ARTHUR D. LITTLE, INC. CAMBRIDGE, MA 02140 · 2011. 5. 14. · ARTHUR D. LITTLE, INC. CAMBRIDGE, MA 02140 SEPTEMBER, 1982 DTIC "FINAL REPORT -CT•w Volume I CT29 g1E Distribution

REPORT DRXTH-TE-CR-82179

SURFACE SAMPLING TECHNIQUES

"Bruce E. GoodwinJames R. AronsonRobert P. O'NeilMargaret A. RandelEmmett M. Smith

ARTHUR D. LITTLE, INC.CAMBRIDGE, MA 02140

SEPTEMBER, 1982 DTIC"FINAL REPORT -CT•wVolume I CT29 g1EDistribution Unlimited UCleared for Public Release H

prepared forU.S. Army Toxic and Hazardous Materials Agency,Aberdeen Proving Ground, Maryland 21010

DISCLAIMER

The veWs. opinions andlor findings containedIn this report are those of the authors andshould not be construed as an official Depart-ment of the Army position, policy or decision.unless so designated by other documents.

REPORT DRXTH-TE-CR -82179

SURFACE SAMPLING TECHNIQUES

Bruce E. GoodwinJames R. AronsonRobert P. O'NeilMargaret A. RandelEmmett M. Smith

ARTHUR D. LITTLE, INC.CAMBRIDGE, MA 02140

DTg~fSEPTEMBER, 1982 r' E-rC-T 9FINAL REPORT OCTa g ID19Volume IDistribution Unlimited HCleared for Public Release

prepared for

U.S. Army Toxic and Hazardous Materials Agency,Aberdeen Proving Giound, Maryland 21010

69CURITY CLAISIVICATIOw OF ?'MIS PAOE r"..m 04til EAW010e)

REPORT DOCUMENTATION PAGEREDISUCON1 2, oovY ACCESSION No. 3. ANECiPICHT'S CATA609 NWMIER

4. T1rL1(omi SMAml)S. TYPE OF RIVONT A PERIOD COVEREO

Surfaze Samipling Techniques Final6ip1q.1 PqOAuiiNO GOR. REPORT NUMBER

7. AU ""o( 4. CONTRAACT OR GRANT UMUE@R(I)

pJames R&. Aronbon Margarct A. ~ lDAAK-81-C-.0014Robart P. O'Neil Emm~ett M. St

FIRPS046o04Z OG OANIZATION NAME AND AL00011s'. 10. PRAONAM ELEM9NT, PROjeCT. TASIK

Arthur D. Little, Inc. AnAaWORK UNMIT NUMBERSAcorn Park.Cambridge, 4A 02140

Me 1 C04'100L.140 OFFICE NAME AMC) ADORSSS IS. REPOnT OATK

US. Army Toxic and Iiggardoux Materilsh Agency 5epterlber 3.98?Attantivn: tileen Reillhy-Wiedow I). MUMMER OF PAOESAberdwan P'roving Ground (F.A), Maryland

?1JT5R7.hTV~AMT AOOESS(I EIIIIAl free, Cee~m~Ind Of/#@*; 16, 89CWRITY CL.ASS. (a/ 014l tepuff)

unclassified-30, U7-fF /OWNRADINO

Distribution unliraited.Chcarad .for public rci.Qaze.

17. 0161 AISUTI~oN 6 ATIMCNT (.1 IA. &h.omesO fieted InASleeok 20, l~iffoIIentrwt fipo

Explowive rowsidiuseSurf acesSamupling Tachn~riepiu

12 ApernAc?' tvwie asmwo .4p it~e.w mweIr ndeftMil &F Sleek ~~inmowo)Qiial i at ivv and q'10n11L3 Li iv WUi6j1116 .111- J.I lyb1 IIULIU.. for tile detectionand JaLsrmIinatioli of explosivei/saxp]Os±'/C resi~dues onl building materlaln;ourtaios are dvocribud, Qu~nLiWA1Ve ý:rtiflcatlon testing data tor a solventex1.iactivti - MIXH mathod are Inc ludrd.

* DD 1473 IUIITIQN Or I NOYV 46 IS OVISCLtTILSECURITY C-ý.A1111PICATIQWt OP TiigS PAGE (W~o Data Enleped)

r-

SUbLARY

Under Contract No. DAAK1l-81-C-0014 to the U. S. Army Toxic and Hazardous

Materials Agency, Arthur D. Little, Inc. has developed sampling protocolsfor the determination of explosives/explosive residues on building materialssurfaces. The Army's need for this sampling and analysis capability hasarisen in connection with the release of surplus government property (e.g.,former ammunition plants) and the specified requirement that as part of

h.these release programs, a determination be made as to whether the propertycan be released for unrestricted use. In the case of buildings known orsuspected to have been contaminated with explosives, the Army is seekingfor this purpose sampling and analytical procedures which would permit:(1) rapid qualitative determination with 90% confidence of the presence/absence of the compounds of interest down to a level of 5 pg/l0 cm2 in agiven building, and (2) if any of these- compounds is detected, precise,accurate quantitative determination of the amount of cuntamination downto the same 5 pg/10 cm2 level.

-- This study resulted in the development of several muethods for the samplingand analysis of explosives/explosive residues on building materials sur-faces. A method for qualitative determination based on detection ofcharge-transfer complexes formed between the explosive/explosive residueand a visualization reagent applied to the surface has been evaluated inthe field at two Army Ammunition Plants. A method for quantitativedetermination based on solvent extraction of samples collected in thefield followed by high pressure liquid chromatographic analysis has beenevaluated on samples prepared in the laboratory, and found to givepromising results. The theoretical and practical feasibility of anothermethod for qualitative determination based on UV irradiation of a suspectedcontaminated surface with subsequent detection of any explosives/explosiveresidues present has been demonstrated. This approach may provide themeans for "scanning" an area to determine whether explosives are presenton a real-time basis. Further development of this approach is recommended.

Accession For--__

NTS

E'-'K81q'T }

U

Di.Arthur 1).Little.ic

TABLE OF CONTENTS

17 Page

LIST OF TABLES iii

LIST OF FIGURES iv

S " SUMMARY 1

I. INTRODUCTION 2

II. SAMPLING AND ANALYSIS ,METHOD SELECTION 4

A. INTRODUCTION 4

B. LITERATURE SEARCH 4

1. Approach 42. Sources and Coverage 43. Results 5

C. REVIEW OF SPOT TEST CHARACTERISTICS 21

II.D. FINDINGS AND CONCLUSIONS 33

SE. SAMPLING PROTOCOL SELECTION 34

III. QUALITATIVE METHODS DEVELOPMENT 35

A. INTRODUCTION 35

B. ANALYTE DETECTION USING CONTINUOUS VAPOR 35PHASE MONITORING

C. ANALYTE DETECTION USING FORMATION OF CHARGE 37TRANSFER COMPLEXES WITH VISUAL IDENTIFICATION

1. Background 372. Preliminary Experiments 383. Formation of Charge-Transfer Complexes 40

Directly on Surfaces4. Solvent Lift Technique for Sampling 40

of Surfaces5. Field Evaluation 43

Arthur L)Little Inc

TABLE OF CONTENTS (Continued)

FA Page

D, ANALYTE DETECTION USING UV IRRADIATION AND 48THERNAL DLAGING

i. Background 482. Theoretical Analysis of the Method 553. Analyte UV Absorption Characteristics 604. Laboratory Demonstration 63

E. ANALYTE DETECTION USING LW IRRADIATION AND 65I UV PHOTOGRAPHY

1. Background 65"2. Preliminary Experiments 65

IV. QUANTITATIVE METHODS DEVELOPI•fENT 71

A. INTRODUCTION 71

B. SW'IQUANTITATIVE CERTIFICATION TESTING 71

C. QUANTITATIVE CERTIFICATION TESTING 71

1. Introduction 712. Analytical Methods 743. Results and Discussion 74

V. CONCLUSIONS AND RECOMMENDATIONS 90

DISTRIBUTION LIST 91

Arthur D Littleý. Inc

r

LIST OF TABLES

Table No. Page

II-i BIBLIOGRAPHY OF USEFUL CITATIONS 6

11-2 ANALYTES DETERMINED IN LISTED CITATIONS 16

. 11-3 ANALYTICAL METHODS DESCRIBED IN LISTED 17CITATIONS

11-4 SAMPLING AND SAMPLE PREPARATION PROCEDURES 18DESCRIBED IN LISTED CITATIONS

- 11-5 SPOT TEST MXETHODS FOR ORGANIC ANALYTES 23

11-6 REAGENTS FOR ORGANIC ANALYTES 28

11-7 SPOT TEST METHODS FOR INORGANIC ANALYTES 31

11-8 REAGENTS FOR INORGANIC ANALYTES 32

III-I LABORATORY ANýLYTE DETECTION LIMITS 42(MICROGRAM/cm ) OBTAINED BY FORMATIONOF CHARGE TRANSFER COMPLEXES DIRECTLY

15ON SURFACES

111-2 LABORATORY ANALYTE DETECTION LIMITS 44(MICROGRAMS/cm 2 ) USING SOLVENT LIFTTECHNIQUES

111-3 INVENTORY OF SAMPLES COLLECTED AT HOLSTON 45AND JOLIET ARMY AMMUNITION PLANTS

111-4 PARAMETERS USED IN SAMPLE CALCUILATIONS 59

I11-5 UV ABSORPTION DATA FOR ANALYTES IN SOLUTION 61

111-6 SOLID STATE AINALYTE UV ABSORPTION COEFFICIENTS 64

111-7 RESULTS OF UV PHOTOGRAPHIC EXPERIMENTS 68

IV-1 SEMIQUANTITATIVE CERTIFICATION TESTING 72ANALYTICAL METHOD CONDITIONS

IV-2 SEMIQUANTITATIVE CERTIFICATION TESTING 73STATISTICAL DATA SUMMARY

IV-3 QUANTITATIVE CERTIFICATION TESTING 75STATISTICAL DATA SUMMARY

IV-4 RESULTS OF ANALYSES FOR RDX, TNB, TNT, AND 88TETRYL IN CONDUCTIVE NON-SPARKING FLOORINGSAXPLES

Arthur 1) Little, Inc

LIST OF FIGURES

Figure No. Page

I1-1 iHolston AAP: BuildIng I-1; RDX Nanufactur 49Manufacture

111-2 Joliet AAP: TNT Melting Area, 2nd Floor 50

111-3 Joliet AAP: TNT Loading Area, ist Floor 51

"111-4 Joliet AA: DNT Sweathouse 52

111-5 Joliet AAP: TNT Wash Building 53

111-6 Joliet AAP: Tetryl Packaging Building 54

111-7 UV Photography of 130 vg/cm2 Each of 662,4-DNT, Tetryl, 2,6-DNT, RDX, and TNTon TLC PlaLe

111-8 UV Photography of TNT, Tetryl, and 662,4-DNP on Concrete

III-9 Analytes on Silica Gel and Metal. 365 nm 70

III-10 Analytes on Silica Gel (top) and Transite 70

(bottom). 254 nm

iv

Arthur 1) Littlc Inc

I. INTRODUCTION

Under Contract No. DAAK1l-83.-C-0014 to the U.S. Army Toxic and HazardousSMaterials Agency, Arthur D. Little, Inc. has developed sampling protocolsfor the determination of explosives/explosive residues on buildingmaterials surfaces. The Army's need for this sampling and analysis capabilityhas arisen in connection with the release of surplus government property(e.g., former ammunition plants) and the specified requirement that as partof these release programs, a determination be made as to whether the propertycan be released for unrestricted use. In the case of buildings known orsuspected to have been contaminated with explosives, the Army is seekingfor this purpose sampling and analytical procedures which would permit:(1) rapid qualitative determination with 90% confidence of the presence/absence of the compounds of interest down to a level of 5 ug/10 cm2 in agiven building, and (2) if any of these ccmpounds is detected, precise,accurate quantitative determination of the amount of contamination down tothe same 5 wg/10 cm2 level.

The specific compounds and surface types of interest were the following:

Compounds

1. 2,4,6-trinitrotoluene (TNT)2. 2,4- and 2,6-dinitrotoluene3. Cyclotrimethylenetr initramine (RDX)4. Pentaerythrite tetranitrate (PENT)

Le 5. Nitroglycerine6. 2,4,6-trinitrophynylmethylnitramine (Tetryl)7. Diphenylamine8. 1, 3,5-trinitrobenzene9. 2,4-dinitrophenol

10. Cd11. Pb12. Hg13. Cr+314. Cr+6

Surface Types

1. Concrete--unpainted2. Brick--glazed and unglazed3. Transite4. Wood5. Metal6. Conducting non-sparking floor

The approach that was taken to the development of these sampling protocolsinvolved the following steps:

A. Literature SearchB. Sampling Protocol SelectionC. Analytical Method Selection/DevelopmentD. Certification TestingE. Development of a Spike and Recovery Test PlanF. Spike and Recovery TestingG. Interference Testing

L 2 Arthur 1) l-ttle. Inc

The Literature Search provided an overview of existing sampling andanalysis methods. The results of the literature search suggested thatmost existing methods would not satisfy the Army's requirements as statedabove. Thus, in the Sampling Protocol Selection step, several approachesbased on technology developed for applications other than explosivesanalysis were proposed for further development. Those approaches involvedmethods intended specifically 'or qualitative analysis, and a methodintended specifically for quantitative analysis of explosives/explosiveresidues on surfaces.

Development of the quantitative analysis method proceeded with an Analy-tical Method Selection/Davelopment step, in which advantage was takenof the fact that several methods for the dtermination of explosiveswere present in the USATHA&L data base. These methods were modified as

- necessary to permit their application to the present study, and theprecision and accuracy for each of the resulting methods were assessedin preliminary Certification Testing (semiquantitative).

Subsequent to approval by the Technical Project Officer of a Spike andRecovery Test Plan, meýhcds for each analyte on each surface type wereevaluated by Spike and Recovery Testing. Finally, methods were subjected

* to Quantitative Certification Testing according to the requirements inthe 1980 USATHAMA QA Plan. Since most of the analytes are determined ina single extract, the Laed for Interference Testing in the laboratory waslargely eliminated, interferences from other sources which may be present

Le •in samples collected in the field were not addressed in this study.

This study resulted in the development of several methods for the samplingand analysis of explosives/explosive residues on building materials sur-faces. A method for qualitative determination based on detection ofcharge-transfer complexes formed between the explosive/explosive residuer and a visualization reagent applied to the surface has been evaluated inthe field at two army ammunition plants. A method for quantitativedetermination based on solvent extraction of samples collected in thefield followed by high pressure liquid chromatographic analysis has beenevaluated on samples prepared in the laboratory, and found to give pro-mising results. The theoretical and practical feasibility of anothermethod for qualitative determinatJon based on UV irradiation of a suspectedcontaminated surface with subsequent detection of any explosives/explosiveresidues present has been demonstrated. This approach may provide themeans for "scanning" an area to determine whether explosives are presenton a real-time basis. Further development of this approach is recommended.No methods for inorganic species which would represent substantiveimprovement over existing qualitative and quantitative methods wereidentified.

Detailed discussions of these areas of investigation are presented inthe remaining sections of this report.

L

3

Arthur [) Little. Inc

II. SAýMPLING AND ANALYSIS METHODS SELECTION

A. INTRODUCTION:

The available literature on the sampling and chemical analysis of ex-

plosives/explosive residues was reviewed to provide an overview of

approaches for further development. The findings of that revicw are

described below.

B. LITERATURE SEARCH:

Work was initiated by conducting a literature search to identify exist-ing sampling and analytical methods for explosives which might be

applicable to the particular problem of detecting and determining thosesame materials on building materials surfaces.

1. Approach.

There were basically two problems that had to be resolved in order todevelop appropriate strategies for the computerized literature search-ings: (1) if the search strategy involved the fifteen defined analytes

in combination with the defined surfaces, as well as with general termssuch as surface sampling, recovery, sample selection, detection, etc.,the search results were nil; and (2) if the search strategy involvedthe use of such general terms as explosives, ammunition, and dynamite,as well as the Chemical Abstracts Section 50 (Explosives and Propellants),combined with terms such as analysis, sampling, recovery, identification,determination, etc., the search results were not applicable to this study.

The techniques used in the forensic sciences appeared appropriate for therecovery and analysis of the analytes under consideration; therefore, thesearch strategy that yielded the most relevant citations involved thecombination of the Chemical Abstracts Forensic Analysis Subsection, theRegistry Numbers of the 15 analytes, and the terms explosives, dynamite,and ammunition.

2. Sources and Coverage.

The search for applicable techniques for surface sampling for explosives/explosive by-products has been performed on:

Source Coverage Number of Citations

Chemical Abstracts 1967 - present 164L

The 164 citations were reviewed for duplication and extraneous material

and were reduced to an output of 45 citations. This output was reviewed

and the 45 citations for document retrieval.

Arthur D Little, Inc

In addition to the computerized search of Chemical Abstracts, a manualreview of d National Technical Information Service (NTIS) bibliographywith abstracts, entitled Pollution Caused by Ammunition Manufacturing,was undertaken. Of 237 abstracts, 9 appeared appropriate for reviewand were ordered.

The following computerized data bases were also searched:

APTIC

ENVIROLINEPOLLUTION ABSTRACTSSCISEARCVNTISCRIS (Current Research Information Systems)SSIE Current ResearchCONFERENCE PAPERS INDEX

3. Results.

The results of the literature search ire summarized in Tables II-1 through11-4. Table II-I is a bibliography of useful citations including (1)the authors' abstracts or brief summaries taken directly from the article,describing thA purpose and results of the work described in the article,and (2) comments describing potential applications of the work describedin each article to this study; Table 11-2 lists the analytes of interestin this study for which sampling and/or analysis procedures are describedin each article with detection limits, when included in the article,listed under the respective headings in the table; Table 11-3 lists theanalytical methods used or recommended in each article for determinationof analytes of interest in this study; Table 11-4 lists (1) the samplingprocedures used or recommended in each article for analytes of interest

OF in this study, (2) any additional sample preparation procedures requiredsubsequent to sampling and prior to analysis, and (3) any interferencesdescribed in the article with the sampling and/or analytical procedures.

Arthur l) I.,tt! Inc

00

I.

6. I z

- -a

ip I

aa

.'0 S

-60

IL tt.Lhi19~v b~ webe

Ll 01 1 V--e-i

Ti 06t~ aI I-UU

Ie . ,

Arhu V. Lite a

I' r

6)~!. I - .6)

N C 0~ CiC I

6) 62, IQD6w4)0 C s

fa,.~ ccf. 0

140 4,-

aIW 41 0

01 . U6 0 )1I 6

= 0. Z6 6 1

be 3,*

mj 6) W Q)4 " , T;4 - >~IL c~ m .0!x Ct .0 -0 -

Is) 06 CCI 6) c a -m - 4)bem- m .

66'C . Q WO a )I06 .C a ce-l- a a J:Z4)

0 4 3 0a C1

;-' 0 -64 -Q 41 2'0 . C )- ) 4C 4 104 fl.. . 6 6 0 ) u

to :-.) C fC)).6 . 0cm

14600

S - c w m 41 c" 421 1c 6)I Z' e-C - r M C IU-6,~ -- 0j 0) 0. hJ "m

P-4 u -C.,* 4i f.1Uo c* 6,~ z66 v

)1 C 4 U 66)4D ... 6)fC 6) j 4).4', U.

C C C 0 .,.a Co ' U 6 CJ C w4, in .. 21j~ - 0 t6) 0 v 6 45 43 *600,

6)6)0T 6). 0) 4o 2 16 0C6-6 6) a a0~ -0w m6 ) :: Z .0. .' ) r ,s mo--G 0. c C

CL t3cL' C z xz v- ~ 6 ) g ; " 0 L . - -A)6 *6C. .00 2. w -I_

-a 0

6)- cc 6, "c .= u . . - 4dL -=. 9 L:

0.-

~~£ U~~ k6.- ~ ~ * ~6 6

*~rtu D. Lite Inc..~ .

a 20-404,

00-

-2

.. " 0 $ .411

X .,, 1,. ",-;_2

.1 'Z 4Ao 3N£ *40 b & ~ f B 5 ~

-~Atu D,~ -tc In ~ -

Io,4 .6

at a-Zoo I04

N IIn-~~-xI w - g,~:w~i w c 6~

O -

_6 1

I~ ~ rI:L

44 -e

... hu D. LE.OC .- U£

4) C

Uý C 0.

-0 4oe 0. x07

CDC

0 m40 ;K 41- ) o W

4. 02 I4

IV 401 m . -41 10 41 ~ ~

C: 60U2. 4.0 2, _ '. 0 C

C 0 U. 06 .

6.2 = * Z L. a, 4 1 c 0 cc 0uvO j~

4,-w- ~ c~V 4w me UN2.C O.C 10 4j4 6C 6 C r>

do a a2 - a, li w ~

0 c -eC e im U z4 -'A J

IX .0OCu ~ J ~ u ~~O-u ~ ~ jCC~U '-C2 . . ~ ~ S C..':0

o- 'c4

a (U

060

~~~ 0 .2os r- c. C

Arthu D. Little,

X 'a -

C' a. C u w

wm14 CL 4

wc c a- ' cc; 0P C34. -. 0go cC Q~

1-4 c.- 0

ýQ C4-Cu3C

m, ej 4 ,c

6 x c

ai 4w 41 u .e4

ZC - . 1 Q U - - ~ 4'1 1 44'

4-0 90 -4,~. .to

'041 20 , 6 0 u -

i!I - " -- L0-

V 41 &- .. 30.'4 0 'a C 'U' 0. j1 1 4; E'a. 4.*04

we ..4-. -- c.' ~ 4.0 a *4.C- 00)-4 4 4 ~ 40 1 1 4 4CC C U - 4o .0 . 4- V4 0 m a00 0..CL 4 T4p r

C 4 .

4 1 N1 41 -C ~ .UJ -0U 'c 1. 0 446, C6D03 4 0. ~ - - -~~

'04UN L .I 4

-, 4-. 0-'~

4-' C.0

r ~ 6 -7 0

12- Arhu D. Lite In

C: 0S* 4o

V CL, C 0Nh.~ Ci- -b0- .

'C 4, -

:~F 1 6u,~ 0 -

'IN 'Cla 5 t -0 - )

-. 11 C' 'CQ At - !. A EC0 . -

0h~ C ý ~ CN -1

~~W 4. ccCS N ~ N

0h.) I0 C6S a N zEO 1. -Oh1 C -. i. jag - Qh U

w 6~4 _C _w. 72 OD4,~~j GJr ,4. 04

U~: go C CL 7,5 N£.e.- .,C.9e 2,0 -*2 LO5C vN a Cf20-

w ZC ZI) Q6 41oi0, 0d

C- m 5jC1

V = 5- 6C -4,C 6

-1CD1 6,4 C .C 4 2L =41 f 4, . 'C

~~---~ C - C-P4 I. .Z w 41t, - V, N- C4

C 41CC 55 C.L~ Z: 0 6,I jj:~E0;;4, 04 I" Ca* SC C

c0c 0 C 1 41

1. Is . or CCI -C -04 0 w~h U 'x C- C 0, C05~~~~~ to NC . 5 40 0A ~ 4, 5 3 - -

C.L.N U.N CO C. D SCN -4j *.

- ~~~ ~ I 5 5 NC -1 N y 4; C V NI sý

I S.

C a

al - 's, ! .

*~6 z~ 4). *

0~~1 Arhu D. Litle Inc N, .~~ -g

4j 4qZi

a c b. C.

- 411 4 )G

m4 1. w c 40X. c 0 CL .

-~. 4b1 44 to4 .x 9. c4' U4 14 Aou -0 1 4 .

%.i ~i. I 0f

.0' L.4 11 amo P% b

P-4 414 rAc0 'a

-41 " -elF-W~41 a )I-N1 j9 ,'

C WG C& f X1

0 Q c1 4 0 IF4A. I. a C 91~ 4) 0 A9

41 Q 0 a4 c--~Id4 4.41- -I '

a c 1410 41. "C .41 r2 TU 6 41 , ' C-.64,- .

* 14 so 0 u ~ h. 41 C41 6 1 r 4.

204. i4141 to .40

9- 9L 401 -c

W 0 C6 I-, w 04 J2 cp.0 J.g 241 60.I a ~ a a

oV :0U 1 60 a6 -- 416 w.A 46 aC. 4 -7I a

-9 :I.I 06 46 : 1~ I.. i 06 6-44 41co w C 4 011g

co c i --!i - -v-1-

-3u

41 60 or- 1-. o

1.41 *41 41 04a- 4' 6,U 4C V£4 V 44 A

*~ .4. 41 4 '.14 4 -6w 9.U1 -. 41 -a a1441~ 4 ~ 4~q -l. . .'- .c

ýI 414. 1. 0. 1 *L44 (6 06

S6. - 0.-C141 .4

c .1.41 14 "7 m

44

14 AthurLi L~tIC In

0~

IS/

4C 0 C4/-

so x~

w- -cc,Z 0Uh

4r=1

Arhr1 Lti n

'-i r ! .-

C- 4.i

- --: M e

It U

H.7: .S--

t ,.. -I -4

- -

* ,

j . .zV

* A

z]- • -. AturD itlr nq

i !i

'7, I *

- k.

ii L

EI -• ,,....I z :'. i... . "a.-.,.a

I, LIh

-- _

. !, * • .' -

I..~ L.,I'.

i-I

Is

!L1.7 A rthur D . Little, I nl(

I.

,.,1

S. - __

C:_u:.,.§

o -- •-

¥ - - '

-. a. C,•• • - :

- a

-' -I--,. ,• ,-

- -

z:c

2"Atu D i~ie n.

I' . .

or t

,61

I=.4

-.

.6 ..- _

;do3

-• . ,.. Vt

6 0 C

K- ;5'L I! 4' ;;"['7 . - .•

Q..

- -•-", ,

9rt

44'

P-4.

C-, -0 a

19 Arthur D. Litle -Inc.

1

I

2'he.gI

L� .�.

0

zo - b- II-. - 4.- 2I-.- 3. 1.C-,

L

m � 1.!� I- L --� I, *�

4. 4.. ':4m- I.i.� S -.

H IV

C"a

A

z p 1o r-C

I' bib.F .1 18.,.-

D.. ma *

CV

4. 1 S F- A 1- ,�

.� �S 'I .. --

- &. - I --J

C � �j I '� 113II A-

-' m � .� - IS.. ml. fr *

� ; $ � � : 'H '� V TI �- r.I-

� �L S - ,. *-. 4 .4 4 .4 -

L At hut [). Little, JIlL,

C. REVIEW OF SPOT TEST CHARACTERISTICS:

AL the raquest of the Technical Project Officer, a review of the spottest# used for detection of explosives/explosive residues was carriedout concurrent with~ the literature search to determine whether thosetoost could be used for in-situ detection of explosives on the surfacetypes of interest in this study.

Tableo 71-5 through 11-8 list and describe selected characteristics of spotP teats which may affect their suitability for use for qualitative analysis

of explosiveu/explosivo residues oil the surface types of interest. Tables 11-5and 11-6 list spot teot methods and characteristics by analyte. Tables 11-7and 11-8 list spot teas methods by reagents used in the respective methods,and lint and describe solected properties and characteristics which may beeiiXhibited by those reagants when used under the conditions specified inthe m•thod,

#u evalunte the data Included in these tables, we defined a set of minimumcriteria which a spot toot proposed for use for the contemplated qualitativeanalysis should witisfy. Theso criteria were: (1) the spot test shouldpermit rapid screening of large areas; (2) it should be sensitive down to theagreed upon detection limit of 0.5 ug of anslyte/cm2 under various surfacecondiLlonb; (3) it should be specific for the analyte of interest; (4) itshould not result In an irrGversIble chemical reaction which would make theonslyLe unavailable for subsequent quantitative tesring; (5) its use shouldntf. pre "tit an ununual hazard to the operator due to the toxicity, flamma-bility. or other hazardous characteristics of reagents, rcaction products,i i-st.LIun cundLfiuns, and; (6) it should not result in a net increase inthe amount of CofltamJt:.!tifLn present in and on the surface tested and shouldnot in any other way affect the suitability of that surface for any projectedfuture use.

p The overall finding of this evaluation based on the available data was thatthtee Is no single spot teat or combination of a few such tests which satisfiesell of these criterla. More significantly, most available spot teststypIcally fail totally to moet one or more of these criteria. It is possiblethat modifications could be made to available tests to improve theirperformance in chis regard. However, the developmental effort required forthat purpose may be substantial, and it io not clear that improvement inthe psrforpialUce Of a given spot text with respect to one criterion wouldbe accompanind by similar improvement in other areas.

Lii opu=r*Y. our view is that avnil~ble spot tests do not satisfy minimumcriLezis for qualitative analycis of explosives/explosive residues onhuiJdlng material surfaces. Ve do not recommend that developmental effortbe expended on spot test methods except for those analytes or conditionsfur which nu more- or equally-promising methoda can be identified anddoviluped.

21

Arthur I) Littlc. Inc

Notes for Cclorimetric Spot Test Tables

Literature numbers:

1 through 47 correspond to Journal articles 1 through 47 listedin Table II-1.

F is Feigl, F. and Anger, V. Spot Tests in Inorganic Analysis,Elsevier Publishing Co., Amsterdam (1972) (6th English Edition).

"2 is Feigl, F. and Anger, V. Spot Tests in Organic Analysis,Elsevier Publishing Co., Amsterdam (1966) (7th English Edition).

(a) is Christensen, H.E. and Luginbyhl, T.T. (eds) Toxic SubstancesList: 1974 Edition, U.S. Department of Health, Education andWelfare, Public Health Service, Center for Disease Control,National Institute for Occupational Health and Safety,Rockville, Maryland, June (1974).

(b) is Sax, I.N. Dangerous Properties of Industrial Materials,Reinhold Book Corporation, New York (1968).

(c) is International Technical Information Institute (ITI)Toxic and Hazardous: Industrial Chemicals Safety Manual,Tokyo (1976).

Abbreviations:

cat - catdog - doggpg - guinea pigh~m - humaninh - inhalationipr - intraperitonealivn - intravenousLD5O - lowest dose 50% killLDLo - lowest published lethal doseLC50 - lethal concentration 50% killLDLo - lowest published lethal concentrationinky - monkeymus - mouseorl - oralrat - ratrbt - rabbitscu - subcutaneousTVL - threshhold limit value; the concentration

of an airborne constituent to which workersmay be exposed repeatedly without adverse effects.

22 Arthur D i-ttle. Inc.

II I I

, 11

x

to 4 - F-4-

M .I -r -

SI S ..... -4- -.. .. ' . I K - , d

-- 16

v I

-. . Ar D.

S-; -.- ,, , I

': : • I I"-

'4 - 2 .'-.

Arthur-D .Little,:I 3

-.. . . . .. . .i

. -I

- I

- .. 3 1 I I

* -. I . I -. -.. "

u1 - A" •

*-- - - -,, ,

- -

I. ! -i-- -- I"

I .I

] - --

i f ,• . .!Arhu D itle I.

C 4

i d .-- L

-.--- --- I -• -

to.

- _ *

- •

C

N

-! -- ---

-rU

-

-

- -

- --

-I -

I aa

-I

40

e' cc

P-.=4

CIIPC

I

A2 5

A r t hrID .- I**l , I

6 a

z . * h.*h

? ~ -'-- .-

a Z.

- . - - --

.. . , '. ,: - --. • .*

P-0-

at Q

Arthur- •- 1. , 1n,.r226

- --. . .. - . o . . .. ..- I

I

733

Z Q-

.0a

ImI

dcI

27 Af thu )Iitv jK

r I ! .i -

Z t I ii 3

I -_

S• i.. A I .M ',I

I z I

.2 -1

-- -- 1±-... - -i i i

.'. ,

t8 A r7,

11

id w w

IA -

1.4

-. - I -

-~~ -- 0 P.

±1I - -, .4-~'

.1 -- ~- -~

28 Arhu D. Lite Inc

r-

.9-

K'Ax

e4_ ;__ _ _j__ _ _ _ _

C. z

29 Arhu D. LiteIc

b.22'i• I U -: -,

a 0

_6" = 'A A 3'; *n adL

* Z

-I - I , I

-i

i i i a

r I

_- _ •.* ,I*-• -U.

o Cd-

a , I•

--"

-7

30 Arhur D. Litte, I

,,,,Z.

x. J

-1*- I -

0-4i

I -

-, 2i -c I eN .CA V, -

Atu D. Lite I

i" •II _ _-__ _-_ _ _ _ __I

, 0 • i ICz. - I I iI ~ i -

I -I 0 -i -

- _S. a - .:., . ,

_ - - - -I

3- , Arthur D. Little, Incl

TABLE 11-8. REAGENTS FOR INORGANIC ANALYTES

;#ln sl t AI011t In. 04CetrAtK ti"80 ADPLIC&tLOA ILPLCI? IX - PlAMab I '.t! Daq

.~ lzai ented Cc- Cant!0 Iie1 :3 A. Ccý &Ahod outa 'hosting

.4.aLI -foIaIL-'7 ic Acid.

-r oxItd t yNt cap o to L z - r

Acetc1 ,CI ar j neoesceI L.t* (b)(b ICAhn abi

:Crioo..s

@I~~~rldo tDL.*OI I

I0 I4 k

- I II I b~h I3

q..~ *t l vLp om lkaLl nstOL-I ct inleEI Ms" -1ýtob-v$ X-- UiihCI4h .zplou

alal I~rL-~f eri-a 1-3 1 h~azard (b)

(c) )M iL oe~~

tICbe Ilv g

402 !otsald*IIlde I

hos -- . _ _

C'06 For,.S 10CSLlorj

Ilt - 5 ;.P In sit -uses. hI I c)(b) 1=bi b rub',.? v;I0. (cl

4ro.~AJj) i &, colored Rfi no-

I orl0-rat S ,:OC.,Lut

k-st Acidi ?b 'AI

32Arthur D. Little, Inc.

p

D. FINDINGS AND CONCLUSIONS:

Among the overall findings of the literature search and spot test revieware the following:

* The literature search indicates that most of the prior work inthis area has been done by law enforcement agencies, and thoseagencies have, for the most part, concentrated on tCe application

P of spot test methods for the detection of explosives: quantitativedetermination of explosives has not been a major concern, exceptto the extent that these investigators have attempted todemonstrate the range of concentrations to which their qualitativedetection methods could be applied. The Army, also, has reliedmainly on spot test methods up Lo now in explosives contaminationassessment work. however, as noted above, available spot testsdo not satisfy minimum criteria for qualitative analysis ofexplosives/explosive residues on building material surfaces.

# Other investigators have reported explosives determinationtechniques involving analytical methods ranging in sophisticationfrom visual examination of residue under low power magnificationto mass spectrometry, including in between almost all readilyavailable modern analytical techniques. Sampling for quantitativeanalysis remains a weak spot. Thus, different methods for qualitativeand quantitative analysis may be required to satisfy the Army's

r requirements.@ The use of a qualitative survey obviously must not preclude the sub-

sequent use of a quantitative method if two different methods areused for qualitative and quantitative analysis, application of thequalitative method should not result in the destruction or loss ofanalyte. Many of the methods described in articles collected inthe literature review result in loss or destruction of analyteand thus may not be applicable to the present problem.

* Recommended sampling and analytical methods should not result ina net increase in the amount of contamination through the applica-tion of toxic or hazardous reagents to the surface to *"e terced,nor should they present any other hazard to the operator. Manyof the materials described in the literature review require theuse of toxic reagents and thus for this reason also may not beapplicable to the present problem.

* Surface contamination of the type the Army is concerned with is mostlikely to have resulted from spillage or dusting of solids or fromspilla-,e of process liquors or liquid wastes. The latter situationpresents the problem of sampling compounds which may have penetratedinto porous surfaces such as wood or concrete. Most existing methodsdo not address this issue.

P

33

Arthur D Littl. Inc

r

E. SA.MPLING PROTOCOL SELECTION:

Based on the review of the available literature and discussions with theTechnical Project Officer and other persons having expertise In traceanalysis and explosives technology, the following sampling protocolswere identified as candidates for developmentaJ teoting.:

1, continuous monitoring of organic analyto vapors using aportable &a& chromtograph;

2. evaluation of existing U.S. Army equipment developed forthu vapor phase detection of CW agents;

3. in-situ formation of charge-tranafer complex*s withvisual identification;

4. UV irradiation of suspected contaminaLed surf4ces withsubsequent detection based on thermal limaging or UVphotography of the Irradiated surtfre;

5. Solvent extracLion using alterinative procedures toconventional wipe or swab methods.

Protocols 1, 2, 3, and 4 were candidates for jualitatlve detection ofexplosive-:/explosivev residues; proLocol 5, com1oinewith modifiedversions of existing USATHAMA methods for the determination of explosivuswas intended for nuantttative determination. Detailed discussion. ofthe developmental testing of these qualitative and quanLitativa methods arepresented in the respective sectiono a; this report.

3r.

MIMIu I) IIttIC Irn'.

-III. QUALITATIVE METHODS DEVELOPMENT

A. INTRODUCTION:

R The qualitative methods selected for developmental testing included:1. continuous monitoring of organic analyte vapors using

a portable gas chromatograph;

2. evaluation of existing U.S. Army equipment developed foror the vapor phase detection of CW agents;

3. in-situ formation of change-transfer complexes withvisual identification;

4. UV irradiation of suspected contaminated surfaces with"- subsequent detection based on thermal imaging or UV

photography.

Detailed discussions of the developmental testing of these methods arepresented in the respective sections below.

The objective of thin testing was development of procedures for the rapidqualitative d4termination with 90% confidence of the presence/absence ofthe compounds.of interest Jown to a level of 5 ug/lO cm2 in a givenbuilding. The approach used to achieve this objective involved in eachcase the spiking wiLh known amounts of analytes of new, uncontaminated

L samples of each of the surface types of interest obtained from buildingmaterials dealers. Samples of conductive non-sparking flooring were notavailable foi this purpose.

Practical determination of the confidence interval associated with aqualitative analysis performed in the field would have required access to

I. a building where the nature and extent of contamination was known. Thiscondition could not be satisfied by those AAP's to which access wasobtained. Therefore, detection limits for positive compound identificationare reported.

At the direction of the Technical Project Officer, emphasis was placed onthe development of procedures for organic analytes. No methods for in-organic species which would represent substantive improvement over existingspot test methods were finally identified.

B. ANALYTE DETECTION USING CONTINUOUS VAPOR PHASE MONITORING:

The detection of explosive vapors using gas chromatography and otheranalytical techniques has been the object of considerable investigativeeffort. In fact, several of the commercially-available explosives detectorsare based on this principle. Available explosives vapor detection methodshave been used with varying degrees of success for the detection of bulkexplosives as, for example, in airport and aircraft security operations.

Arthur D Little, Inc, " ,• ..,',• - v •- -' = • ' ... . .-" "' .• " ' -l n • . . • " : - - ' - 1 • 1 -' . . . . . . . . . .

Detection of explosives/explosive residues on building materials surfacesusing continuous vapor phase monitoring could offer the following advantages.

* By means of one or a few measurements, all the surfacescomposing an entire enclosed area could be effectivelyscreened for the presence of explosives;

a The analyses could be performed on or close to a real-timebasis, making it possible to conduct "walk-through" surveys.

Unfortunately, however, there are several potential difficulties withusing continuous vapor phase monitoring for detection of explosives onsurfaces, including the following:

* The concentrations of analytes on surfaces which are of interestin this study are very low;

* Further, the vapor pressures of most of the analytes of interestare very low under amlient conditions;

0 Most of the analytes of interest are strong.A.y polar ani thusadsorb strongly on surfaces with which they come in contact.

These factors have, in fact, largely precluded the effective use of vaporphase detection methods for many applications where the detection of otherthan bulk quantities of explosives was attempted.

Two types of recently-developed analytical instruments were evaluated in

this study to determine whether their operating and performance character-istics made possible the detection of vapors derived from explosives inair immediately adjacent to surfaces spiked with explosives/explosiveresidues. One was the Photovac lOAlO Portable CGs Chromatograph (Photovac,

Inc., Thornhill, Ontario, Canada), a portable gas chromatograph with aphotoionization detector for which detection levels down to parts perbillion for various compounds including nitro-compounds are claimed. Atechnical representative of Photovac, Inc. visited Arthur D. Little, Inc.laboratories to discuss the application of the Phctovac 1OAl0 PortableGas Chromatograph to continuous monitoring of organic explosive vapors andto demonstrate the performance of that instrument used in its continuousmonitoring mode for sampling of air immediately above a surface spikedwith known amounds of selected explosives. During that demonstration,the Photovac lOA10 failed to give any observable signal when used to samplethe air immediately above t~e bottoms of Pyrex glass beakers spiked with

the equivalent of 125 ig/cm of RDX, 2,4-DNT, and DPA.

The Arthur D. Little, Inc. Project Manager and the Technical ProjectOfficer also visited U.S. Army C.S.L. Laboratories to view and assess theperformance of U.S. Army CW agent field detection equipment when used tosample explosives vapors. Three ionization detectors, including the modelin current use and two prototype instruments, were evaluated by samplingthe air immediately above the bottoms of Pyrex glass beakers spiked withvarious concentrations of RDX, 2,4-DNT, and DPA. Each of the detectors gavewhat appeared to be an obqervable response at or near the agreed-upon detection

36


i2

limit of 0.5 jig/cm2 . However, in no case was that response sufficientlylarge to assure unambiguous detection. It was judged that none of theseinstruments appears at this time to possess sufficient sensitivity towarrant further investigation in this project.

C. ANALYTE DETECTION USING FORMATION OF Ci!RGE TRANFER COI' MEXESWITH VISUAL IDENTIFICATION:

1. Background..

Charge-transfer complexes, particularly molecular additior .c'" V "- ,

have been used for years in the isolation, purificatlon a.a6 id,! l:ationof organic compounds. Among the better known examples arn :he Atl*xcompounds of 2,4,6-trinitrophenol (picric acid), l,3,5-tr .. trobeiz;-rvand 2,4,7-trinicrofluorenome. These nitroaromatic compoum.ý a~e cl3args.transfer "electron-acceptors." Their complexes with "elvtron-dor rs'are usually formed in 1:1 ratios of acceptor:donor, and th c,',i .:ionexhibits properties differing from the ind4.vidual c-mponencs, e.g.,color, crystal structure, melting ý.)int, etc. The oond 4.rength of theaddition compounds varies from very weak--on the order of van -lo- •aa1 'sforces--to moderate strength such as in hydrogen bonding. The 6'mplk.aromatic hydrocarbons, e.g., durene, naphthalene, and anthracer , aretypical donor compounds whose complexes have the weaker type of bondssuch that the solid complexes can be readily dissociated by lot , of thedonor through volatilization at or only slightly above room temperature.Solvent action can also be used to remove the donor compound from thecomplex.

In the interaction of nitroaromatic hydrocarbons with the simpleraromatic materials, visible color due to complex formation would bethe simplest method of detection. In the absence of a good colorcontrast, the known property of nitroaromatics to quench thefluorescence of the other aromatic compounds could be used as the basisfor a less simple detection scheme. Of the donor compounds mentionedabove, however, only anthracene has a fluorescence at visible wavelengths such that it could be used without an instrumental detector.Still another combination of a physico-chemical interaction of afluorescent donor with the quenching nitroaromatic acceptor compoundas part of an energy-transfer system might also be used to increasethe sensitivity of detection. Work performed by Arthur D. Little, Inc.on detection of polycyclic aromatic hydrocarbons indicates that lessthan nanogram quantities of anthracene can be detected by virtue ofthat compound's energized fluorescence. The interference of nitro-aromatic compounds with such a test should thus permit their detectionat a corresponding level (cf. Interagency Energy/Environmental ReportEPA 600/7-78-182, September 1978).

In summary, in-situ formation of charge-cransfer complexes for thedetection of explosives on building materials surfaces offers thefollowing advantages:

' 3 7 Arthur D Little I,-c

r

(1) Sensitivity down to the desired detection limit of0.5 jg of analyte/cm2 ;

(2) Speed: the chemical reactions used in this applicationoccur under ambient conditions; relatively easy andstraightforward methods for dispersal of the requiredreagents over large areas of the surface types of interestare available; and visual observation of colors orquenching is used as the detection method;

(3) Manageable hazard during and subsequent to use ascompared to similar spot test methods; and

(4) Reversibility: the charge-transfer complexes used in this appli-cation can be destroyed relatively eaisily leaving the originalanalyte intact and available for suLsequent quantitativetesting.

2. ?reliminary Experiments

Whatman No. 42 filter paper was used as a substrate in place of samplesof the actual surface types of interest in initial spiking experiments.Filter paper or the equivalent is customarily used in this type ofwork since it provides a convenient means for manipulating the smallamounts of chemicals involved and also provides a nearly ideal visualbackground for visualizing the colors or fluorescence of reactionproducts.

Acetonitrile solutions of three of the organic analytes--2,4,6-TNT,P.2,4-DNT, and RDX--were applied to Whatman No. 42 filter paper in

quantities sufficient to yield concentrations equivalent to 0.5 Pgand 50 ug of analyte/cm2 (- Ix and lOx the agreed-upon detection limitof 0.5 jig/cm2). The acetonitrile was allowed to evaporate, and the spikedfilter paper was then treated with cotton swabs which had been immersedin acetonitrile solutions containing 100 uig/mL of the electron-donorcompounds durene, hexamethylbenzene, naphthalene, and anthracene.

The filter paper was then examined for (1) the presence of coloredreaction products, or (2) quenching of reagent fluorescence when examinedunder a UV lamp. First priority was assigned to identifying a reagentwhich gave a colored reaction product which, at all concentrations,was clearly visible to the unaided Eye under ambient lighting conditionsand was clearly distinguishable from any potentially interfering colorsor surface irregularities. Detection of a reaction by observation ofquenching of reagent fluorescence was considered acceptable only ifa colored reaction product could be identified; in that case, theobserved quenching should satisfy the same criteria as above.

38

Arthur Dhittle In,

I -. - - . -j--~- ,- - - - . - - -- -

ri

In an effort to identify an electron-donor compound which would yielda colored reaction product with these analytes, the compoundsN,N-dimethylaniline and diphenylamine were also evaluated using theprocedures described above. While diphenylamine is also an analytein this program, it was nevertheless considered useful to evaluateits electron-donor properties since it is often cited in the literatureas a strong electron-donor and could thus serve as a useful modelcompound.

At the 50 vg/cm2 level, 2,4,6-TNT formed a reddish-orange spot withboth N,N-dimethylaniline and diphenylamine; 2,4-DNT formed a pale yellowspot with diphenylamine only; and RDX failed to form an observable coloredspot with either compound. At the 0.5 jig/cm2 level, none of the analytes

-, formed an observable colored spot with either compound. However, whenspiked filter paper treated with anthracene was observed under ultra-violet illumination (254 nm), quenching of the anthracene fluorescencewas observed for all analytes at both concentrations.

On the basis of these initial positive findings, additional experimentswere performed on the nine analytes listed below:

Analytes Tested

NG NitroglycerinPETN PentaerythritetetranitrateRDX CyclotrimethylenetrinitramineTNT 2,4,6-trinitrotolueneTNB 1,3,5-trinitrobenzene2,4-DNT 2,4-dinitroLoluene2,6-DNT 2,6-dinitrotolueneDNP 2,4-dinitrophenolTetryl 2,4,6-trinitrophenylmethylnitramine

Each analyte was spiked on Whatman No. 42 paper at levels correspondingto 0.5, 1.0 and 10OX the specified detection limit (0.5 pg/cm2 ).As in the preceding experiments, cotton swabs which had been immersedin an acetonitrile solution containing 100 ug/mL of anthracene wasgently drawn across the filter paper. After the acetonitrile hadevaporated, the paper was examined under 254 nm UV illumination.

All of the analytes except NG and PETN were detecr.ed at the 10OX levelwithout appl.-cation of anthracene; the detections were as dark spots onthe lighL paper background. (In visible white light, the DNP and Tetrylspots were yellow.) When treated with anthracene, all of the analytespots were seen as dark spots; on the fluorescent anthracene background.

39

Arthur I) L•t•ic. .ncS. .-~A~(-_

At the iX (0.5 wg/cm2 ) level, again only NG and PETN spots were not

evident until treated with anthracene. Even at this level, they didshow a weak quench of the anthracene fluorescence. The other compounds,evident without anthracene, showed increased contrast (dark/white)with the reagent.

At the 1/2X (0.25 ijg/cm2 ) level, none of the materials were detected. until treated with anthracene. PETN, NG, and RDX detections are

questionable at that level. Taken together, these findings suggestedthat detection of the three analytes tested at the agreed-upon detectionlimit of 0.5 iig/cm' was well within the capabilities of the technique.

3. Formation of Charge-Transfer Complexes Directly on Surfaces.

In an additional set of experiments, analytes were spiked directlyon clean samples of the surface types of interest. The acetonitrileianthracene solution was sprayed on the spiked surface sample using aspray bottle. In each case in which the quenching of the anthracenewas positively identified, a lower concentratiov of the analyte wasspiked on a clean sample of the surface type being examined on theexperiment repeated. The lowest concentration at which quenching of theanthracene fluorescence could be positively ident.ified was taken asthe detection limit ior that analyte-surface combination, The resultingdetection limits are listed in Table IllI-.

An additional finding of this work was that several spray applications(i.e., 3-4) of the anthracene/naphthalene reagent are necessary toachieve a uniform fluorescence background on brick and concrete. Witha single application of reagent, surface irregularities appear asslightly darker areas and cannot be distinguished from analyte.

It was also observed in these experiments that the presence of dirt orother foreign matter on metal and wood surfaces prevented uniformspreading of the reagent over the surface. Also, care is required onmetal and wood surf4ces to avoid puddling or running of the reagent,which may result in removal from or dilution within the ezea beingexamined of any analyte present.

4. Solvent Lift Technique for S~amling of Surfacss.

Tre analyte detection limits shown in Table 111-1 are higher for allanalyte-surface combinations than for the same analytem on filter paper,In an effort to determine whether lower detection limits could beobtained. alternative procedures for isolating the atnlyte from a surfaceprior to treatment with the acetonitrile/anthracene solution wereevaluated.

40

Arlt Ir i ) littiI :I•.

as4 % ) C) 4

1.3 0 0 00C0 0 I-

z U.4.0 0

0j 0 m

04,,sC

., 0)4 0 %7 %

4J4 46) "4 wC

411

0~" 0 .

di

oj .94 4141 j w o

I.. 4) 4) 4w 4

P- = a' a F4 0

P4 (f 4) w0~~l 4. 0 O C .

*~~P * v 4) ' %

UU0 v w v 4

P.. ca uC -4

-P4~ r.C j4

I--

'0 Ar-4)

Arthur 1)itk' Itic

The best results were obtained using the following procedures:1) Whatman No. 42 9.0 cm filter paper circles were saturated with0.5-10 mL acetonitrile; 2) the wetted filter paper was pressed againstthe surface of interest; 3) the filter paper was allowed to remain inplace until the acetonitrile had evaporated; 4) the filter paper wasremoved and stored in 100 x 15 mm disposable plastic Petri dishes(Fisher Scienctific Co. Cat. No. 8-757-12) until analyzed.

Using these procedures, the detection limits are shown in Table 111-2were obtained. Comparison of these results to those in Table III-I indi-cates that the detection limits obtained using the solvent lift techniqueare lower in almost all cases than those obtained by formation anddetection of charge-transfer complexes directly on surfaces. The sol-vent/lift approach also eliminates the analyte dilution and reagentpuddling problems which may accompany the accidental application ofexcess reagent.

5. Field Evaluation.

During the week of April 19, 1982, visits were made to Holston ArmyAmmunitior Plant in Kingsport, Tennessee and Joliet Army AmmunitionPlant in Joliet, Illinois to evaluate under field conditions thecharge-transfer complex formation solvent lift sampling protocoldescribed above. A total of 195 filter paper lift samples were col-lected at the two installations using the procedures described above.Seventy filter paper lift samples were obtained at Holston AAP and 125were obtained at Joliet AAP. Six locations in five different buildingsat tho two installations were sampled in this manner. Solid sampleswere a±lo collected from four buildings at the two installations. Ineach case, the specific locations sampled were those which personnelfamiliar with present or former manufacturing processes performed atthese installations suspected were contaminated with the explosivesof interest. A complete inventory of all samples is included inTable 111-3.

The variety of buildings and locations sampled represents most of thesampling conditions likely to be encountered in the field. It wasnoted, however, that even among nominally identical manufacturing linesand buildings at a given installation, substantial differences wereencountered between equipment types and configurations, previous types andof intensities of usage, etc. The significance of this observationis that the development of a generalized sampling plan would probablynot be possible or useful.

Due to safety restrictions and the absence of readily accessible ACpower supplies, only a few filter paper lift samples were analyzed inthe field for demonstration purposes. In all other respects, however,the filter paper lift aprroach proved to be easy and efficient to use.All samples were returned to Arthur D. Little, Inc. laboratories andwere analyzed within cwo deeks. Several samples which were suspectedto be contaminated with explosives were reanalyzed at various times over

43

Arthur D Little. Inc

-4

U

C-C)

1.4

C-4

F- 1' 0 0 n 0 0

E-1

Ln 0fL('C U) )

.Atu 1)Ltt n

I-I

TABLE 111-3. INVENTORY OF SAKPLES COLLECTED AT HOLSTON ANDJOLIET ARMY AMMUNITION PLANTS

Site Buildin& Area Sample No. Sample Type

Holston : top of 1 filter paperwooden drain 2 lift

R cover 34

Holston ii bottom of 5 filter paperwooden drain 6 liftcover 7

8

Holston II bottom of 9 filter paper

concrete 10 liftdrain basin 11

Holston Ui wall frame 12 filter paperbase (wooden) 13 lift

Holston Il metal wall 14 filter paper

frame 15 lift

Holston Ui base of 16 filter paperconcrete wall 17 lift

Hoiston 11 two feet from 18 filter paperfloor on liftconcrete wall

Holston Il corner of concrete 19 filter paper

floor 20 lift

Holston Ii one foot from 21 filter paper

floor on a 22 liftconcrete wall 23

24252627

Holston 11 bottom of 28 filter paperwooden door liftframe

Holston Ii one foot up 29 filter paper

on a concrete 30 liftwall 31

3233

SArthur D Lit de Inc

TABLE 111-3. INVENTORY OF SAMPLES COLLECTED AT HOLSTON ANDJOLIET ARNY AMMUNITION PLANTS (CONTINUED)

Site Building Area Sample No. Sample Type

Holston Ii concrete floor 34 filter paper(area covered15 1/2' x 14') 55

Holston Ii concrete pump 56 filter paperbase 4 lift

63

Holston Ii metal strip 64 filter papera long face 65 liftof pump base 66

Holston Ii drain basin 67 white powderpump base 68S concrete

69S concrete70S concrete

Holston D4 flooring 71S asphalte

(RDX wall 72S brickreaction baseboard 73S tar materialbuilding) door frame 74S woodtank base 75S white powder

materialequipment base 76S concretepump base 77S concretepump base 78S concretewall base 79S black tar

"materialdrain base 80S black powder

Joliet TNT secont floor 101 filter papershell melting bay + liftloading area (concrete) 162

Joliet TNT first floor 163 filter papershell loading bay liftloading (concrete) 186

Joliet DNT first floor 187 filter papersweathouse drain area lift

(concrete) 194

Joliet DNT transite 200 filter papersweathouse wall paneling lift

204

Joliet TNT concrr:te 205 filter paperwashhouse floor + lift

209

Arthur D Little Ink46

TABLE 111-3. INVENTORY OF SAMPLES COLLECTED AT HOLSTON ANDJOLIET ARMY AMMUNITION PLANTS (CONTINUED)

Site Buildin& Area Sample No. Sample Type

Joliet TNT concrete 210 filter paper

washhouse sump basin + liftP 213

Joliet Tetryl non sparking 216 filter paper

packaging floor 217 lift218

Joliet Tetryl concrete 219 filter paper

packaging drain basin 2 lift222

Joliet Tetryl non sparking 223 filter paper

packaging floor lift

Joleit Tetryl shaet metal 224 filter paper

dust collector 225 lift

Joliet TNT second floor 3015 wood floor covering

unloading flooring

Joliet Tetryl floor 3025 non sparking floor

packaging

Joliet Tetryl drain basin 3035 white powder

4

4• ~Arhur DLittile no

an additional two-week period and in each case the observed resultswere indistinguiishable, indicating that no observable loss of analyteoccurred upon storage of samples for up ýo one month.

The solid sample 675 collected from a •loor drain in Building I-I atHolston AAP was analyzed separately by x-ray diffraction and was foundto consist mainly of RDX. Acetonitrile washes of several filterpaper lift samples collected from the same general area were alsoanalyzed Independently by a wet chemical method for identification ofRDX (ref. Dept. of the Army Technical Manual TM 9-lo00-214, MilitaryExplosives, No. 1967, p. 12-4) and were found to give positive resultsfor that analyte. Several additional filter paper lift samples fromnearby areas which contained amounts of dirt and debris similar inquantity and appearance to that on the samples described above gaveno indication of contamination. Taken together, these observationssuggest that certain of the areas sampled in Building 1-1 at HolstonAAP were indeed contaminated with RDX, that the charge-transfer complexformation sampling protocol gave a positive indication of contaminationin and around areas where contamination was known to exist, and, inareas further removed from the known contaminated areas, the charge-transfer sampling protocol gave fewer or no indications of contamination.The latter observation, in particular, suggests that the presence ofdirt and debris of the type found throughout the building did notresult in false positive findings. This is precisely the outcome whichhad been desired.

S The results of analyses of all filter lift samples are representeddiagramatically in Figures III-i through 111-6. In those f.gures, smallcircles indicate the locations from which filter paper lift samples werecollected, Open circles 0 represent a finding of no apparentexplosives contamination; crossed circles @ represent a findingthat the location sampled was contaminated with explosives. Findingswl.ich were questionable due to very faint fluorescnece are denoted bythe appropriate notation next to the corresponding location in thefigure.

D. ANALYTE DETECTION USING UV IRRADIATION AND THERMAL IAGING:

1. Background.

An analytical approach utilizing UV irradiation and thermal imagingtechnology was evaluated for its potential applicability to thedetection of trace levels of explosives on building materials surfaces.The principles underlying this approach involved the following:

The area of interest is irradiated with an ultraviolet illuminationsource matched to the absorption chaiacteristics of the analyte. Theradiation that is absorbed must necessarily heat the materiel. A smallchange in temperature should result and thermal imaging technology might

be used to sense that change in temperature, provided that diffusionof the heat into the substrate does not proceed too rapdldy.

48

Arthur D lhittle Inc

*. . . . .

..........

44,

'Jl.._(,9(DQQQQ

o K>C7O

i--4

~c\1 6L,

or~E

Arhu D iteUn

P O OEQ O 0 0

0 0 $ 0 1 00 ED G e•

0

Q0

o-J - J

>' 00

~4.5 0

oQ000O

oooe@@ @o±

50 . . ....

0 G (9)

- overhanging equipment

IGI I

concrete floor

Svery (Tfaint 0

Figure 111-3Joliet AAAP: TNT Loading Area, Ist Floor

51 Arthur D Iiae Inc

CL

- 0) .• o o p

ccc

I.3t

.4 - 0

4 • - 0-

0D () 0

.(n

Cj Dj

0

o : ()0

0 0

7

~4.

I-,

-4 i..- .-

0 () 0

0 0.

Aw

52 ArthurDL~ttk~nc.

446

0.0

004),

L. G j

.~oc

C (~2cc

C 'J - w

u cno-ClT -

53 ArhrDIC.ll-e

0E

cJ --

-U0

00

'100

s *i jaC Uo

54 AthurD I tilo I4

The technical feasibility of this approach was evaluated by a theoreticalanalysis which showed that the estimated temperature increases accompany-ing TJV irradiation of trace levels of the analyr.es on surfaces wereapparently within the measurement capabilities of commercially availabl.thermal imaging instrumentation. Subsequent laboratory experimentsdemonstrated the practical applicability of the approach. The workcompleted in each of these steps is suwarized in the respective

- sections below.

2. Theoretical Analysis uf the Method.

a. Principle. The explosives of interest are known to have absorptionbands in the middle IR and in the UV but Are not fluorescent in theUV. Thum, irradiation of these compounda with UV or IR light of theproper wavelength will result in energy absorption accompanied by arise in temperature. That temperature rise may provide the meansfor deteccing the premence of the explosives on building materlalssurfaces. Because modern instrumentation is capable of precisemeasurement of even very small temperature changes (e.g., AT , O.10C),the method may be extended to the detection of trace leveic ofexplosives. Discrimination of analyte signal from backgroundtemperature fluctuations :an, in theory, be achieved by carefulseloction of the incicent radiation.

A detection metho.i based on this principle would involve illuminatingthe wall wioh UV irradiation of specific wavelengths tuned to the suspectedmaterial's abovrption bands and sensing the temperature rise due to the

absorbed radiation. The absorptiou coefficients are generally so muchgreater in the UV than in the IR that a UV method is to be preferred.To maximize the temperature rise, one could illuminate the wall with anarrow beam of radiation which slowly scanned the wall, The temperaturerise can be acnsed by a Lherwal-infzgLed deLUCtor that scans the wall insynchronism with the illuminating beam. The illuminating benm can beeither chopped or unmodulated. Because the observed temperature changemay be strongly dependent on which beam type is chosen, both methods willbe examined,

b. ,Chppped-Radiation Methods

By meano of a suitable chopper, the incident radiation intensity, I,

as a functiorn of time, t, can be made to have the form shown below:

0

o t

Chopped Incident Radiation as a Function of Time

55 Arthur L) Little. Inc

t,

This function can be represented by the formula

I - +1 (i+cos Wt) (1)0

consi.sting of a steady component

is -I 0 (2)

and an alternating component

I I et (3)a oiwt

where for mathematical convenience we have replaced cos wt by e

If a lock-in amplifier tL. ed to the angular frequency w is used withthe detector, only the effect of the component la is observed. Itis to be noted that since Ia takes on both positive and negativevalues, this component of the intensity implies formally that poweris both put into and taken out of the wall.

If the thin absorbing layer is of thickness d and absorption coefficienta, at the wavelength of the incident beam, the intensity transmittedthrough the layer is

it M (l-R) Ia e-'d . 12 R 1o e-ad eiwt (4)

where R is the reflectance of the film at the irradiation wavelengthand the power per unit area absorbed in the layer is therefore

W 2 o ( -ed) (1)

where the factor e has been suppressed.

"We assume that the transmitted intensity, given by Equation (4), isslowly absorbed by the underlying wall and therefore does notappreciably affect the temperature distribution set up in the wall.

The temperature distribution in the wall is given by cue heat flowequation

K 2 -T C 3PT - iw C PT (6)

where x is distance into the wall, and K, C, p are, respectively, thethermal conductivity, specific heat and density of the wall material.

56""Arthur I I vlc ln.

The solution of Equation 6 is

T - To e4 x (7)

where k is the thermal diffusivity, given by

k - (8)

and T0 is the surface-temperature amplitude.

The heat flux density at x - p is

q -K -LT =K 17 T (9)

x 0

This flux density must equal the power W given by Equation 5. There-fore, the temperature amplitude at the surface is

-cxd) -ad(l-R) I (I e- e (l-R)I( -ad

2K V-- 2

4 KC

' c. Unmodulated-Radiation Method

In the case of an unchopped incident beam the surface temperature iscontrolled by the thermal spreading resistance r into the concretewall. This quantity is given by

rI - ---- (deg/watt) (11)S- RaK

where a is the radius of the incident beam and K, as before, is thethermal conductivity of the concrete wall.

In addition there ic another resistance due to thermal radiation fromthe heated surface. This effect can be represented by a thermalresistance r 2 given by

1r2 - (12)

4 c T3 a2

where E is the thermal emittance of the wall and a is the StefanBoltzmann constant.

57

Arthur 1) ,tlkU IrxL.

Let P be the power delivered to the absorbing layer by the incidentbeam. Then the temperature rise T is given by "Ohm's Law" for heat

(l-R) (1i d) 1 d

T = (I-R) rP = (13)K + 4 ca T

3

a

where 1 -1 + 1 (14)r r1 r 2

The sample calculations below show that the unmodulated-beam methodis greatly be to preferred. Equation (13) applies to the case of anon-scanning beam. The temperature rise will decrease with increasingscan rate. In actual practice, a number of implementations of theprecise methodology are possible. One would be to irradiate arelatively large patch of the wall and view it at the same time.The entire area could be surveyed by moving the irradiated patchand thermal viewer together. Another possibility would be to irradiateonly a small moving spot (more radiation per unit area) and view alarge patch over the time period necessary to irradiate the wholepatch. A third possibility is to view only the irradiated small spotand move the viewer with the subject spot. The precise implementationof the scheme will depend on a number of engineering tradeoffs con-cerning the expected signal strength, time available to survey thearea in question, arA so forth.

d. Sample Calculations. Based on the formulas derived above, the AC andsteady-state temperature changes were calculated using ultravioletirradiation and thermal viewing on concrete and wood. The parametersof the calculation are given together with their sources in Table 111-4. Theabsorption coefficient used in this calcuation was for PETN at awavelength near 2,000 angstroms. This wavelength is shorter than theUV irradiation source peak used for calculation and some effort wouldhave to be made to match the two. The PETN data were obtained fromthe Journal of Physical Chemistry, 77 910, (1973) showing a value ofthe absorption coefficient of 5.6 x-104 cm- 1 between approximately45,000 and 55,000 cm- 1 (near 2000A). TNT in the UVI region (near

* 2325A) has an absorption coefficient similar to that of PETN. Mostof the other parameters are somewhat variable depending on the source,

* as the materials (concrete and wood) are variable in composition.

A spectrum of TNT was run in the infrared which shows a maximum absorptioncoefficient of about 7,000 cm- 1 between 6.2 and 6.7 iim. These data,which are expected to be typical of most of the compounds in the tworegions, are the reason for our choice of the ultraviolet region forsurface irradiation.

58

IrthuI I) (tN:c InC_

TABLE 111-4. PARAMETERS USED IN SAMPLE CALCULATIONS

Parameter Value Source and Comments

R 0.1 Estimate from the Infrareduv

Handbook

• PM 5.6 x 104 cmM J. Phys Chem, 77, 910 (1973)

L cTNT (peak) 1.38 x 105 cm- 1 ARLCD-TR-78025 (1978)

d 0.56 x 10-7 cm Taken from 1 pg/cm2 , IPETN 0 1.77

I 5 x 10- watts/cm2 Estimate by P. von Thuna, ADL

K concree 8 x j1-3 watts/cm deg C AIP Handbook

Kwood 2 x 10- 3 watts/cm deg C AIP Handbook

Cconcrete 0.65 j/gm deg C Marks Standard Handbook for"Mechanical Engineers

Cwood 1.75 J/gm deg C Handbook of Chem & Physics

Pconcrete 1.6 g/cm3 AlP Handbook

0 wood 0.6 g/cm3 API Handbook

a 1 cm

w 6.28 Hz For 1 Hz chopping frequency

0.9 Estimate from related materials

T 300 0K

o, the Stefan Boltzmann Constant 5.67 x0-12 ( watts )

"This value will depend on the precise wavelength, bandwidth,and geometry of UV optics.

59

Arthur 1) I iýt jn,

Using formula (10) for modulated MV irradiation (PEn)S~(1-R)(l-e a-d) I

T 3.0060C (concrete)0 2 / - 0.0100 (wood)

Using formula (13) for unmodulated UV irradiatio (PETN)

S(1-R) (1 - e-' IA T K+ 0 0.300 (concrete)

-- + 4E T3 = 1.10 (wood)a

The values for TNT would be approximately the same as those for PETNas the listed value is at the peak of the absorption band whichextends out to near 2 800A°. The fact that the latter value arewell within the measurement capabilities of commercially availablethermal imaging equipment forms the basis for our conclusion thatthe method is technically feasible.

3. Analyte U`V Absorption Characteristics

UV absorption data for the analyte PETN were presented in the precedingdiscussion. Data on each of the other analytes of interest was alsoassembled and is in Table 111-5.C,t should be noted that most of the data in Table 111-5 are for solutionsof the analytes since these properties are customarily measured and re-ported in that manner. It is not clear that solution data are reliableindicators of the strength or location of UV absorption bands of thesolid samples because these characteristics are often strongly !.nfluencedby the solvent. However, laboratory comparison of UV absorption spectraof the analyte 2,4-DNP in acetonitrile solution and in a KBr pellet andshowed that the wavelengths corresponding to maximum absorption and thecalculated absorption coefficients are quite similar. To the extent thatthis is also the case for the other analytes (and depending on the solvents),the data in Table 111.-5 may, in fact, indicate that relative locationsand strengths of UV absorption bands. In that case, the data in Table 111-5suggest the following:

1. With the exception of NG, all analytes exhibit comparableabsorption strengths (i.e., E and a in Table 111-5 are on the sameorder of magnitude for all analytes). Thus, the proposedapproach should be equally applicable to all analytes;

2. The absorption bands of all analytes are sufficiently close to2500 Angstroms that an illumination source capable of providingsufficient power output over some wavelength range centered on orabout this wavelength may permit detection of all analytes (exceptNG). Commercially available mercury sources which emit thecharacteristic 2537 Angstrom line may be suitable for this purpose.

60


rF,ý

z 00 0uc000 0 0n 00 C0

~ - CI N ~ c -

K zz4-i

A000

C14 C1 1 1 4CN CN NN N NN N~ N- C14 -4~N

0, - -

1-4C - .

14 Z4

-Z-

'-L

7;: *-

- 61

Arhr- itlI n

Table 11-5 (Continued

REFEKENCES

1. "Organic Electconic Spectral Data," Vol. 1, X.J. Kamlet, Ed.Interscience, New, York, 1960.

2. S.D. Haman & M. Linton, J. Chem. Soc. Faraday Trans, 2243 (1974).

3. P.L. Marinas, J.E. Mapes, D.S. Downs, P.J. Ketone and A.C. Forsyth,Molec, Cryst, Liq Cryst, 35 15 (1976).

4. 0. Sandus and N. S1agg, Tech Repot ARLCD-TF.-78C25, May 1978.

5. K.H.S. Bagdasayan, Yu 1 Kiyukhin and Z.A. Sinitsina, J. Photochem,1 (1972/73).

6. T.H. Mourey and S. Siggia Analytical Chern. 51, 763 (1979).

7. T.J. S~ek, R.J. Osiewicz and R.J. Bath, J. Forensic Sciences,525 (1976).

8. P.A. Mullen and I.K. Orloff, J. Phy.,s. Chem. 77, 910 (1973).

It-up,

Arthur D Uttle, Inc

UV Absorption data for solutions of analytes were presented in Table 111-6.We also obtained solid state UV absorption coefficients of six analytes toP Fdetermine whether the tentative conclusions based on examination of solu-tion data could be confirmed. A technique used extensively in infraredanalysis--the preparation of KBr pellets containing the analyte--wasattempted and found to be satisfactory through the Cary 219, 4 00 -200nm,ultraviolet range. Macropellets of 13-mm diameter were pressed under20,000 pounds pressure, using oven-dried KBr to prevent moisture fromclouding the pellet.

Initial attempts to uee a Wig-L-Bug device to thoroughly mix sample andKBr proved unsatisfactory--inexplicable clouding occurred, even inthoroughly dried containers. A successful method finally used involvedgentle but thorough grinding in a mortar and pestle. Some experimentationwas necessary to determine the right concentration to yield a measurablepeak.

Two methods of measurement were used, both giving comparable results.The first method involved plotting an air vs. air baseline, thenobtaining both blank KBr pellet and sample KBr pellet traces. At thesample peak, the blank reading was subtracted. In the second method,a baseline was obtained with blank KBr pellets in each beam. When thesample pellet was substituted on the sample side, the absorbance at thepeak could be read directly.

Thickness of the pellets was measured with a micrometer and calculationsmade as noted in Table lII-6. These findings confirm the tentative con-clusions noted earlier.

4. Laboratory Demonstration.

The thermal imaging equipment required for laboratory demonstration of thepractical utility of the UV illumination/thermal imaging sampling protocolwas rented from Inframetrics, Inc. for one week. Using this equipment andUV illumination sources already on hand In Arthur D. Little, Inc. labora-tories, we were able to observe the presence of small quantities of solidanalyte on all of the available surface types of interest except brick.However, the observed contrast between analyte and surrounding surfaceswas less than would be desirable for routine field use. Independentmeasurements indicated that the power output of the UV illumination sourcesused was on the order of 150 microwatts to 1 milliwatt, which is lessthan that assumed in calculations above. Substitutions ofthe experimental values for UV power output in those calculationsresults in a predicted temperature rise on the order of that whichwas, in fact, observed. Thus, while these experiments confirm thepractical utility of the method, they also indicate that for optimumperformance a U1 illumination source having a higher power output inthe desired wavelength region should be used. We expect that a UVsource with the desired performance characteristics and operationalcharacteristics that would permit its safe use under field conditionscan be obtained commercially.

63

Arthur D L.tale inc

Un 0nV) 0 a0 0 0 0 -4 1-4

312 -'7 x U

LuM

-4 4)

0 -4 0

-4-

-4 C14

-%) 04 -T 00

*~~~~0 co D . :C-4 -4-P

U coo 0 0

'z G .C4 4 C4 C4C40

w L V

&0

C4)

0- 0

00

CC

C). UM

p.4 no7 eN7 1-4 -4

0

-r4AC 0' 0o C4 V) r uN -4-

-~ .04)Q

0 e- 0 Q

U,0 04

~0V 2 U01

00 P En

-4 04 -

64' Arhu D ittle

E. ANALYTE DETECTION USING UV IRRA.DIATION AND U`V PHOTOGRAPHY:

i. Background,

UV light directed at analytes present on a surface may be 1) transmitted,2) reflected, or 3) absorbed. Measurement of the temperature rise

resulting from UV absorption is the principle underlying the thermalimaging approach. Any LUV absorption will necessarily be accompanied by

a decrease in the UV reflectance, and observation of the latter quantity

may also be a feasible detection method. In this case, surfaces on which

analytes are present would appear as dark areas against a lighter back-

ground when illuminated with a UV source and viewed with a UV selectivedetector. The detector could consist simply of black and white film in a

large format camera equipped with a UV filter centered on or about 2500

Angstroms. The advantages of this approach would include the simplicity

and ease of operation of the required equipment and the fact that the

resulting photographs could be maintained as a permanent record ofthe investigations. .

2. Preliminary Experiments.

Two types of experiments were performed to assess the practical utility

of this approach. In one experiment the diffuse UV reflectance from

a concrete surface to which 60 ug/cm of 2,4-DNP were added was measuredin a U'V spectrophotometer. About a 59% reduction in reflectance at

3600 Angstroms was observed. This observation suggested that for

concrete and probably for other surface types as well, the anticipated

effects were, in fact, observed and that the measurements could

probably also be made at much lower analyte concentrations.

In a second set of experiments, UV photographs of several analytes

spiked on surfaces were obtained. Figure 111-7 is a photograph showing

the contrast obtained for 2,4-dinitrophenol, tetryl, 2,6-dinitro-

toluene, RDX and TNT in 1.3 x 10- g/cm2 amounts on a thin layer

chromatography silica gel substrate. The photograph was obtained usinga 2400 A filter and a Ziess HBO 200 W source with the glass lensremoved.

Figure 111-8 shows the concentration sensitivity for differing amounts

TNT, tetryl and 2,4-dinitrophenol on concrete using a 3650 A filter.

The top row has concentrations of approximately 6 x 10-5 g/cm2 for each

material (including the amount lost by diffusion into the substrate).

The second row has about twice the amount of TNT and half the amountsof tetryl and 2,4-dinitrophenol although the spot size is somewhatvariable.

Characteristics of the ULV irradiation and viewing filter would have to

be optimized for maximum contrast. Additional experiments using

different VV illumination sources and narrow bands pass UV filters were

performed to establish the optimum conditions for UIV photography,

65

"Arthur [) Litthe Inc

let"

Figure 111-7. LTV Photography of 130 iglcm 2 Each of2,4-DNT, Tetryl, 2,6-DNT, RDX, and TNTon TLC Plate

Figure 111-8. UV Photogcaphy of TNT, letryi, and2,-.''on Concrete

66 Axthur D. Lictie, 111C.

r

The camera and lens for these experiments consisted of long bellowsBausch and Lomb microscope camera that could be converted to a horizontalposition. The camera has a variable shutter to which we could attacha simple, 250 mm focal length, quartz lens. A variable Lris was usedto reduce the aperture and thus effectively increase the depth offield. This procedure compensated for the :hauge in =ffective fo-.qllength encountered when focussing with visible light then Photograph-

1P ing at a shorter wavelength, thus throwing the image out of focus.

This camera has a 4 x 5 sheet fil.m formar i4ith which we used PolaroidP/N 55 sheet film providIrng both the positive print and a negstive. Thefilm is a slow high resolution type film with apparent adequate set-sitivity to the ult-caviolet to allow a relatively short exposure time.Five seconds was adequate with the Zeiss lacp.

The experiment.- described previously uti1ized a high pres-sure mercury arc (Zeisr HBO ?00 '4art) which produces a substantialamount of energy centered cn .ae 365 nm mercury emission line. Thisradiation was .3ele.•tively transmtttted to the camera lens with the use ofa barrier filter O-th a maximum transmission at 365 nm.

In additional erperiments, a .iinralite Model S-61 sourc6 was

obtained for evaluationr. This is a low pressure merc..ury lamp which isreported 1', have a higner intensity cf 254 nm radiation. This lamp isof particular Interest as a radiation source for the thermwl imagingexperiwaents where it i% esAencial to optimize; thc absorption ofradiativn to achieve the maximum 6uZqequent thermal emission. Withthiz lamp .ks a lighc &rsurce. a series of photographs were taken of thetest spots of explosivey on various substc-ates using differentbarrier 'ilters. 9nperimentally it was derem.ned that a 10-minuteexposure was requirod to produce an adequate imagc usisng a 254 imvand pass barrier filter )ver the cnimer" lens.

The results of these experiments are shown in Table I-7.

These data i2ndicate that it takes a much longer expos.ire ir.terval toobtain a suitable image with tho shurter 254 run radiptlon. This wc'i.d.not be a significant handicap if there were a dilstiwut advantage Zo begSaned. However, we ohserve approximately the same level of gray imngefor each of the representative substrates for both rsdiadton sources.Itn orde, to successfully photograph any enplosive reiidue, it is obviousthat the image muat appetr darker than the ba-kground or edbmtrata material.In our nva1Ination we have applied the explosive r-ompound tc silica thinlayer chrcmatography substrates as a standard of comparison. A goodcontrast imege was obtained for most of the explosivis on this substrateand 50 microgram spots could be detected for most of the compoundo.Nitroglycerine and PFTN did not abs.,rb at this wavele'ugt'i and do notphocograyh well under these c-onditiona.

.lt-:avioleL Prcjucts, Inc., South Faaadena, CA

At tit I) jrdc Im

TABLE 111-7. RESULTS OF UV PHOTOGRAPHIC EXPERIEfNTS

High PressureHg Arc Low Pressure

- Radiation Source HR Arc, Zeiss Hg Arc, Hineralice

Most intense UV radiation 365 nM 254 riM

Optimum barrier filter for camera 365 rnM 254 nM

Explosure level (film response) 5 seconds 10 minutes

qbstrate Response:

Silica, Chromatograph plate White White

Concrate block Light gray Light gray

Transi te Medium gray Medium gray

Galvanized stell Reflective Reflective

Brick Dark gray Dark gray

Wood Dark gray Dark gray

Glas Black

68

Arthur D h.tte. Inc

Typical examples of these results are shown in attached photographs.These photographs were taken using the low pressure mercury source withan exposure of 10 minutes. Figure 111-9 was taken with the 365 nm trans-mission filter over the lens. We note excellent contrast for DNP, tetryland TNT. Figure III-10 shows the effect of using the 254 nm tansmissionfilter which changes the focal length with the shorter wavelength.It is interesting to note that the DNP and tetryl have less contrast atthis shorter wavelength, while RDX has slightly better contrast.

Figure III-10 includes a sample of transite containing 50 micrograms quantitiesof applied explosives. Recognizable spots are noted for DNP, tetryl, RDXand TNT while TNB is barely visible.

It has been reported that normal optical glass will transmit enough 365 nmradiation that suitable photographs can be produced with a conventionalcamera lens. However, this would be a less than optimum iondition andtotally ineffective for the 254 nm radiation and therefore not investi-gated.

Taken together, these results suggest that further investigation of thisapproach appears to be warranted. In particular, it would be desirableto evaluate under actual field conditions the ability of this approachto detect those analytes for which these experiments indicate it is bestsuited.

I

69

"Arthur D Little Iric

r

Figure 111-9. Analytes on Silica Gel and Metal. 365 m

rdo ff

Figure III-10. Analytes on Silica Gel (top) and Transite (bottom) 254 nrn

70 Aru -Titp

r

IV. QUANTITATIVE METHODS DEVELOPMENT

p, A. INTRODUCTIONThe quantitative method selected for developmental testing involvedsolvent extraction using alternative procedures to conventional wipeor swab methods. Existing USATHAMA methods for the determination ofexplosives, modified as necessary for this particular application, were

p_ used for analysis of the resulting extract.

The objective of this testing was development of procedures for quanti-tative determination of specific compounds down to concentrations aslow as 5 Pg/10 cm2 . The approach used to achieve this objective involvedin each case the spiking with known amounts of analytes of new, uncon-taminated samples of each of the surface types of interest obtained frombuilding materials dealers. The samples of conductive non-sparkingflooring that were available for this study were known to have been con-taminated with explosives/explosive residues, thereby precluding thespike and recovery approach. The results obtained by applying theprocedures developed for other surface types to samples of this material"are described in Section IV.C.3.

At the direction of the Technical Project Officer, emphasis was placedon the development of procedures for organic analytes. Methods based onan extraction approach similar to that described in this report fororganic compounds are used widely for the determination of inorganicspecies, including those of interest in this study, in soils, sediments,sludges, etc. No methods for inorganic species which would representsubstantive improvement over those methods were finally identified.

B. SEMIQUANTITATIVE CERTIFICATION TESTING

Preliminary semiquantitativn certification was accomplished by spkinganalytes into the quantity of acetonitrile specified in the existingUSATHAMA method prior to any concentration or solvent exchange steps,and continuing through the steps of the analytical method. The certifi-cation test involved analyzing five analyte concentrations plus a blank,one time each on three or more days. The resulting data is summarizedin Volume II of this report.

"Tables IV-l and IV-2 summarize the analytical method conditions andthe statistical data, respectively. (Tables IV-l and IV-2 arereproduced as Tables I-1 and 1-2 in Volume II of this report).

C. QUANTITATIVE CERTIFICATION TESTING

1. Introduction

Quantltative certification was accomplished as noted above by spikinganalytes in acetonitrile soluuion directly onto new, uncontaminatedsamples of each of the surface types of interest, extracting the spiked

71

Arthur D Little. Inc

C C c

a~ cc: C4 oe at ga a i ccU - c -r ! f -Q -, -W-,

C: v -0 U - -a c al In - 40r .r, ~ i

o. z o x 10 11 7- -.- 112

wi .0 C CO -1 -

C. = - - . - = C ~ L ( j

mm m4 bek Ot

C~~7 OC z

"I4 U I

- W 6#j 16 UR 146 lo w 6

Q 9 a ': =U a -(.4 N c4 C cU 0U NU r4- r4CI~C x x- C - - C -- C - - - -

I tj Ci C. I U

C, m 4 ' b N U a. 2. r-CUCU6 V) V) r) r)InJ CA - Cn -l V

.r ic- = C - h . -

L7UrhrDLjte na

3' -, .a .*A* 7 00 z 0 0 0

> E-

14a

Ce

'-4C

-i - - - - - - --

4. 4 4 6 2 6 73U U

* ~ ~ ~ ~ ~ Atu D Utt - . l .

samples with acetonitrile using ultrasonic agitation, and analyticalprocedures used for preliminary certification. The analytical methodsthat we developed were evaluated according tv the procedures specifiedin tbc 1980 USATHANA QA Plan. The resulting data is summarized inVolume II of this report.

2. Analytical Methods

The analytical procedures that were developed and evaluated aredescribed in detail on pages 76 through 87 of Volume I of this report.(These analytical methods are also presented on pages 46 through 57of Volume II of this report.

3. Results and Discussion

Table IV-3 summarizes the statistical data obtained from QuantitativeCertification Testing. (Table !V-3 is reproduced as Table 11-1 inVolume II.)

Documents

ARTHUR D. LITTLE, INC. CAMBRIDGE, MA 02140 · 2011. 5. 14. · ARTHUR D. LITTLE, INC. CAMBRIDGE, MA 02140 SEPTEMBER, 1982 DTIC "FINAL REPORT -CT•w Volume I CT29 g1E Distribution