation ofMEKP Hence line flow rates FLIPAfrac140046 g min1 andFLHClfrac14149 g min1 were chosen from Table 1 for the optimum aciddegradation of MEKP
322 Neutralisation of acidic hydrolysateAs it was found in this study that the flow rate in NaOH line
(ie FLNaOH) did not change with working pressure the flow ratein this line was kept the same as shown in Table 1 Additionallythe HCl flow rate could not exceed 149 g min1 as it causes the pHof the effluent at the outlet of FIA system to fall below 9 whichresulted poor chemiluminescent signal from the detector
323 Chemiluminescent reactiondetection of H2O2
For optimal chemiluminescent detection of H2O2 the flow ratesin copper and luminol lines (ie FLCu and FLLum) were chosen tocompromise between the reagent usage and Ruzickarsquos DispersionCoefficient (Dmax) As the chemiluminescent reaction betweenH2O2 and luminol is pH dependent FLCu and FLLum could not beindependently chosen to produce a minimum Dmax Table S2 insupplementary information illustrated that the chemiluminescentsignal was significantly reduced at pH 8 and was absent at pH 25However Dmax only varied between 7 and 875 at pH 102 and108 respectively As the corresponding FLCufrac14008 g min1 andFLLumfrac14 012 g min1 at pH 108 represented minimal reagentusage these two line flow rates were chosen for the optimumchemiluminescent reactiondetection of H2O2
The calculation of Dmax [21] for total flow rates at the outlet ofthe FIA system is described in Supplementary information (Table
S2) The calculations of optimum concentrations of Nathorn Cl Cu2thorn
and luminol during the chemiluminescent reaction are also pre-sented in Section 7 in Supplementary material
33 Acid degradation of HMTD TATP and MEKP
Samples containing different concentrations of TATP MEKP orHMTD were tested in the FIA-CL system The flow rates in theindividual lines were adjusted within the experimental range asdescribed in the previous section The measured line mass flowrates were then converted to volumetric flow rates by dividing bythe corresponding densities of the reagents at 20deg C Four sets ofexperimental line flow rates in Table 1 resulted total flow rates of162 mL min1 180 mL min1 200 mL min1 and 220 mL min1 atthe outlet The HClIPA flow ratios at these set ups were 188 3235 and 438 (Table 1) The corresponding contact times betweenthe sample plug and concentrated acid were calculated (see TableS3 in Supplementary material) Fig 2 shows detector responses for175 mM injection of MEKP 14 mM injection of TATP and 175 mMinjection of HMTD at contact times of 165 min 182 min 201 and215 min
From Fig 2 it can be seen that the detector responses for 14 mMTATP and 175 mM HMTD increased sharply at acid contact timesover 201 min To the contrary the detector response for 175 mMMEKP actually decreased with a contact time greater than201 min providing the maximum response at 201 min
In this study injections of equimolar concentrations of H2O2
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Table 1Experimental working pressures and corresponding measured mass flow rates in the FIA-CL system the mass of the reagents dispensed in 8 min from each bottle weremeasured and the experiments repeated for five days working pressures in the system were represented by PIPA PHCl PNaOH PCu and PLum flow rates in the system wererepresented by FLIPA FLHCl FLNaOH FLCu and FLLum
Operating pressure MPa Flow rate g min1 RSD Total FlowRateμL min1
Contact time be-tween acid andsample min
HClNaOHflow ratio
HClIPAflowratio
HCl concentrations inreactor coil ()
pH
PIPA PHCl PNaOH PCu PLum FLIPA FLHCl FLNaOH FLCu FLLum
002 007 002 004 002 0043 081 003 008 012 12ndash30 162 215 27 188 304 108003 009 003 006 003 0046 149 0031 01 023 23ndash35 180 201 48 32 311 102004 01 004 008 004 0051 179 003 013 044 09ndash15 200 182 60 35 3107 81005 012 005 01 004 0056 245 0032 015 006 16ndash28 220 165 77 438 3128 35006 015 006 012 007 0067 357 003 018 07 12ndash22 290 138 119 53 3145 25
Fig 2 Detector responses for 175 mM MEKP 175 mM HMTD and 14 mM TATP atdifferent contact times between OPEs and 32 HCl in the 1 m knitted reactor coil attotal flow rates 162 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac14188)180 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac1432) 200 μL min1 (cor-responding line flow ratio FLHCl FLIPAfrac1435) and 220 μL min1 (corresponding lineflow ratio FLHCl FLIPAfrac1443) Concentrations of HCl acid in the reactor coil corre-sponding to the flow rates are also shown
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎4
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
for sample collection from different surfaces It was shown thatsuch mixture provides the best extraction of various types of in-organic and organic explosives [20] However as a result of ele-vated viscosity of this mixture the FIA system in this study need tooperate at pressures higher than those normally provided byperistaltic pumps Also initial experiments using peristaltic pumpsdemonstrated poor stability of the silicone tubing when in contactwith solutions of isopropanol hydrochloric acid and sodium hy-droxide requiring the frequent replacement of the tubing andleading to poor reproducibility of data For this reason a pneumaticFIA-CL system consisting of five pressure lines equipped withprecision low-pressure regulators (maximum operating pressure02 MPa) was constructed as shown in Fig 1
The acid fumes and the heat generated from the reaction be-tween concentrated HCl and NaOH was a practical challenge inselecting the acid and base concentrations in this study Ad-ditionally the solubility of NaCl formed as a product of the neu-tralisation reaction of HCl and NaOH in isopropanol is very low(0013 g g1 at 235 degC) To avoid frequent regulator malfunctionand system blockage due to NaCl precipitation 5050 vv deio-nised waterndashisopropanol mixture was used to carry the 100 mLsample plug into the cheminert mixing chamber The system op-erated optimally without blockage from NaCl precipitation andsystem malfunction from acid fumes and excessive heat when 32HCl (concentrated) and 18 NaOH (wv) were used
32 Optimisation of the operating pressures and calculation of in-dividual reagent flow rates
The optimisation of reagent concentrations and optimum flowrates influencing degradation degree of OPEs and maximum re-sponse of chemiluminescent detection is not trivial task in mul-tichannel FIA system with pressure driven flows This is due todifficulties in control of flow rate in separate lines of pressuresystem when changes in backpressureflow rate in one linemay effect on flow rates in other lines Therefore the exact
concentration of used reagents should be measured in a separateexperiment
Initially the system was investigated to establish the optimuminput pressure at the regulators to provide the required flow rateswith the least variance The mass flow rates of deionised water(DIW) in all five lines were determined at 003 007 01 015 and018 MPa by precise measuring the weight losses in the containersafter 8 min operation The experiments were repeated for 5 con-secutive days The relative standard deviations (RSD) of the flowrates in five lines ranged from 25 to 51 for 04 MPa supplypressure whilst RSDs ranged from 15 to 30 for 06 MPa supplypressure As 06 MPa supply pressure resulted in the lower rangeof RSDs of flow rates for a working pressure range of 003ndash018 MPa in the FIA system 06 MPa supply pressure was used inall further experiments To understand the precision of reagentdelivery in the FIA system the relative standard deviations of themass flow rates of IPADIW 32 HCl 18 NaOH 05 mM Cu2thorn and088 mM luminol in five lines were calculated and presented inTable 1
In-depth investigations of flow rates in each line and workingpressures revealed that mass flow rates were linearly varied withthe working pressures in copper and luminol lines of Fig 1 The linemass flow rates remained almost constant in the NaOH line andvaried non-linearly with working pressures in acid and IPADIWline The variations of mass flow rates and working pressures areplotted in Figs S1ndashS5 in supplementary information As the FIAsystem was designed to work at very low Reynolds number (o20)and the temperature and all geometric characteristics of the tubingsremained constant throughout the experiment only viscosities of32 HCl (173 mPa s) 18 NaOH (278 mPa s) DIW water (1 mPa s)and isopropanol (237 mPa s) at 20deg C may have caused the non-linear nature of the mass flow rates To further confirm the non-linear nature of flow rates in the individual lines of the FIA systemshown in Fig 1 an experiment with an ion-chromatographic (IC)column was undertaken to separate and measure concentrations offive different transition metals added in the effluents of the FIA
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 1 FIA-CL network for acid degradation of HMTD TATP and MEKP
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 3
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
system by varying the pressure in one individual line linearly Theexperimental conditions of the IC column FIA system as well asvariations of flow rates applied pressure and corresponding metalconcentrations are detailed in the Supplementary information(Section 6 in SI Tables S5ndashS6 and Figs S6-S9)
Regarding selection of optimal flow rates in individual streamsthree separate reactions including decomposition of OPEs to pro-vide maximum yield of H2O2 neutralisation of acidic hydrolysateand chemiluminescent reaction detection of H2O2 with luminolwere considered in this work
321 Acid degradation reactionThe concentrated HCl was used to provide maximum H2O2 yield
in a short time The flow rates in IPADIW line and 32 HCl linewere varied to optimise the contact time between acid and samplein the reaction coil to achieve this goal The reaction times of aciddegradation were calculated between 215 and 138 min (Table 1) A215 min reaction time corresponded to maximum possible H2O2
yield from acid degradation of TATP and HMTD with HClIPA flowratio of 188 Hence line flow rates FLIPAfrac140043 g min1 and FLHClfrac14081 g min1 were chosen from Table 1 for the optimum acid de-gradation of HMTD and TATP However we observed that a contacttime 201 min with corresponding HCLIPA flow ratio of 32 wererequired for maximum possible H2O2 yield from acid degradation ofMEKP Hence line flow rates FLIPAfrac140046 g min1 andFLHClfrac14149 g min1 were chosen from Table 1 for the optimum aciddegradation of MEKP
322 Neutralisation of acidic hydrolysateAs it was found in this study that the flow rate in NaOH line
(ie FLNaOH) did not change with working pressure the flow ratein this line was kept the same as shown in Table 1 Additionallythe HCl flow rate could not exceed 149 g min1 as it causes the pHof the effluent at the outlet of FIA system to fall below 9 whichresulted poor chemiluminescent signal from the detector
323 Chemiluminescent reactiondetection of H2O2
For optimal chemiluminescent detection of H2O2 the flow ratesin copper and luminol lines (ie FLCu and FLLum) were chosen tocompromise between the reagent usage and Ruzickarsquos DispersionCoefficient (Dmax) As the chemiluminescent reaction betweenH2O2 and luminol is pH dependent FLCu and FLLum could not beindependently chosen to produce a minimum Dmax Table S2 insupplementary information illustrated that the chemiluminescentsignal was significantly reduced at pH 8 and was absent at pH 25However Dmax only varied between 7 and 875 at pH 102 and108 respectively As the corresponding FLCufrac14008 g min1 andFLLumfrac14 012 g min1 at pH 108 represented minimal reagentusage these two line flow rates were chosen for the optimumchemiluminescent reactiondetection of H2O2
The calculation of Dmax [21] for total flow rates at the outlet ofthe FIA system is described in Supplementary information (Table
S2) The calculations of optimum concentrations of Nathorn Cl Cu2thorn
and luminol during the chemiluminescent reaction are also pre-sented in Section 7 in Supplementary material
33 Acid degradation of HMTD TATP and MEKP
Samples containing different concentrations of TATP MEKP orHMTD were tested in the FIA-CL system The flow rates in theindividual lines were adjusted within the experimental range asdescribed in the previous section The measured line mass flowrates were then converted to volumetric flow rates by dividing bythe corresponding densities of the reagents at 20deg C Four sets ofexperimental line flow rates in Table 1 resulted total flow rates of162 mL min1 180 mL min1 200 mL min1 and 220 mL min1 atthe outlet The HClIPA flow ratios at these set ups were 188 3235 and 438 (Table 1) The corresponding contact times betweenthe sample plug and concentrated acid were calculated (see TableS3 in Supplementary material) Fig 2 shows detector responses for175 mM injection of MEKP 14 mM injection of TATP and 175 mMinjection of HMTD at contact times of 165 min 182 min 201 and215 min
From Fig 2 it can be seen that the detector responses for 14 mMTATP and 175 mM HMTD increased sharply at acid contact timesover 201 min To the contrary the detector response for 175 mMMEKP actually decreased with a contact time greater than201 min providing the maximum response at 201 min
In this study injections of equimolar concentrations of H2O2
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Table 1Experimental working pressures and corresponding measured mass flow rates in the FIA-CL system the mass of the reagents dispensed in 8 min from each bottle weremeasured and the experiments repeated for five days working pressures in the system were represented by PIPA PHCl PNaOH PCu and PLum flow rates in the system wererepresented by FLIPA FLHCl FLNaOH FLCu and FLLum
Operating pressure MPa Flow rate g min1 RSD Total FlowRateμL min1
Contact time be-tween acid andsample min
HClNaOHflow ratio
HClIPAflowratio
HCl concentrations inreactor coil ()
pH
PIPA PHCl PNaOH PCu PLum FLIPA FLHCl FLNaOH FLCu FLLum
002 007 002 004 002 0043 081 003 008 012 12ndash30 162 215 27 188 304 108003 009 003 006 003 0046 149 0031 01 023 23ndash35 180 201 48 32 311 102004 01 004 008 004 0051 179 003 013 044 09ndash15 200 182 60 35 3107 81005 012 005 01 004 0056 245 0032 015 006 16ndash28 220 165 77 438 3128 35006 015 006 012 007 0067 357 003 018 07 12ndash22 290 138 119 53 3145 25
Fig 2 Detector responses for 175 mM MEKP 175 mM HMTD and 14 mM TATP atdifferent contact times between OPEs and 32 HCl in the 1 m knitted reactor coil attotal flow rates 162 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac14188)180 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac1432) 200 μL min1 (cor-responding line flow ratio FLHCl FLIPAfrac1435) and 220 μL min1 (corresponding lineflow ratio FLHCl FLIPAfrac1443) Concentrations of HCl acid in the reactor coil corre-sponding to the flow rates are also shown
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎4
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
system by varying the pressure in one individual line linearly Theexperimental conditions of the IC column FIA system as well asvariations of flow rates applied pressure and corresponding metalconcentrations are detailed in the Supplementary information(Section 6 in SI Tables S5ndashS6 and Figs S6-S9)
Regarding selection of optimal flow rates in individual streamsthree separate reactions including decomposition of OPEs to pro-vide maximum yield of H2O2 neutralisation of acidic hydrolysateand chemiluminescent reaction detection of H2O2 with luminolwere considered in this work
321 Acid degradation reactionThe concentrated HCl was used to provide maximum H2O2 yield
in a short time The flow rates in IPADIW line and 32 HCl linewere varied to optimise the contact time between acid and samplein the reaction coil to achieve this goal The reaction times of aciddegradation were calculated between 215 and 138 min (Table 1) A215 min reaction time corresponded to maximum possible H2O2
yield from acid degradation of TATP and HMTD with HClIPA flowratio of 188 Hence line flow rates FLIPAfrac140043 g min1 and FLHClfrac14081 g min1 were chosen from Table 1 for the optimum acid de-gradation of HMTD and TATP However we observed that a contacttime 201 min with corresponding HCLIPA flow ratio of 32 wererequired for maximum possible H2O2 yield from acid degradation ofMEKP Hence line flow rates FLIPAfrac140046 g min1 andFLHClfrac14149 g min1 were chosen from Table 1 for the optimum aciddegradation of MEKP
322 Neutralisation of acidic hydrolysateAs it was found in this study that the flow rate in NaOH line
(ie FLNaOH) did not change with working pressure the flow ratein this line was kept the same as shown in Table 1 Additionallythe HCl flow rate could not exceed 149 g min1 as it causes the pHof the effluent at the outlet of FIA system to fall below 9 whichresulted poor chemiluminescent signal from the detector
323 Chemiluminescent reactiondetection of H2O2
For optimal chemiluminescent detection of H2O2 the flow ratesin copper and luminol lines (ie FLCu and FLLum) were chosen tocompromise between the reagent usage and Ruzickarsquos DispersionCoefficient (Dmax) As the chemiluminescent reaction betweenH2O2 and luminol is pH dependent FLCu and FLLum could not beindependently chosen to produce a minimum Dmax Table S2 insupplementary information illustrated that the chemiluminescentsignal was significantly reduced at pH 8 and was absent at pH 25However Dmax only varied between 7 and 875 at pH 102 and108 respectively As the corresponding FLCufrac14008 g min1 andFLLumfrac14 012 g min1 at pH 108 represented minimal reagentusage these two line flow rates were chosen for the optimumchemiluminescent reactiondetection of H2O2
The calculation of Dmax [21] for total flow rates at the outlet ofthe FIA system is described in Supplementary information (Table
S2) The calculations of optimum concentrations of Nathorn Cl Cu2thorn
and luminol during the chemiluminescent reaction are also pre-sented in Section 7 in Supplementary material
33 Acid degradation of HMTD TATP and MEKP
Samples containing different concentrations of TATP MEKP orHMTD were tested in the FIA-CL system The flow rates in theindividual lines were adjusted within the experimental range asdescribed in the previous section The measured line mass flowrates were then converted to volumetric flow rates by dividing bythe corresponding densities of the reagents at 20deg C Four sets ofexperimental line flow rates in Table 1 resulted total flow rates of162 mL min1 180 mL min1 200 mL min1 and 220 mL min1 atthe outlet The HClIPA flow ratios at these set ups were 188 3235 and 438 (Table 1) The corresponding contact times betweenthe sample plug and concentrated acid were calculated (see TableS3 in Supplementary material) Fig 2 shows detector responses for175 mM injection of MEKP 14 mM injection of TATP and 175 mMinjection of HMTD at contact times of 165 min 182 min 201 and215 min
From Fig 2 it can be seen that the detector responses for 14 mMTATP and 175 mM HMTD increased sharply at acid contact timesover 201 min To the contrary the detector response for 175 mMMEKP actually decreased with a contact time greater than201 min providing the maximum response at 201 min
In this study injections of equimolar concentrations of H2O2
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Table 1Experimental working pressures and corresponding measured mass flow rates in the FIA-CL system the mass of the reagents dispensed in 8 min from each bottle weremeasured and the experiments repeated for five days working pressures in the system were represented by PIPA PHCl PNaOH PCu and PLum flow rates in the system wererepresented by FLIPA FLHCl FLNaOH FLCu and FLLum
Operating pressure MPa Flow rate g min1 RSD Total FlowRateμL min1
Contact time be-tween acid andsample min
HClNaOHflow ratio
HClIPAflowratio
HCl concentrations inreactor coil ()
pH
PIPA PHCl PNaOH PCu PLum FLIPA FLHCl FLNaOH FLCu FLLum
002 007 002 004 002 0043 081 003 008 012 12ndash30 162 215 27 188 304 108003 009 003 006 003 0046 149 0031 01 023 23ndash35 180 201 48 32 311 102004 01 004 008 004 0051 179 003 013 044 09ndash15 200 182 60 35 3107 81005 012 005 01 004 0056 245 0032 015 006 16ndash28 220 165 77 438 3128 35006 015 006 012 007 0067 357 003 018 07 12ndash22 290 138 119 53 3145 25
Fig 2 Detector responses for 175 mM MEKP 175 mM HMTD and 14 mM TATP atdifferent contact times between OPEs and 32 HCl in the 1 m knitted reactor coil attotal flow rates 162 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac14188)180 μL min1 (corresponding line flow ratio FLHCl FLIPAfrac1432) 200 μL min1 (cor-responding line flow ratio FLHCl FLIPAfrac1435) and 220 μL min1 (corresponding lineflow ratio FLHCl FLIPAfrac1443) Concentrations of HCl acid in the reactor coil corre-sponding to the flow rates are also shown
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎4
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
were used for comparison of the analytical responses from TATPHMTD and MEKP standards allowing the ratios of the peakheights to be used as an indicator for the degree of degradationArmitt et al [22] established that quantitative degradation of TATPcan produce 3 to 4 molecules of H2O2 Initially 14 mM TATP 175 mMHMTD and 175 mM MEKP were injected at 162 mL min1
flow rateFig 3 shows the degradation of TATP and HMTD as compared toequimolar concentrations H2O2
MEKP could not be degraded under the conditions reported forFig 3 The degradation of MEKP was only possible at flow ratesgreater than 162 mL min1 when the HClIPA flow ratios wereZ32 The chemiluminescent reaction of H2O2 with luminol is pHdependent [10] and flow rates greater than 180 mL min1causedpH at the outlet to fall below 9 (Table 1) resulting poor or nochemiluminescent signal from the detector The increased totalflow rates also imposed the use of higher flow rates in the reagentlines especially in the Cu2thorn and luminol lines which was notdesirable for a cost effective FIA-CL system Additionally Cu2thorn
might act as a catalyst for removal of H2O2 generated from theOPEs degradation at an increased flow rate in Cu2thorn line and de-crease the detector response Pham et al [23] reported removal ofnM level H2O2 and formation of Cuthorn in reaction between 02 mMH2O2 and 04 mM Cu2thorn As 180 mL min1
flow rate in this study
resulted the maximum possible degradation of MEKP the de-gradation performance of the FIA-CL system for MEKP and HMTDat 180 mL min1 is illustrated in Fig 4
Despite the modest degradation performance of 9 this is thefirst report of acid-catalysed degradation of MEKP for FIA The lowsolubility of MEKP in water (65 g L1 at 20 degC) and high degree ofresistance towards HCl acid decomposition [24] were attributed tolow degradation of MEKP in this study As a consequence ofincreasing the flow rate the degradation of HMTD decreased from54 (Fig 3b) to 15 (Fig 4b) most likely due to a reduced contacttime with HCl At 180 mL min1 chemiluminescence response forTATP was very poor indicating the FIA system is not suitable forthe simultaneous detection of HMTD and TATP at high flow rateconditions
At 180 mL min1 the time to reach the maximum peak heightwas 15 min from injection (Fig 4a and b) compared to 25 min ata flow rate of 162 mL min1 (Fig 3a and b) There was a linearrelationship between the chemiluminescence detector responsesand concentration of H2O2 HMTD and TATP over the range of 1ndash200 mM (r2frac14098ndash099) at both flow rates The linear range forMEKP at 180 mL min1 was 20ndash200 mM (r2frac14097) The instru-mental limits of detection (LOD) at a flow rate of162 mL min1were 05 mM for H2O2 HMTD and TATP For a 100 mL
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 3 Degradation of 14 mM TATP (a) and 175 mM HMTD (b) sample as compared to equimolar concentrations of H2O2 Total flow rate was 162 μL min1 (RSD 12ndash3)Line flow rates FLIPAfrac1443 μL min1 FLHClfrac14704 μL min1 FLNaOHfrac1425 μL min1 FLCufrac1480 μL min1 and FLLumfrac14120 μL min1 calculated degradation degree is about 70 forTATP and 54 for HMTD
Fig 4 Degradation of 175 mM MEKP (a) and 175 mM HMTD (b) sample as compared to 175 mM H2O2 Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow ratesFLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 and FLLumfrac14230 μL min1 calculated degradation degree is about 9 for MEKP and 15for HMTD
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 5
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
injection these LODs correlate to 11 ng for TATP 10 ng for HMTDand 17 ng for H2O2 per injection The LOD of MEKP at180 mL min1 was 10 mM (equates to 210 ng in a 100 mL injection)
For TATP LODs reported by GCndashMS ranged from 01 to 5 ng[2526] LODs reported by using Ion Mobility Spectrometry (IMS)ranged from 800 to 1900 ng [27] The lowest LOD for H2O2 bychemiluminescence using Cu2thorn as a catalyst was reported at03 mM [28] which is slightly lower than the 05 μM obtained inthis work The LOD value of 05 μM (11 ng) obtained for TATP inthis study is 10 times higher than GCndashMS reports but 1000 timeslower than with IMS There are no literature data on the detectionof MEKP and HMTD with GC or IMS systems but LODs of 05 mMfor HMTD and 1 mM for TATP have been reported by HPLC-IRmethod [3] Parajuli and Miao [6] reported a LOD of 25 mM forTATP using direct electrogenerated chemiluminescence The de-tection times in these approaches varied from 5 to 12 min sig-nificantly longer than the 25 min achieved in this work RecentlyJohns et al [5] reported the sensitivity of detection of HMTD as831 mM and that of TATP as 878 mM using micellar electrokineticchromatography in post blast scenario which were 166 times and176 times higher than this study for HMTD and TATP respectivelyThe analysis time for OPEs were not explicitly mentioned in [5]
34 Interferences from common household products and metal Ions
It is reasonably expected that common house sources of ketonescan be present with OPEs during their preparation at homeTherefore household products containing benzophenone-1 andacetone such as surface and glass cleaners shoe polish nail polishremover hair cream aftershaves furniture polish and WD-40 lu-bricants were investigated for interference with the acid degrada-tion of MEKP TATP and HMTD Only the nail polish removershowed interferences with the system In order to characterise theinterferences from the problematic compounds in nail polish re-mover separate solutions of acetone in water and benzophenone-1in acetone each of 02 mg mL1 05 mg mL1 and 1 mg mL1wereprepared and injected into the FIA-CL system It was observed thatthe interference from these compounds were not present below1 mgmL concentrations Additionally injections of 005 mg mL101 mg mL1 and 02 mg mL1 of Ba2thorn Mn2thorn Mg2thorn Fe2thorn andZn2thorn separately prepared in 1 mM HMTD 1 mM TATP and 1 mMMEKP samples resulted in a maximum absolute relative error of 3in the peak heights in this 05 mM Cu2thorn catalysed chemilumines-cence reaction
Additionally the possible CL reaction with peracetic acid wastested as it can be present in laundry detergents as well as in otherhousehold bleaching agents The decomposition mechanism ofperacetic acid can follow three potential pathways in aqueous so-lution namely spontaneous decomposition alkaline hydrolysisand transition metal catalyserd decomposition [2930] The spon-taneous decomposition reaches its maximum at pH 82 while boththe alkaline hydrolysis and metal ion catalyserd reactions increasewith increasing pH At pH 105 or higher alkaline hydrolysis be-comes dominant when the metal ion catalyserd decomposition isminimised by metal chelating [2930] As the pH was maintainedbetween 102 and 108 for maximising the chemiluminescent signalfrom the reaction between luminol and hydrogen peroxide in ourFIA analysis the unlikely event of peracetic acid decomposing intohydrogen peroxide through alkaline hydrolysis cannot be ruled outThe products directly containing peracetic acid are exteremely rareavailable in supermarkets around Australia So laundry soaking andstain remover product Ecostores containing sodium percarbonateand tetraacetylethylenediamine (TAED) In aqueous alkaline solu-tion the perhydroxyl anion HO2
(from H2O2) reacts with TAEDand releases (roughly) two equivalents of peracetic acid [31] Sixdifferent solutions of Ecostores of 9864 mg L1 2466 mg L1
1233 mg L1 09864 mg L1 04932 mg L1 and 009864 mg L1
were prepared in deionised water and tested No chemiluminescentemission was observed below 1233 mg L1 of the Ecostores
solutionThe possible effect of degradation products formed during acid
hydrolysis of OPEs on chemiluminescent reaction was also con-sidered According to Armitt et al [22] the identified degradationproducts of TATP when exposed to vapours of HCl in a sealed vialinclude acetone 11-dichloroacetone 111-trichloroacetone DADP13-dichloroacetone 113-trichloroacetone hexachloroacetone andchloroacetone along with H2O2 Acetone has been reported as aninhibitor of the chemiluminescent reaction between luminol andH2O2 by Weber et al [32] In analogy higher chloro-derivatives ofacetone are also expected to inhibit the chemiluminescent reac-tion in this study Additionally Coche and Moutet [33] mentionedthat increasing the number of α-chloride atoms in a carbonylderivative decreases its reduction potential but increases hydra-tion of the carbonyl group which renders the compound moredifficult to be reduced Reduction of the carbonyl group could notbe observed before complete dehalogenation Therefore α-poly-halocarbonyl derivatives of acetone will not oxidise luminol toproduce interferences in the chemiluminescence reaction in thisstudy To the best of authorsrsquo knowledge the acid degradationproducts of HMTD and MEKP have not been investigated by theresearchers and this is beyond the scope of this manuscript Onlyknowledge of thermal degradation products of HMTD [34] andMEKP [35] are available to date
35 Detection of OPEs in real samples
Due to limitations in logistic support inside the chemistry la-boratories and occupational health and safety issues the home-made TATP samples were prepared by officers from the TasmaniaPolice and the home-made HMTD were prepared by scientists atDefence Science and Technology Organisation (DSTO) of Australiausing proprietary methods and supplied to us as dilute solutionsin organic solvents under strict regulatory conditions Sampleswere further diluted in the laboratory (1100) before injecting intothe acid degradation based flow injection system The TATP andHMTD traces are shown in Fig 5 The quantitation of the TATP andHMTD home-made samples are illustrated in Supplementarymaterial
4 Conclusions
A rapid FIA-CL system was developed for the degradation oforganic peroxide explosives TATP HMTD and for the first timeMEKP to H2O2 followed by Cu2thorn catalysed chemiluminescencedetection with luminol Optimisation of the flow rates yielded anoptimum rate of 162 μL min1 for TATP and HMTD and180 μL min1 for MEKP with detection times less than 3 min Flowrates were precisely controlled at different mixing points of apneumatic pressure driven FIA system to ensure the exact con-centrations of reagents The maximum degradation performancewas 70 for TATP 54 for HMTD and 9 for MEKP with detectionlimits from 05 to 10 μM These are the fastest and lowest detec-tion limits (with the exception of lower LOD for TATP in GCndashMS)for these organic peroxides to date This method indicates greatpotential for fast screening of organic peroxide explosives in ex-tracts and liquid samples thanks to its high speed good sensitivityand experimental simplicity
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎6
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
Acknowledgements
The authors especially acknowledge Prof Zhenggui Wei ofAustralian Centre for Research on Separation Science at Universityof Tasmania for supplying the metal concentration data through anIC system Special thanks to Tasmania Police and Defence Scienceand Technology Organisation of Australia for supplying home-made OPE samples The authors also acknowledge the Departmentof Prime Minister and Cabinet of Australia for funding this project
Appendix A Supplementary material
Supplementary data associated with this article can be found inthe online version at httpdxdoiorg101016jtalanta201505052
References
[1] Intelligence and security committee report into the London terrorists attacks on7 July 2005 langhttpwwwcabinetofficegovukmediacabinetofficecorpassetspublicationsreportsintelligenceisc_7july_reportpdfrang (Accessed 101112)
[2] RM Burks DS Hage Current trends in the detection of peroxide-based explosivesAnal Bioanal Chem 395 (2009) 301ndash313
[3] R Schulte-Ladbeck A Edelmann G Quintaacutes B Lendl U Karst Determination ofperoxide based explosives using liquid chromatography with on-line infrared de-tection Anal Chem 78 (2006) 8150ndash8155
[4] D DeTata P Collins A McKinley A fast liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF-MS) method for the identification of organicexplosives and propellants Forensic Sci Int 233 (2013) 63ndash74
[5] C Johns JP Hutchinson RM Guijt EF Hilder PR Haddad M Macka PN Nesterenko AJ Gaudry GW Dicinoski MC Breadmore Micellar electrokineticchromatography of organic andQ2 peroxide-based explosives Anal Chim Acta (2015)httpdxdoiorg101016jaca201502070
[6] S Parajuli W Miao Sensitive determination of triacetone triperoxide explosivesusing electrogenerated chemiluminescence Anal Chem 85 (2013) 8008ndash8015
[7] L Yuan W Lin S Zhu K Zheng L He Single Fluorescent Probe DistinguishesHydrogen Peroxide and Nitric Oxide in Cell Imaging in E Cadenas L Packer (Eds)Hydrogen Peroxide and Cell Signaling Part A 2013 pp 83ndash106
[8] H Cui Q Li R Meng H Zhao C He Flow injection analysis of tannic acid withinhibited chemiluminescent detection Anal Chim Acta 362 (1998) 151ndash155
[9] GM Greenway T Leelasattarathkul S Liawruangrath RA WheatleyN Youngvises Ultrasound-enhanced flow injection chemiluminescence for de-termination of hydrogen peroxide Analyst 131 (2006) 501ndash508
[10] L Marle GM Greenway Determination of hydrogen peroxide in rainwater in aminiaturised analytical system Anal Chim Acta 548 (2005) 20ndash25
[11] D Price R Fauzi C Mantoura PJ Worsfold Shipboard determination of hydrogenperoxide in the western Mediterranean sea using flow injection with chemilumi-nescence detection Anal Chim Acta 377 (1998) 145ndash155
[12] S Girotti E Ferri E Maiolini L Bolelli M DrsquoElia D Coppe FS Romolo A quanti-tative chemiluminescent assay for analysis of peroxide-based explosives Anal
Bioanal Chem 400 (2011) 313ndash320[13] R Schulte-Ladbeck P Kolla U Karst Trace analysis of peroxide-based explosives
Anal Chem 75 (2003) 731ndash735[14] R Schulte-Ladbeck U Karst Determination of triacetonetriperoxide in ambient air
Anal Chim Acta 482 (2003) 183ndash188[15] JC Oxley JL Smith J Huang WJ Luo Destruction of peroxide explosives Forensic
Sci Sep 54 (2009) 1029ndash1033[16] M Bali Niche threat Organic peroxides as terrorist explosives Aust Army J X
(2013) 35ndash48 (accessed 290613)langhttpwwwarmygovauOur-futureLWSCOur-publications mediaFilesOur20futureLWSC20PublicationsAAJ2013AutumnBali_Niche20Threatpdfrang
[17] P Kubaacutentilde S Liu PK Dasgupta in M Trojanowicz (Ed) Electroosmosis-Driven FlowAnalysis In Advances in Flow Analysis copyWiley-VCH Weinheim 2008
[18] DJ Malcolme-Lawes GA Milligan A novel approach to non-segmented flow ana-lysis Exp Syst J Clin Lab Automat 9 (1987) 179ndash183
[19] M Valcarcel MDL De Castro Flow Injection Analysis Principals and ApplicationsEnglish language edition copyEllis Horwood Limited Chichester England 1987
[20] Eacute Tyrrell GW Dicinoski EF Hilder RA Shellie MC Breadmore CA Pohl PR Haddad Coupled reversed-phase and ion chromatographic system for the si-multaneous identification of inorganic and organic explosives J Chromatogr A 1218(2011) 3007ndash3012
[21] J Ruzicka EH Hansen Flow Injection Analysis 2nd edition copy John Wiley and SonsNew York 1988
[22] D Armitt P Zimmermann S Ellis-Steinboener Gas chromatographymass spec-trometry analysis of triacetone troperoxide (TATP) degradation products RapidCommun Mass Spectrom 22 (2008) 950ndash958
[23] AN Pham G Xing CJ Miller TD Waite Fenton-like copper redox chemistry re-visited hydrogen peroxide and superoxide mediation of copper-catalyzed oxidantproduction J Catal 301 (2013) 54ndash64
[24] J Tseng Y Chang T Su C Shu Study of thermal decomposition of methyl ethylketone peroxide using DSC and simulation J Hazard Mater 142 (2007) 765ndash770
[25] A Stambouli A El Bouri T Bouayoun M Bellimam Headspace-GCMS detection ofTATP traces in post-explosion debris Forensic Sci Int 146S (2004) S191ndashS194
[26] A Kende F Lebics Z Eke K Torkos Trace level triacetone-triperoxide identificationwith SPMEndashGCndashMS in model systems Microchim Acta 163 (2008) 335ndash338
[27] J Oxley J Smith L Kirschenbaum S Marimganti S Vadlamannati Detection ofexplosives in hair using ion mobility spectrometry J Forensic Sci 53 (2008)690ndash693
[28] H Ma U Jarzak W Thiemann Synthesis and spectroscopic properties of new lu-minol-linked calixarene derivatives Anal Chim Acta 362 (1998) 121ndash129
[29] Z Yuan Y Ni ARP Van Heiningen Kinetics of Peracetic decomposition Part Ispontaneous decomposition at typical pulp bleaching conditions Can J Chem Eng75 (1997) 37ndash41
[30] Z Yuan Y Ni ARP Van Heiningen Kinetics of the Peracetic acid decompositionPart II pH effect and alkaline hydrolysis Can J Chem Eng 75 (1997) 42ndash47
[31] DM Davies ME Deary Kinetics of the hydrolysis and perhydrolysis of tetra-acetylethylenediamine a peroxide bleach activator J Chem Soc Perkin Trans 2(1991) 1549ndash1552
[32] K Weber Z Prochazka J Spoljaric Die Wirkung yon Fremdstoffzusatz auf die Lu-minolreaktion Croat Chem Acta 28 (1956) 25ndash31
[33] L Coche J Moutet Selective electroanalytic reduction of hexachloroacetone on aviologen polymer modified electrode in aqueous media J Electroanal Chem 245(1988) 313ndash319
[34] JC Oxley JL Smith H Chen E Cioffi Decomposition of multi-peroxide compoundsPart II Hexamethylene Triperoxide Diamine (HMTD) Thermochim Acta 388 (2002)215ndash225
[35] M Yuan C Shu AA Kossoy Kinetics and hazards of thermal decomposition ofmethyl ethyl ketone peroxide by DSC Thermochim Acta 430 (2005) 67ndash71
123456789
101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566
676869707172737475767778798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
Fig 5 Degradation of home-made TATP sample supplied by Tasmania Police (a) and home-made HMTD sample supplied by DSTO (b) samples were compared to 29 mMH2O2 standard Total flow rate was 180 μL min1 (RSD 23ndash35) Line flow rates FLIPAfrac1443 μL min1 FLHClfrac1413 mL min1 FLNaOHfrac1426 μL min1 FLCufrac14100 μL min1 andFLLumfrac14230 μL min1
P Mahbub et al Talanta ∎ (∎∎∎∎) ∎∎∎ndash∎∎∎ 7
Please cite this article as P Mahbub et al Talanta (2015) httpdxdoiorg101016jtalanta201505052i
