AN

ALYST

CO

MM

UN

ICA

TIO

N

THE

www.rsc.org/analyst

A field test for the detection of peroxide-based explosives

Rasmus Schulte-Ladbeck,a Peter Kollab and Uwe Karst*a

a University of Twente, Department of Chemical Analysis and MESA+ Research Institute, P.O.Box 217, 7500 AE Enschede, The Netherlands

b Bundeskriminalamt, KT 16, 65173 Wiesbaden, Germany

Received 9th July 2002, Accepted 7th August 2002First published as an Advance Article on the web 16th August 2002

A rapid and simple field test for the detection of triacetone-triperoxide (TATP) and hexamethylenetriperoxidediamine(HMTD), two explosives which find significant illegal use,has been developed. Unknown samples are first treated witha catalase solution to remove hydrogen peroxide traces, inorder to provide selectivity towards peroxide-based bleach-ing agents which are contained in commercial laundrydetergents. Subsequently, the peroxide-based explosives aredecomposed via UV irradiation, thus yielding hydrogenperoxide, which is determined by the horseradish peroxidase(POD) catalysed formation of the green radical cation of2,2A-azino-bis(3-ethylbenzothiazoline)-6-sulfonate (ABTS).The limits of detection for this method are 8 3 1026 moldm23 for TATP and 8 3 1027 mol dm23 for HMTD,respectively. As an option, p-hydroxyphenylacetic acid(pHPAA) may be used as peroxidase substrate, resulting inlower limits of detection (8 3 1027 mol dm23 for TATP andHMTD). The complete method uses a mobile setup to beapplied under field conditions.

Introduction

Legal authorities are currently observing an increasing numberof incidents involving the two explosives triacetonetriperoxide(TATP) and hexamethylenetriperoxide diamine (HMTD).There are different areas of occurrence, the most prominentfields being terrorism1,2 and drug related crime.3 Accidents withTATP and HMTD4 are a subject of interest as well. Theseperoxide-based explosives are very easy to synthesize startingfrom readily available chemicals. TATP was first prepared byWolffenstein in the 19th century.5 It is almost as powerful asTNT and one of the most sensitive explosives known. One ofthe reasons why TATP is not applied for any commercial ormilitary purposes is its tendency to sublime within a few days.6HMTD was first synthesized by Legler in 1881.7 It is not assensitive to impact as TATP, but nevertheless it is an explosivewhich is quite difficult to handle. HMTD has also no knownmilitary or commercial applications.

In those cases in which unknown explosives are involved,governmental agencies and police forces share a commoninterest. They have to make quick on-site threat evaluation,regarding both public and their own safety, respectively. Thisevaluation starts with the fast detection and especially theidentification of an unknown substance as an explosive. This isvery difficult with respect to TATP and HMTD, as these are noteasily detectable with common field methods for the identifica-tion of explosives, which is normally done by the use ofchemical sensors or by the help of dogs. TATP and HMTD havealso a very unsuspicious appearance because both are whitepowders with no obvious characteristics.

There are very few direct methods to determine thesesubstances, as they are neither fluorescent nor do they havesignificant absorption within the UV range. The methods, whichare currently used, are based on IR spectroscopy or chemicalionization mass spectrometry (CI-MS)3,4,8 for TATP. HMTDcan either be identified by means of IR spectroscopy,8 CI-MS9

or LC-MS with an atmospheric pressure ionisation (APCI)

interface.10 These methods have the disadvantage that they arenot applicable to a fast, on-site test, which could be carried outunder field conditions. The aim of this work was to develop atest scheme which is sufficiently rugged to deploy in the fieldand which can give a clear identification as fast as possible.

Experimental

Safety note

TATP and its homologues as well as HMTD are extremelydangerous materials, which may lead to severe and spontaneousexplosions under impact, friction and temperature changes. Thesynthesis of these substances may only be carried out by highlyqualified personnel, using appropriate safety measures (re-inforced goggles and gloves, splinter-proof vessels, protectiveshield, etc.) and in small quantities. For this work, thesubstances were synthesized according to literature procedures(see below) in quantities not exceeding 100 mg. Working withlarger amounts of the substance significantly increases thedanger of spontaneous explosions.

Instrumentation

UV/VIS. A hand-held photometer SQ118 from Merck(Darmstadt, Germany) was used for field measurements. For therecording of the UV/VIS spectra, a diode array photometerModel HP 8453 with the software UV–visible Chemstation845x from Hewlett-Packard (Waldbronn, Germany) was used.

Fluorescence spectroscopy. The microplate fluorescencespectrometer Fluostar from BMG (Offenburg, Germany) wasused for all fluorescence measurements.

Chemicals

All chemicals were purchased from Aldrich Chemie (Stein-heim, Germany), Merck (Darmstadt, Germany), Sigma (Dei-senhofen, Germany) and Fluka (Neu-Ulm, Germany) in thehighest quality available. Acetonitrile was Merck gradientgrade. Horseradish peroxidase [EC 1.11.1.7] and catalase [EC1.11.1.6.] were purchased from Sigma (Deisenhofen, Ger-many).

As laundry detergents for interference studies, the followingproducts were used: Ultra Biz Bleach (Procter and Gamble;Cincinnati, OH, USA); A&P Compact Reiniger (HamburgerWaren Kontor GmbH; Hamburg, Germany); Skip Waschsys-tem (Lever GmbH; Hamburg, Germany); Sun Bleach (HuishDetergents Inc.; Salt Lake City, UT, USA).

Synthesis. The synthesis of TATP and HMTD was per-formed as described in the literature.5,11 For this work, thesynthesis of the analytes was carried out in such a way that 100mg of the respective explosive was obtained, provided that a

This journal is © The Royal Society of Chemistry 2002

1152 Analyst, 2002, 127, 1152–1154 DOI: 10.1039/b206673b

Publ

ishe

d on

16

Aug

ust 2

002.

Dow

nloa

ded

by U

nive

rsity

of

Sydn

ey o

n 07

/10/

2013

07:

29:3

2.

View Article Online / Journal Homepage / Table of Contents for this issue

quantitative reaction had taken place. For safety precautions,refer to the Safety Note above. Excess amounts of theexplosives can be destroyed according to ref. 6.

Rapid UV/VIS-based test for the identification of HMTDand TATP. Solid samples were first washed with a solution of0.25 mg of catalase in 50 mL of distilled water and then withpure water. The samples were subsequently extracted withacetonitrile and filtered off. The solution was irradiated withUV light of 254 nm for 15 min. Afterwards, it was diluted withwater (1+10) to avoid the deactivation of the peroxidase byacetonitrile in the following step. The reagent (2.6 mg POD and25.2 mg 2,2A-azino-bis(3-ethylbenzothiazoline)-6-sulfonate(ABTS) in 50 ml 0.01 mol dm23 acetate buffer with a pH of 5.5)was mixed with the sample in a ratio of 1+1. Detection wasperformed with a hand-held photometer at a wavelength of 415nm.

Fluorescence-based test method for the identification ofHMTD and TATP. Solid samples were first washed with asolution of 0.25 mg of catalase in 50 mL of distilled water, andsubsequently with pure water. The sample was extracted withacetonitrile and filtered off. The solution was irradiated withUV light, 254 nm, for 15 min. Afterwards, it was diluted withwater (1+10) to avoid the deactivation of the peroxidase byacetonitrile in the following step. 60 mL of the diluted samplewere mixed with 50 mL of reagent solution [2.5 mg POD and 7.6mg p-hydroxyphenylacetic acid (pHPAA) in 50 mL of ammoniabuffer (pH = 9.5; 0.01 mol dm23)]. After an incubation periodof 15 min at 25 °C, the fluorescence was determined on amicroplate reader using an excitation wavelength of 320 nm andan emission wavelength of 405 nm.

Results and discussion

This rapid testing scheme is intended to be used for theidentification of unknown white powders that are suspected tobe explosives. Selectivity is an even more important issue for

this test than sensitivity, because both false positive and falsenegative results may cause severe consequences with respect tohealth risks or costs, respectively. In the analysis of peroxide-based explosives, false positives could easily result fromlaundry detergents that contain peroxide-based bleachingsystems. Therefore, more reactive peroxides, especially hydro-gen peroxide, have to be removed prior to the application of thedetection scheme for peroxide-based explosives. This is easilyperformed by adding an aqueous solution of catalase, anenzyme which rapidly decomposes hydrogen peroxide, but doesnot decompose TATP or HMTD. The analytes are almostinsoluble in water, providing even more selectivity towards thewell water-soluble peroxide-based bleaching constituents.

After removal of the enzyme by extraction of the solid sampleusing acetonitrile, UV irradiation of 254 nm is applied todecompose the peroxide-based explosives to hydrogen peroxide(see Fig. 1). The irradiation time should not be longer than 15min, as it was found that the UV-irradiation additionallydecomposes the H2O2 with time. If the irradiation time is longerthan 15 min, the H2O2 produced from the decomposition of theanalytes will also be destroyed to a significant extent. Thus, theirradiation time is a compromise between quantitative decom-position of the analyte and the destruction of the formedhydrogen peroxide. Then the formed hydrogen peroxide isdetected with an enzyme-catalyzed method for peroxidedetermination. Therefore, 2,2A-azino-bis(3-ethylbenzothiazo-line)-6-sulfonate (ABTS) was used as a substrate for horse-radish peroxidase (POD) in this detection scheme. ABTS is oneof the most popular substrates for peroxidase. ABTS iscolourless, and it reacts readily with hydrogen peroxide inpresence of peroxidase to yield a green coloured radical cation(see Fig. 2), which absorbs light at 405, 415, 650, 730 and 810nm. In most cases, the strongest absorption at 415 nm is used fordetection purposes (see Fig. 3). The other wavelengths, 650,730 and 810 nm, are more effective in coloured solutions wheredetection at lower wavelengths does not provide sufficientselectivity. The enzyme-catalyzed reaction step provides se-lectivity towards other oxidizers as only hydrogen peroxide andprimary hydroperoxides are activated by the peroxidase used.

Fig. 1 Decomposition of triacetonetriperoxide (TATP) and hexamethyle-netriperoxide diamine (HMTD).

Fig. 2 Oxidation of 2,2A-azino-bis(3-ethylbenzothiazoline)-6-sulfonate (ABTS) by means of peroxidase (POD) and hydrogen peroxide.

Fig. 3 UV-spectrum of ABTS+4 after photochemical decomposition of 13 1024 mol dm23 TATP and enzymatic reaction of the hydrogen peroxideformed with peroxidase and ABTS.

Analyst, 2002, 127, 1152–1154 1153

Publ

ishe

d on

16

Aug

ust 2

002.

Dow

nloa

ded

by U

nive

rsity

of

Sydn

ey o

n 07

/10/

2013

07:

29:3

2.

View Article Online

Blanks recorded without the addition of peroxidase can be usedto verify the absence of other oxidizers, which might directlyreact with ABTS. For this method, a portable UV-photometerand a hand-held UV-lamp were used. The whole setup fits intoa small suitcase.

While the ABTS method is fast and reliable, a more selectiveand sensitive possibility, especially in complex matrices, couldbe advantageous. The use of p-hydroxyphenylacetic acid(pHPAA) as a peroxidase substrate in a detection scheme basedon fluorescence spectroscopy is fulfilling these requirements.pHPAA is known to dimerize in the presence of POD andhydrogen peroxide (see Fig. 4). The dimer is highly fluorescentand has already been used for the detection of several differenthydroperoxides.12–14 Especially for TATP, where the LOD ofthe pHPAA method is by one decade lower than that of theABTS method, this variation is attractive, although a morecomplex and time-consuming experimental setup has to beused.

For the semi-quantitative estimation of the concentration ofHMTD and TATP on the ABTS method, the absorption at awavelength of 415 nm was used. For the pHPAA method, anexcitation wavelength of 320 nm and an emission wavelength of405 nm was used (see Fig. 5). The linear ranges for calibration,LODs and LOQs for both methods are summarized in Table 1.For the concentration range from 3 3 1025–1 3 1024 mol dm23

the relative standard deviation (RSD) for HMTD and TATPwith the ABTS method was 7–11% for HMTD and 6–10% forTATP, respectively. The pHPAA method had RSDs of 2–8%for HMTD and 2–5% for TATP for the concentration range of3 3 1026–1 3 1024 mol dm23.

More than thirty real samples comprising both peroxide-based explosives and cleaning agents were investigated. For alllaundry detergents, no false positives were observed for bothvariations of the rapid testing scheme. Regarding the investiga-tion of real samples containing explosives, TATP and HMTD indifferent matrices, mostly contaminated with earth and dust,were used. In all tests involving peroxide-based explosives,correct identification was observed.

Conclusions

A new rapid and simple field test for the identification ofperoxide-based explosives has been developed. This testscheme can be performed by a person without a chemicalbackground and will give results within less than 30 min. Future

work will be directed to the development of a method for thedetermination of TATP in the gas phase.

Acknowledgement

Financial support by the German Minister of the Interior(Berlin, Germany) is gratefully acknowledged.

References

1 The Independent, London, 1996, Oct. 8, p. 7.2 R. T. Cooper, Los Angeles Times, 2001, Dec. 29, p. A12.3 G. M. White, J. Forensic. Sci., 1992, 37, 652–656.4 H. K. Evans, F. A. J. Tulleners, B. L. Sanchez and C. A. Rasmussen,

J. Forensic. Sci., 1986, 31(3), 1119–1125.5 R. Wolffenstein, Chem. Ber., 1895, 28, 2265–2269.6 A. J. Bellamy, J. Forensic. Sci., 1999, 44(3), 603–608.7 L. Legler, Chem. Ber., 1881, 14, 602–604.8 S. Zitrin, S. Kraus and B. Glattstein, Proceedings of the International

Symposium on the Analysis and Detection of Explosives, U.S.Government Printing Office, Washington, DC, 1984, pp. 137–141.

9 D. Suelzle and P. Klaeboe, Acta Chem. Scand., 1988, A24,165–170.

10 A. Crowson and M. S. Beardah, Analyst, 2001, 126, 1689–1693.11 W. P. Schaefer, J. T. Fourkas and B. G. Tiemann, J. Am. Chem. Soc.,

1985, 107, 2461–263.12 A. L. Lazrus, G. L. Kok, J. A. Lind, S. N. Gitlin and S. E. McLaren,

Anal. Chem., 1985, 57, 917–922.13 H. Sukugawa and I. R. Kaplan, Atmos. Environ., 1987, 21,

1791–1798.14 J. Meyer, A. Büldt, M. Vogel and U. Karst, Angew. Chem. Int Ed.,

2000, 39, 1453–1455.

Table 1 Analytical figures of merit of the ABTS and pHPAA methods for TATP and HMTD

UV/Vis (ABTS) Fluorescence (pHPAA)

LOD/mol dm23 LOQ/mol dm23 Linear range/mol dm23 LOD/mol dm23 LOQ/mol dm23 Linear range/mol dm23

TATP 8 3 1026 3 3 1025 3 3 1025–1 3 1024 8 3 1027 3 3 1026 3 3 1026–5 3 1025

HMTD 8 3 1027 3 3 1026 3 3 1026–5 3 1024 8 3 1027 3 3 1026 3 3 1026–1 3 1024

Fig. 4 Dimerization of p-hydroxyphenylacetic acid (pHPAA) in thepresence of peroxidase (POD) and hydrogen peroxide.

Fig. 5 Fluorescence spectrum of pHPAA after photochemical decomposi-tion of 1 3 1024 mol dm23 HMTD and the dimerization of pHPAA inpresence of peroxidase (POD) and hydrogen peroxide.

1154 Analyst, 2002, 127, 1152–1154

Publ

ishe

d on

16

Aug

ust 2

002.

Dow

nloa

ded

by U

nive

rsity

of

Sydn

ey o

n 07

/10/

2013

07:

29:3

2.

View Article Online