TI. · 9E. LIST OF FIGURES ... Page Figure 1. The Evaluation of Irwin Clays in Adsorbing DECP Using the Headspace Method (Section II.A) 17 Figure 2. The Evaluation of Bentonite Clays

Decontamination of Casualties from BattlefieldCJ Under CW and BW Attack

FINAL REPORT,

Robert E. Lyle, Ph.D.Henry F. Hamil, Ph.D.

Edward P. McGovern, Ph.D.David A. Trujillo, Ph.D.

'November 15, 1984

U.S. ARMY MEDICAL RESEARCH AND DEVELOPMENT COMMANDFort Detrick, Frederick, Maryland 21701-5012

Contract No. DAMDi7-83-C-3128

TI. C , Southwest Research Institute

6220 Culebra RoadAUG17 1"3896 P.O. Drawer 28510

San Antonio-, TX 78284

'Approved for public release; distribution unlimited.

The findings in this report are not to be construed as an officialDepartment of the Army position unless so designated by otherauthorized documents.

0 711

SECURITYIO 5-T P_ýAGEForm Approved

REPORT DOCUMENTATION PAGE OajNo. O704-018U

Ia. REPORT SECURITY CLASSIFICATION lb RESTRICTIVE MARKING5Unclassified

2a. SECURITY CLASSIFICA IION AUTHORITY 3I ISTRIBUTIONiAVAILABILITY OF REPORTApproved for public release; distribution

2b. DECLASSIFICATION/DOWNGRADING SCKEDULE unlimited

4. PERFORMING ORGAN!ZATION REPORT NUMBER(S) 5 MONITORING ORGANIZATION REPORT NUMBER(S)

6.. NAME OF PERFORMING ORGANIZATION 6b OFFI:CS SYMBOL 7a NAME OF MONITORING ORGANIZATION

Southwest Research Institute

6c. ADDRESS (City, State, and ZIP Code) 7b AODRESS(City, State; and ZIPCodo)

6220 Culebra Road, P.O. Drawer 28510San Antonio, TX 78228-0510

1e. NAME OF FUNDING/SPONSORING Sb OFFiCE SYMROL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORCANIZATION U.S. Army Medical DAMD(7-83-0-3128

Res & Dev Command ,_D___ ___7-83-__ ____8

Sc. ADORESS (City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMPERSPROGRAM PROJECT TASK WORK UNIT

Fort Detrick, Frederick, MD 217C(-5012 ELEMENT NO. NO 3M16 NO ACCESSION NO62736A 2734A875 AP 436

11 TITLE (IMClUde Security Clew/*Ca Jn0)

Decontamination of Casualties from Battlefield Under CW and BW Attack

12, PERSONAL AUTHOR(S)

R.E. Lyle; E.P. McGovern; D.A. Trujillo; H.F. HamilI]&. TYPE OF REPORT lib TIME COVEREO 4 DATE OF REPORT (Year. WOnth. Day) 1S PAGE COUNT

Final FROM44l/8j3 rTO 10/1 /. 1984 November 1516. SUPPLEMENTARY NOTATION

11 COSATI CODES 18 SUIoECT TERMS (Cont•m• on revors of nfnctn'y and odennfy by Woc*k niwmteFIELD GROUP SUe-GROUP Modified Clays; Clay, DiethyLene triaminet DS-2 analog; CW

06 17 simulanta; Grafted polymers, Fuller's Earth; Kinetics of,

07 03 decomposition; Reactive polymer, Analytical methods; DS-2;19 ABSTRACT (Comtoiue on revere of moeuary and oyntIloy by bdo(k numbeJi)

A series of activated clays were evaluated for their ability to adsorb and/or reactwith simulants of mustard (HD). The simulant was chloroethylethylsulfide (CEFS). Using thenewly dev'Ioped "washer test" for total adsorption-decomposition, the clay sample was con-siderably more effective in stopping CEES than was Fuller'# Earth. A measured decompositionrate of 4 minutes for the simulant CEES showed the clays to be much better than ilcr'sEarth or samples of proprietary polymers. The results Wiri 11,, .ýmuianr diethvl chlorophos-phate (DECP) were not clear-cut, since rprrai- clay samples appeared to encaputulate theagent without des:roying 1. ,,ttew,,,ts to identify products of zhe decomposition were onlypartly successful, since the products seemed to be bound to clay. Hydrolysis products ofthe DFCP were detected.

Reactive monomers of vinyl benzyl chloride were grafted onto polvethvlene film% andconverted to quaternary salts. The anion of these salts was exchanged to give highlv nucleo-

20 01STRIlUTIONiAVAILAItILITY OF ASSTRACT .1 AISTRACT SECL:RITY CLASSF-CATION

CUAIC(LASSiF1(OIJNt4VITFD U~ SAME AS OFT Q3 I.)YC )Sf*S UI'CIpllif-iedIla NAME Of RESPI)N'SRIJ NO~V:OVAL 22 (IEP.ON( I(Nxtude Are* Coe) 22c U$V(f SYMOOL

Mrs. Virrinla M. Miller 301/6f)3-7325 SC:RP-RMI-S

DO Form 1473, JUN 84 •viooa tonre'oore SarITe 604.to

b

18. (continued)

Nucleophiles; Quaternary salts; Indicators; Decontamination of CW agents;Oxidizing agents

19. (continued)

philic hydroxide, azide, or thiosulfate salts. Oxidative anions such as permanganatechromate, perchlorate, chlorate, or hypochloride were also formed. These films hadlow capacity for decontamination but could be treated with indicators to formdetection films.

Polymers having the diethylene triamine side chain were prepared. Attempts toprepare a polymeric form of DS-2 (a sodium hydroxide-containing decontaminant) wereinterrupted by the close of the project. ,

AD

Decontamination of Casualties from BattlefieldUnder CW and BW Attack

FINAL REPORT

Robert E. Lyle, Ph.D.Henry F. Hamil, PhD.

Edward P. McGovern, Ph.D.David A. Trujillo, Ph.D.

November 15, 1984

U.S. ARMY MEDICAL RESEARCH AND DEVELOPMENT COMMANDFort Detrick, Frederick, Maryland 21701-5G12

Contract No. DAMDI7-83-C-3128

Southwest Research Institute6220 Culebra RoadP.O. Drawer 28510

San Antonio, TX 78284

Approved for public release; distribution unlimited.

The findings in this report are not to be construed as an officialDepartment of the Army position unless so designated by otherauthorized documents. A

NTiS CRA&I tDlIC TH 0

By-- . .. ,4,I.- -1;', •r,}" . .. .. . .

r

e

2

FOREWORD

Citations of commercial organizations and trade names inhis report do not constitute an official Department of thermy endorsement or approval of the products or services ofthese organizations.

3

4

-a

TAPLE O? CONTENTS

Page

FOREWORD 3

TABLE OF CONTENTS 5

LIST OF TABLES 8

LIST OF FIGURES 11

I. INTRODUCTION AND BACKGROUND 11

A. Reaction, Adsorbent Clays 11

B. Reactive Grafted Polymer Films 12

II. EXPERIMENTAL METHFDS 14A. Evaluation of Clays with Nerve and Mustard

Simulants Using Headspace Analysis 14

B. Evaluation of Clays by Extraction 14

C. Procedure for Comparing Reactivities ofClays with CEES 14

1. Method 1 14

2. Method 2 14

3. Kinetic Runs 23

D. Procedures for Comparing Reaction of DECPwith Clays 23

1. Method 1 23

2. Method 2 23

3. Method 3 23

4. Method 4 23

5. Kinetic Runs with DECP 26

5

Page

E. The Washer Test for Evaluation of Reactivity ofTreated Clays Using Chemical Agent DetectorPaper, VGH ABC-M8 26

1. Packing of Washers With Treated Clay 26

2. Addition of Simulant to Clay PelletContained by Washer 28

F. 4 Minute Kinetic Runs with CEES and DECP 28

G. Modified Washer Test 36

H. Procedures for Film Preparation 36

1. Grafting 36

2. Preparation of Periodide Films 38

3. Alternative Preparation of Periodide Films 38

4. Preparation of Azide Films 38

5. Preparation of Hydroxide Films 39

6. Preparation of Thiosulfate Film 39

7. Preparation of Chlorochromate Films 39

8. Preparation of Perbromide Film 40

9. Quaternization of Grafted Films 40

10. Ion Exchange Capacity 43

I. Nerve Agent Simulant Diethyl-p-Nitrophenylphosphate 43

1. Preparation of Standards 43

2. Visible and Ultraviolet AbsorptionSpectra 43

J. Preparation of Indicator Treated FilmBD-5-Q-NMe 3 +OH- 44

6

Page

K. Reactions of Simulants wit! Grafted Films 44

1. Diethyl Chlorophosphate 44

2. Mustard Simulant-Chlor.DethylEthyl Sulfide 44

L. VX Simulant O-Ethyl-S-ethylmethylphosphonothionate 49

III. RESULTS AND DISCUSSION 50

A. Introduction 50

B. Choice of Simulants 50

C. Clays 50

1. Analytical Methods 51

a) Headspace analysis 51b) Kinetic analysis - 6 minute test 52c) Kinetic runs - Variable time 53d) "Washer test" - Estimation of

adsorption and decomposition 54e) Kinetic test - 4 minute test 55

(1) CEES 55(2) DECP 56

f) Modified washer test 58

2. Attempts to Determine the ReactionProducts 65

D. Grafted Films 70

1. Preparation 70

2. Evaluation 70

3. Detector Films 74

E. Polymeric DS-2 Analogs 80

1. Films with Grafted DS-2 80

2. Polymers with DETA Side Chains 80

IV. FUTURE WORK 82

V. REFERENCES 83

DISTRIBUTION LIST 84

7

a a

LIST OF TABLES

Page

Table I. Simulants for Chemical Agents 15

Table II. Control Clays and Modificatlons Receivedfrom Anglo-American Clays Corporation 16

Table III. Regression Analyses on KineticRuns of CEES 24

Table IV. Comparison of Methods in Section II.C 25

Table V. Regression Analyses on KineticRuns cf DECP 27

Table VI. Results of Spiking Clay Pelletswith CEES 31

Table VII. Results of Spiking Clay Pelletswith DECP 32

Table VIII. Results of Washer Tests on Clays4952-4953 33

Table IX. Decomposition of CEES by Clays

After 4 Minutes 35

Table X. Tightly Packed Washer Test with CEES 37

Table XI. Simulants Tested 41

Table XII. Absorbance at 273 NM of 9.CI x 10- 5MDNPP After Incubation with GraftedDerivatized Films 45

Table XXIII. Color Reactions of Indicator Filmswith Sin.ulants 46

Table XIV. Reactions of Rohm and Haas Resinswith CEES 57

Table XV. Decomposition of DECP by Clays 59

Table XVI. Reaction of Rohm and Haas Resinswith DECP 60

8

Page

Table XVII. Results of Washer Tests on ClaysReceived 8/84 Using CEES as Simulant 62

Table XVIII. The Results of the Washer Test withModified Clays 63

Table XIX. Evaluation of Robm and Haas Resinswith Tightly Packed Washers Using CEES 64

Table XX. Reaction of Grafted Films with CEES 75

Table XXI. Reaction of Grafted Films with EEMPT 76

Table XXII. Reaction of Grafted Films with DECP 77

Table XXIII. Color Reactions of Indicator Filmswith Simulants 78

9E

LIST OF FIGURES

Page

Figure 1. The Evaluation of Irwin Clays in AdsorbingDECP Using the Headspace Method(Section II.A) 17

Figure 2. The Evaluation of Bentonite Clays inAdsorbing DECP Using the Headspace Method(Section II.A) 18

Figure 3. Evaluation of Clays for Adsorption ofCEES Using the Headspace Method(Section II.A) 21

Figure 4. Evaluation uZ Clays 1, 1 and 3 AgainstDECP Using the Headspaci AnalysisMethod (Section II.A) 20

Figure 5. Evaluation of Clays 4 and 5 Against DECPUsirg the Headspace Analysis Method I(Section, II.A) 21

Figure 6. The Analysis of the Extract from theClays Treated with DECP in the 6Minute Test (Section I.1.3) 22

Figure 7. The Effect of Copper Ion on theDecomposition of DECP 22

Figure 8. The Results of the 4 MinuteDecomposition Method with Modified Clays 29

Figure 9. The Results of the 4 MinuteDecomposition Method with Modified Clays 31

Figure 10. The Fourier Transform Infrared Spectrumof the Breakthrough Product of DECP withClay RT008 66

Figure 11. The Fourier Transform Infrared Spectrumof DECP 67

Figure 12. 1 HNMR of the DMSO Extract of CEESAfter Treatment with Clay RT004 68

Figure 13. The 1 HNMR of CEES 69

10

I. INTRODUCTION AND BACKGROUND

The treatment of wound casualties on a battlefield in -. ichchemical and/or biological agents are being used is complicatedfor both the casualty and the medical team that is supplyingaid. In order to provide the optimum conditions fc'L treatmentof wounds in this situation, it is essential thit a method bedevised to protect and decontaminate and cont,,in any chemicalagent remaining on the casulty or his clothing. In most instancesthe casualty will be clothed in an ensemble for protection ag.instagent exposure, but this protective barrier will be compromisedby the munition which caiz-d the wound. Thus it is desirablethat the treatment also seal the ensemble to prevent furtherexposure of the casualty to chemical agents.

Detuermining the method of treatment is further complicatedbecause the decontaminating agent must be sufficiently mildthat it will not cause significant 'damage to the expose.d skin.In general, this would eliminate strong caustic solutions, suchas DS-2, or the highly potent oxidizing agents, such as (supertropical bleach (STB). It is presumed, therefore, that a moresatisfactory approach to reactive decontamination would be theuse of the milder oxidizing agents, such, as soft halogens, orweakcr nucleophilic reagents, such as amines.

The approach proposed by SwRI to meet these criteria istwofold. Consideration of the procedure of the Canadian andBricish Air Forces I to de:cont,,minate with Fuller's Earth sug-gested -the possibility of utilizing a packet or bandaqe, of porouscloth containing a reactive -idsorbent, such as a clay, witha reactive surface coating. This packet would allow the chemicalagent to enter by diffusion for adsorption and r,,action withth- clay. The second approach utilizoq a reactive film, whichcould be' used as a covering or perhips as a thread in a wovenba'ndage. Either of those could be placed over the wound toprovide protection against furthor entry of chemical aq,,nts,to provide decomposition of any agvnt in the vicinity of theentry point, and dotoxifi(,.tion of iny chemical or bioloical,1gent on the ,ensemble for the protection of the medi cal t,,eamtrr,ating thi! casualty.

A. .!acr ive, Ads;orbhnt

A (,,;rrentf y aval Ible act iv.• •d clay has been under inv--st-i.;i" ,_n .. ,i si1rfi(-, for other r .iact iv, functions. Anglo-Am,'r ,can

C j,/ys ef a;,indfrsvi l Io, G;orii•r , hi'i ured a prop-lotiry pr-coi,,I ro, Iruuc, h i gh l y ,1duorb *nt or re- a ct i v e s3irfa c o oi a , t - i t (ed

,rLen'.n t,, clc.y. Prior to this contract several s, mol,-s of th,-;,r",l1Y/ ha.d hoee,l 4 -xIim I ,n, I f(ýr th'i r e, f f,*ct i ve n', s in r,',. 1,A ci nti-ý, vapor pressurr of ',,nzyl br,:r-ido, which was us,,' as a s41',ulaint

for mustard (HD). 2 The effectiveness of the treated clays forreducing the vapor pressure of benzyl bromide was approximate-ly one order of magnitude greater than that of a comparableweight of Fuller's Earth.

The proprietary method used by Anglo-American Clays forintroducing reactive, surfaces on the clay also could be usedto introduce a nucleophilic amine, such as ethylene diamineor diethylene triamine (DETA), on the surface of this clay.They have methods for increasing the surface area of the claycreating large pore sizes, and allowing intercalation of catalyticmetal ions in the layers of the clay. Thus the a.Lion of thetreated clay would be twofold: (1) The adsorption would reducethe vapor pressure of the chemical agent surrounding the wound,and (2) the reactive surface would irreversibly deactivate thechemical or biological agent near the wound by a slower mechanism.The proprietary methods of clay treatment used by Anglo-AmericanClays were not disclosed to SwRI. The properties desired forthe clays were determined by conference between SwRI and Anglo-American Clays and then were furnished by the latter on issueof a purchase order for the sample.

B. Reactive Grafted Polymer Films

The Polymer Chemistry Section of the Chemistry and ChemicalEngineering Division of Southwest Research Institute has shownthe effectLvvne&i of grafting polymierb by uLilizing an ionizingradiation soruce to induce the polymerization. By this technique,reactive surfaces have been introduced on a variety of inertfilms such as polyethylene, Teflon, and similar paraffins.' 3

Utilizing this technique, polyvinyl pyridines have beengrafted to polyethylene. The pyridine nucleus of these filmswas then treated with an alkylating agent such as methyl iodideand the halide ion converted to a variety of reactive polyatomicanions such as 137, Brf-, and N3-. The perhalide forms of graftedfilms are the subject of a patent disclosure and must be con-sidered proprietary.

The films have several interesting properties. (i) As,one might expect, the trihalide foim of the quaternary salthas' a highly intense color, and as the film undergoes reactionwith reduction of the trihalide ion, the color fades. Thusthe film acts as an indicator showing when the reactivity ofthe film has, been expended. (2) A family of perhalides andmixed halogen salts could be prepared from these grafted vinylpyridine polymer films. Thus a range of oxidizing activitiescould be obtained to baWance the reactivity required for thedeactivation of chemical and biological agents with the mildnessrequired for compatibility with human skin. (3) The preliminary

12

studies show that the film can be regenerated to the perhalideform by washing the film with a halogen solution in chloroformor the trihalide solution in water. Because their propertiesseemed suitable, the films were studied for their use in decontam-ination or detection systems.

The objectives of this project were:

o Design and procure the decontamination materials

- Select simulants for the chemical agents

* Develop analytical methods for evaluating the decon-tamination materials

* Introduce a color detector system into the decontami-nation material

13

II. EXPERIMENTAL METHODS

The simulants used are given in Table I and were obtainedfrom Aldrich Chemical Co. The clays are given in Table II andwere obtained from Anglo-American Clays Corporation.

A. Evaluation of Clay with Nerve and Mustard Simulants UsingHeadspace Analysis

A series of e00 mg samples of clay, as recei ed from Anglo-American Clays or modified by methanolic Na.'" .reatment, wereplaced in 4 mL septum-sealed vials. Using a Hamilton syringe,10, 20, 40 or 60 pL of the simulant was injected directly intothe clay and the vial was immediately placed in a vortex mixerfor 1 minute. A 20 pL sample of vapor from the headspace abovethe clay was withdrawn with a 50 pL gas-tight syringe and analyzedby gas chromatography GC). The vapor height for the simulantwas compared to that for a sample from pure simulant with noclay added, about 200 . The results for clays A'-E' are givenin Figures 1, 2 and 3. The data for clays 1-5 were plottedas a percentage of the peak height of the standard with no clayadded. These results ar given in Figures 4 and 5.

B. Evaluation of-Clays by Extraction

As in the headspace method, 40 pL of simulant was addedto 200 mg, of clay, wiich was stirred on a vortex mixer for1 minute and then all wed to stand for 5 additional minutes.At this time 2 mL of ther was added; a 1:100 dilution of theether solution was anal zed by gas chromatography using an alkaliflame ionization detec or (GC-AFID). A solution of untreatedsimulant (no clay) was the standard. The results for clays1-6, as percentage of re:overed simulant, are presented in Figures6 and 7.

C. Procedure for Compa ing Reactivities of Clays with CEES

1. Method 1

A 200 mg samT le of clay was weighed into a vial and40 pL of chloroethyl el hyl sulfide CEES was added. The samplewas mixed on a vortex mixer for 1 minute. After standing for5 minutes, the sample was extracted with 2 mL of. diethyl etherand the extract was anal zed by GC-AFID.

2. Method 2

The procedure of Method 1 was followed, except thatonly 20 pL of chloroethyl ethyl sulfide (CEES) was added andthe mixture was allowed -o stand 15 minutes before extraction.

14

U) afl V) r > C*H 1 -4 0) a~) m 0

M-4-

~-4 V0 0 4 0

4ji

-44 u

1 0

.4-iC 00N

(u mN 0

ad c

Inj 4U U -U

0 >qI -W

-J /

TABLE II

CONTROL CLAYS AND MODIFICATIONS RECEIVEDFROM ANGLO-AMERICAN CLAYS CORPORATION

IdentifyingSymbol Treatment

A' Control clay IrwinB' Control clay sodium bentoniteC' 1% Diethylene triamine (DETA) on A''D' 1% DETA on B'El 10% DETA on B'1 Surfaced bonded ammonia2 or 4374 Intercalated ethylene diamine (EDA)4375T Surface bonded EDA3 Surface bonded urea4 DETA different from C', D', or E'5 Variation of 46 Control :-lay for samples 1-5RT001 Acid activatedRT002 Variation of RT001RT003 Seneficiated clayRT004 Beneficiated clay, calcined at 2500, 60 ma/g

surface areaRT005 Al+3 /Cr+ 3 intercalatedRT006 Cr+ 3 intercalatedRT007 High surface area, 400 m'/g, pore size of

12-14 ART008 Cu+ 2 intercalatedRT009 Cr+ 3 intercalated, 180-200 m'/g surface area4952 Acid wash, dried4953 Surface acidified4954 A1+ 3 intercalated4955 Acidified ethylene diamine treated5030 Zironyl ion intercalated

16

ý4 QJ.4 0

10~ 0

o.1 -4 wJ~

0 ~00

(-m IOJa 194 E-1

U.-. .~ 0

00 -4

V-44

0

~~.'- -41

U

~~ 00

a 4j

44U

0 r

.0

E40

181

-4

r'4

- 4

S0 • 0

0

00 .

4-4 ".6

19 ",4

r4

C-)

0 0 P-

o*4K

Q 4J

U)

u-

0

-'m

0

0 Ou'i

(7 00 Pý 1- 0 Ln T M C 4 ý4Cz

41

20

~~0'Un

-40

O@4 4-)

CD

u

CD.

U)>4-

ro)

-4 u

4.)

ra41

0' OA0 , 1o 4-

*CIS seaorspra UTd~g

Ln

-40

I4Jr21

Decomposition of DECP by Clays70

60 -

U 50 -

= 40 -

..30> 200

10

1 2 3 4 5 6Clay

Figure 6. The Analysis of the Extract from the ClaysTreated with DECP in the 6 Minute Test(Section II.B)

Decomposition of DECP by Copper Treated Clays70

S60W 50 -

> 40 -

> 30

o 20

dO10-

Control 2 + Cu Ccitr + CuClay

Figure 7. The Effect of Copper Ion on the Decompositionof DECP

22

3. Kinetic Runs

Several 200 mg samples of clay were weighed into separ-ate vials and spiked with 5 uL of neat CEES. The mixture wastreated as in Method I; the 'time from CEES addition to additionof 2 mL of ether was measured. A 3 pL sample of the extractwas analyzed by GC-AFIU. A regression analysis4 of the naturalln/ln plot was used to obtain the data given in Table III.

The difficulty with Methods 1 and 2 was the possibilitythat the capacity of the clay had been exceeded. Method 1 re-sulted in peak heights ranging from 75-93% of those standards,which were prepared by adding the standard amount of CEES toan equal volume of extracting solvent. When the amount of simu-lant added to the clay was decreased by 50% (i.e., Method 2)and the standing time increased, the resulting peak heightsdecreased to 30-72% of those of standards. Results of the compar-ison between methods aru presented in Table IV.

D. Procedures for Comparing the Reaction of DECP with Clays

1. Method 1

A sample of 200 mg of clay was weighed into a vialand 40 .L of DECP was added. The mixture was stirred on a vortexmixer and allowed to stand for 5 min.utes. The clay was extractedwith 2 mL of ether and the extract was diluted 1:100 beforeanalysis by GC-AFID.

2. Method 2

A sample of clay was treated with 2 mL of an etherealsolution of 1 pg/mL DECP. A 3 pL sample of the solution wasanalyzed at different times to allow measurement of reactionkinetics.

3. Method 3

A sample of clay was treated with 5 pL of DECP asin Method 1, except that the ether extracted was diluted 1:10.

4. Method 4

A sample of clay was treated with 10 pL of DECP asin Method 3.

23

(D0

)~4 . q ~ ~ qr qr q. qr r qv qr qr q

(N ( N -4 (N (N 0 0 0 0 - 0 (

. .. . . . . . . .

9-b.4 04 (7% 0% 00 0 0 0w 0% 0% 4- (N%

-0 0% E0 %00 -i ,

di w'Menf40M ^

C)4 0%0 t 0 u, -l0 0 0 %0 CD '0m- co en 0% 0% %% 0 ' 0 0 00 ol w% -P co N%

N 4)

~m.~ ~ 0 N, r- r f- 1-1 "1 N U,~ , (Fn '0 0I' % 4 , 0 ~ r

04 4f 0 - 0 0 0

424

TABLE IV

COMPARISON OF METHODS IN SECTION II.C

Method and duration of simulant/clay contactfor CEES. Values expressed as % simulantremaining.

Method I Method 2 KineticClay (5 min) (15 min) (5 min)

Control 78 30 7

4375 79 44 14

4374Tw/Cu 2 + 83 48 19

S.)

5. Kinetic Runs with DECP

Samples of 200 mg of clay were weighed into separatevials and 5 pL of a 20% solution of DECP in ether was addedto each. After 1 minute of mixing using a vortex mixer, thevials were allowed to stand t* minutes and 2 mL of ethyl etherwas added. The extract was mixed 30 seconds using the vortexmixer, and the clay allowed to separate. A 3.0 uL portion ofthe extract was analyzed 'by GC-AFID. The results are givenin Table V.

The difficulty encountered with Method 1 was the possibilityof exceeding the capacity of the clay by adding excessive simu-lant. In this case, differences in reaction rates would notbe observable in a suitable time period. Method 2 wa's abandoned,as we felt the use of a dilute solution did not closely relateto a "real world" situation. 'The method ultimately chosen re-quired addition of only 5 4L of a- 20% ethereal simulant solutionto the clay. The clays reacted rapidly with simulant addedin this mode. (Example: 41% of standard in 5 minutes for 4375T/Cu++). It was felt that the ether added (4 PL) would not inter-fere with or affect the reaction but woulJ allow improved, pre-cisiort as compared to a 1 IL volume of neat material..

Another consideration was the physical effect of addingrelatively large volumes to the clay samples. With both CEESand DECP, larger spikes (20 and 40 wL) led to clumping of theclay which could not be alleviated by agitation, using the vortexmixer.' Clumping could affect precision if it prevented reproduc-ible interaction between simulant and clays.

E. The Washer Test for Evaluation of Renctivity of TreatedClays Using Chemical Agent Detector Paper, VGH ABC-M8B

Steel washers having an inside diameter of 10 mm and thick-ness of 1.5 mm were used for the tests. The experimental proce-dure is outlined below.

1. Packing of Washers with Treated Clay

a) VGH ABC-M. paper* was placed on a glass platewhich, for easy bottom side viewing, was positioned a few inchesabove a mirror.

*The VGE{ ABC-M8 paper was obtained from the Contract Officer'sTechnical Representative, Dr. Millard Mershon.

26

0

4)0

00 c4 co ) w. %0 ýo 40 a

-4

,..

x (7% co -4 ON La in co ~

o, o 0 0 0 0 '-

?0 0 0 0 0 0 0

-.• I I I - I - I I Iz

U77

0-4

co as 0 r- ai 0% 0% ON ON 0

%. 0 0 00 0o 0 0 0 0A

C1C2

0) '0 -4 - -4 0 ~ '- q 0 Lf

-4 - N ~ 40 '.4-27

b) Three washers were secured to each piece of paperwith glue.

c) Enough finely ground clay was added with a micro-spatula to the center of the washer to overfill it.

d) Pressure was applied to the clay with a microscopeslide rotated in the lateral plane, thus packing the clay intoa disc. More clay was added to the center of the pellet andcompressed with the slide until the central hole was completelypacked with clay. This packing method resulted reproducibly,in clay loads of about 200 mg, determined by weighing replicateclay samples.

2. Addition of Simulant to Clay Pellet Contained by Washer

a) The disc of clay was spiked with simulant measuredusing a microliter syringe. The syringe was held verticallywith the tip of the needle less than 1 cm from the surface ofthe clay.

b) A stopwatch was started immediately upon spiking.

c) After 4 minutes, if no sign of simulant was seenon the bottom surface of the clay, a second spike was applied.rhis process was repeated at 4 minute intervals until interactionof the simulant with the detector paper was observed.

d) The time when the simulant was first detectedwas recorded.

Tables VI, VII and VIII present the results obtainedwith the various treated clays. The greater the. reported time,the more effective the clay in reacting with or adsorbing thesimulant.

F. 4 Minute Kinetic Runs with CEES and DECP

Samples of 200 mg of treated clays were weighed into separatevials and 5-100 pL of neat CEES or DECP was added to each.After 1 minute of mixing using a vortex mixer., the vials wereallowed to stand 3 additional minutes; 2 mL of ethyl ether wasadded and the powder extracted well by vigorous stirring for30 seconds with the vortex mixer. The clay was allowed to separ-ate and was centrifuged. A 3.0 UL portion of the extract Wasanalyzed by GC-AFID. A solution of known concentration wasused as a standard. The results, expressed as a percentageof simulant remaining, are given in Table IX and Figures 8 and9.

28

90

40 80

Z 70

S60

= 50

40

o OFuller's earth

0 30 0 Clay 5030To 0 6 Clay 4952T0Clay 4955TClay 4953T

r 20 Clay 4954T200 Clay RT04

U

SI 0I

10 20 30 40 50 60uL of CEES Added to 200 mg of Clay

Figure 8. The Results of the 4 Minute Decomposition Methodwith Modified Clays.

29

90

80

7u

60

S50

40

S03o Fuller's earth0 0 RT001 -0 RT

20Sa RT005

A• RT0060 RT007

SRT00810 0 RT009

10 20 30. 40 50 60

uL of CEES Added to 200 mg of Clay

Figure 9. The Results of the 4 Minute Decomposition Methodwith Modified Clays.

34

TABLE VI

RESULTS OF SPIKING* CLAY PELLETSWITH CEES

Time at Which Colored Spot FirstADpeared (min:sec)Replicate Replicate Replicate

Clay #1 #2 # 3

#1 9:50 9:45#2 5:15 8:00#3 6:40 5:30#4 8:40 5:30#5 12:30 3:45 12:15#6 10:45 9:00Crude Sodium Bentonite 11:15 4:45 8.30Sodium Bentonite

w/l% Amine 13:20 9:30Sodium Bentonite

w/1l0% Amine 3:00 6:45Crude Irwin 4:50 5:45Irwin w/l% Coupled Amine 7:40 7:00Fuller's Earth 8:40 12:304374T 5:30 3:504375T 9:45 6:15

#6 w/Cu 2 + 17:30 19:354374Tw/Cu 2 + 10:15 10:154375Tw/Cu 2 + 11:30 9:15

*Clays were spiked with 10 wL of neat simulant at time zero,then repeat spikes were made at 4 minute intervals (i.e., 0,4, 8, 12 .... ).

31.

1-4.

.4-)*.M L()0 0 00U

0. ena N -0

04 4.)4)

_4 m CmU,00 00 0L L0om 0 tr

00

.-4 a -

0 -N 0 0nL n0 0 o f .D40 00mtn D o0, .0ý ýq :* m4(N a 14 0v

0 ad or=~ez N NC'N ClqN "C14 1-4 -, C-4 - C)Nr4 4

.9 S.

~4) 4.)o

oI - MQ 4-JO '

0q L()U L 0 a t M a a0000 O") i 00 C 0-4

(1) OsNNr- 0 N0rN(0 0 - 10

U) a40 *

~~tn(4.

.4.).- 4..4~ ~~.-$ 0 0 0~.- ( ~-4

00-4

0)~~~U 4J 4) C 1

-4 0404.

* 0 41IA4

0 w0 wr

-4 C 4CF -4 WN40 c U N) .' 4.)

E Vp5 4 u ar

10 - 0 ', ,', -4 - n 3 W C)I E0 9 I z 3:1 - - r *s -40

r- N I ýow0 0 0w .w y , )mu -T

323

TABLE VIII

RESULTS OF WASHER TESTS ON CLAYS 4952-4955

Washer

CEESa

Interactionl Replicate Time (min)

4952 1 27.002 32.303 20:15

4953 1 26:002 23:303 31:00

4954 1 9:002 9:303 9:00

4955 1 25:002 25:303 21:00

Fuller's Earth 1 14:002 12:00

4952T 1 34:302 28:00

DECPb

4952 1 55:452 48:003 51:00

4953 1 52:002 54:303 42:30

33

. 4

TABLE VIII. (cont.)

Interaction

Clay Replicate Time (min)

DECPb

4954 1 17:0012 46:003 70:002

4955 1 50:302 36:303 43:45

1 Appears to have formed a channel down the side of the pellet.2 At around 55:00 minutes simulant seemed to build on uppersurface and drain off to side.

aFuller's Earth showed interaction at about 12 minutes.

bFuller's Earth showed interaction at about 30 minutes.

34

-r r -ER **.r %

c o r- kD 1,0L

441r- n

:4

N) ff m LO 1

~4 -

Q )

u

ol -4 0 rr- dI0

*.D CD C') tgD C: ) CD Coo C) 0 C D ) C

H-4 &a&)4 E- - - 64 E

1435

Determinations with RT004 were repeated at least four timeswith deviation of less than 10% from the values reported inFigures 8 and 9, so long as the ambient temperature of the deter-mination was reasonably constant. Subsequent to the preparationof the Draft Final Report the ambient temperature was changedand the temperature dependence of this test was dissolved.

G. Modified Washer Test

The washer test described in E was modified by using a5 mL round-bottom flask to indent the clay plug uniformly.Attempts to develop a uniform pressure to pack the clay plugshad limited success. The final modification was to pack theplugs until excess clay powder poured over the sides of thewasher. The flask was used to indent the surface and the weightof clay was recorded. The simulant was then added (10 pL ofCEES or 20 pL of DECP) every 4 minutes. The results, expressedas ,a factor, PL simulant X time mg of powder, are given in TableX.

H. Procedures for Film Preparation

1. Grafting

The films are graft copolymers prepared by the gammaradiation-initiated polymerization of the desired monomer ontolow density polyethylene film. The polyethylene (1.0 mil thick-ness, 14 in x 25 ft) was interleaved with cheesecloth and rolledaround a 0.5 inch aluminum rod. The rolled film was placedinto a hydrometer jar, which was closed with a rubber stoppercontaining inlet and outlet tubes. The jar was connected toa manifold and alternately evacuated to 0.5 cm Hg and broughtto ambient pressure with N2 (five cycles) in order to removeoxygen. The jar was evacuated again and the grafting solution(monomer plus solvent), was drawn into the jar to immerse thefilm roll. The filled jar was placed in the irradiation celland exposed to cobalt-60 gamma radiation at 12,000 rads/hourfor 68 hours. Total applied radiation dose was 0.811 Mrad.

After standing overnight, the jar was opened and thegrafted film was separated from the cheesecloth backing, washedthree times with a suitable solvent to remove any homopolymer,and air-dried.

Grafting solution compositions for the films reportedin this period are:

BD-l 20 wt % 2-vinylpyridine80 wt % methanol

BD-? 25 wt % 4-vinylpyridine75 wt % methanol

36

TABLE X

TIGHTLY PACKED WASHER TEST WITH CEES

Breakthrough TotalTime Simulant Simulant PL

Material Wt min:sec VL X sec/mg

Fuller's Earth 190 4:30 20 28186 4:27 20 28

RT004 175 9:00 30 93159 9:35 30 i08

RT006 84 5:00 20 7281 1.:35 10 1294 1:47 10 11

RT007 95 0:13, 10 1108 1:30 10 8

RT008 104 15:10 40 35296 10:30 30 198

RT009 82 6:05 20 9088 5:35 20 76

37

BD-3 25 wt % 4-vinylpyridine1 wt % divinylbenzene

74 wr % methanol

BD-4 25 wt % vinylbenzylchloride75 wt % methanol

BD-5 25 wt % vinylbenzylchloride

75 wt % bE:nzene

2. Preparation of Periodide Films

Samples of the three quaternized films (3 ft', ca12 g) were placed in separate IL wide-mouth jars. To each wasadded 330 mL of KI-I2 solution (0.2 N in K13) along with suffic-ient deionized water to fill the jars to within I inch of thetop. The capped jars were agitated on a reciprocating platformshaker for 48 hours. The films were transferred to 3L stainlesssteel beakers and washed five times with 2L portions of deionizedwater and dried overnight between absorbent paper toweling.Film BD-3-Q-1 3 disintegrated upon washing and Was discarded.The remaining two lustrous, red-black films were stored in amberbottles.

Film designations: BD-I-Q-1 3BD-2-Q-I3

3. Alternative Preparation of Periodide Films

Samples (6 in x 6 in} of each of the three films BD-l-Q,BD-2-Q and BD-3-Q were placed in separate 50 mL Erlenmeyer flasks.To each flask was added 350 mL of chloroform containing 3 gof T. The stoppered flasks were shaken overnight. Each filmwas washed four times with 300 mL portions of chloroform. Thecross-linked film (BD-3-Q) disintegrated upon washing. Thetwo remaining films were dried between absorbent paper toweling.The two remaining films (lustrous, red-black, flexible) werestored in amber bottles.

Film designations: BD-l-Q-1 3 (CHCl 3 /I 2 )BD-2-Q-I 3 (CHCl 3 /I 2 )

4., Preparation of Azide Films

Samples of films BD-l-Q, BD-2-Q and BD-5-Q (1 ft')were placed in separate IL jars. To each jar was added 200mL of 0.2 M sodium azide and 500 mL of deionized water. Aftershaking for 2 days on a reciprocating platform shaker, the filmswere removed from the jars, washed four times with 2L portions

38

of deionized water, and dried overnight between layers of ab-sorbent toweling. Film BD-l-Q (colorless, transparent), FilmBD-2-Q (pale blue, translucent), and Film BD-5-Q were storedin amber bottles.

Film designations: BD-l-Q-N 3BD-2-Q-N 3BD-5-Q-N 3

5. Preparation of Hydroxide Films

Samples of BD-I-Q (0.6 ft'), BD-3-Q, and BD-5-Q (1.0ft') were placed in separate 1L wide-mouth jars. To each jarwas added. 500 mL of deionized water containing 20 g of potassiumhydroxide. The capped jars were agitated overnight (ca. 18hours) on a reciprocating platform shaker. The films were removedfrom the jars and washed with six 2L portions of deionized wateruntil the washings were neutral to pH paper. The colored films(BD-l-Q amber; BD-3-Q blue-green) were air-dried between layersof absorbent paper and stored in amber bottles. BD-5-Q-OH wascolorless.

Film designations: BD-l-Q-OHBD-3-Q-OHBD-5-Q-OH

6. Preparation of Thiosulfate Film

Approximately 2 ft' (ra. 8 g) of quaternized BD-2-films (BD-2-Q and BD-5-Q) were placed in a 1L wide-mouth jaralong with 800 mL of deionized water containing 40 g of sodiumthiosulfate. The capped jar was agitated for 24 hours on areciprocating platform shaker. The film was removed, washedfive times with 2L portions of deionized water, and air-driedbetween layets of absorbent paper. The faint yellow film wasstored in an amber bottle.

Film designation: BD-2-Q-S 2 0 3BD-5-Q-S 2 03

The proco'dur-s in the prceding paragraph were rep,.atodto yield the chlorate, porchlorate, permanganate, dichromate,hypochlorite, m-chioroperbenzoaite and related polymers of BD-5-Qfilm.

'7. Preparation of Chlorochromate Films

Chromium trioxide (CrO 3 , 9.8 g) was added with stirrinqto hylirochloric icid (1.65 M, 60 mL). The resulting clear orinq,sol-ition wis divid,,] into two 30 mL portions and each portion

wa: added to a 11L wide-mouth jar containing 1 ft' (ca. 4 g)of either BD-l r BD-2 film in 450 mL of cold (0-5*C) deionizedwater. After s anding in an ice bath for 30 minutes, the cappedjars were agita ed on a reciprocating platform shaker for 48hours. The fil s were removed from the jars, washed five times(2L/wash) with deionized water, air-dried between layers ofabsorbent paper, and stored in amber bottles.

Film designation : BD-1-HtrO3C1

BD-2-HCrO3 Cl

8. Prepar tion of Perbromide Film

Sample of the films BD-l, BD-2, and BD-3 (3 ft' each,ca. 12 g) were placed in separate 1L wide-mouth jars. To eachjar was added 2 0 mL of the reagent (120 mL of conc. HBr + 40mL of Br 2/L and 500 mL deionized water). The capped jars wereplaced on a reciprocating platform shaker and agitated overnight.The excess reagent was decanted and the films were transferredto 3L stainless steel beakers and washed five times with 2Lportions of deionized water. The washed films were placed betweenlayers of 'abso bent paper toweling and air-dried overnight.After drying, the cross-linked film prepared from BD-3 was ex-tremely brittle and disintegrated upon handling. The othertwo films (reddiOh-brown) were stored in amber bottles.Film designations: BD-I-HBr 3

BD-2-HBr 3

9. Quaterrlization of Grafted Films

Approximately 3 ft' of each film (ca. 12 g) were placedin UL Erlenmeye fla4s. To each flask was added 750 ml ofreagent grade mthanol containing 30 mL of methyl iodide. Thestoppered flasksl were agitated for 3 days on a reciprocatingplatform shaker. The films were removed, washed twice withmethanol, and ai -dried.

Film designations BD-l-QBD-2-QBD-3-Q

Approximately 3 ft' of film grafted with 4-chloromethyl-styrene was trea ed with trimethylamine in methanol. The mixturewas agitated for 3 days and the film washed with water and air-dried.

Film designation: BD-5-Q

The films prepared and tested are given in Table XI.

40

+ 2 + ++

-4)

+. ++

U3)

0++ +

~4 14+ ++ >~>~ 4Nu

m U+t I I I+ I I+ I I4 I I I I I I

.41

E, 0

4.1

x I.

m +

41

0 01

t44 P-4

0 4) 1)

>R 04- i

>4 0w1 41 A

0~ >q0

a ~ 4 1. :4 1

V) = C

n. XC - 4IC 0 + ad 04 04

E-4 (A Lnu0W1

C 04JCI42

10. Ion Exchange Capacity

Portions of each air-dried film were weighed (BD-5-Q-OH,1.414 g, and BD-5-Q-CI, 1.2971 g) and placed in separate clean250 mL Erlenmeyer flasks. To BD-5-Q-OH was added 100.0 mL of0.10 N HCI and to BD-5-Q-Cl was added 1OLO mL of 40% neutralNaNO 3 . The stoppered flasks were shaken overnight and the chlor-ide content of the aqueous solutions was determined by ion chroma-tography.

A sample of 1.4144 g of BD-5 Q-NMe 3 OH, containing9.10% moisture* consumed 78 of the 219 chloride "units" in 100mL of 0.1 N HCI. This equates to an ion-exchange capacity of2.77 mEq/g (dry wt).

A sample of 1.2971 g of BD-5-Q-Me 3 Cl, containing 7.46%moisture,* released 71.5 chloride units when treated with sodiumnitrate. The standard of 3.56 pg/mL gave 219 chloride units.The ion-exchange capacity of the sample chloride was calculatedfrom these data as 2.71 mEq/g (dry wt).

I. Nerve Agent Simulant Diethyl p-Nitrophehylphosphate

1. Preparation of Standards

Aqueous solutions of 4-nitrophenol (Aldrich 13,023-0)and diethyl-p-nitrophenylphosphate (DNPP)! (Aldrich 85,579-0)were prepared gravimetrically to yield stock solutions of 1.02x 10- M and 9.01 x 10-3 M, respectively. Working solutionswere prepared by 1:100 dilution of the Istock with deionized(milli-Q) water. All solutions were stored at 40*C in the dark.

2. Visible and Ultraviolet Absorption Spectra

Absorption spectra were determined using a Varian-CaryModel 219 UV/VISIBLE recording spectrophotometer. All measure-ments were made using fused s.4.lica cells having a nominal pathlength cf 1 cm.

Grafted derivatized films were ýut into 13 x 50 mmstrips. These strips were inserted diagona~lly into the spectro-photometer cuvet, reagent solution was added, and after thedesired time interval the strip was withoirawn and either theentire spectrum scanned or the absorbance atia specific wavelengthmeasured.

*Moisture based on weight loss of film on heating under reduced

pressure.

43

The films were treated with DNPP in aqueous solution.The absorbance at 273 nm, the most prominent maximum exhibitedby DNPP, was followed over time. These experimental resultsare presented in Table XII.

J.. Preparation of Indicator-Treated Film BD-5-Q-NMe 3 +OH-

A 10 cm x 20 cm piece of BD-5-Q-N-Me 3 +OH- was swirled for60 seconds with 10C mL of 100 4g/mL methanolic phenolphthaleinsolution. The film rapidly became pink. The pink film waswashed twice for 2 minutes with 100 mL portions of fresh methanoland air-dried. Other indicator solutions were used to preparea variety of dyed films. These are listed in Table XIII.

K. Reactions of Simulants with Grafted Films

1. Diethyl Chlorophosphate

A 3 cm x 3 cm piece of grafted film was placed intoa septum-sealed 4 mL vial. One or 2 mL of an ethereal solutionof DECP (41.9 4g/mL) was placed in the vial, which was placedon a reciprocating shaker. The concentration of DECP was followedover time by analysis of the ethereal solution using GC. Theinitial rate of the film/simulant reaction was calculated asthe initial slope of the simulant concentration vs. time curve.

The GC analytical conditions for DECP from clay head-space or solutions, used for the evaluation of reactive films,are given below:

Column: 8 ft x 2 mm-I.D. 3% OV-I on Supelcoport(100/120 mesh)

Oven Temperature: 125 isothermal

Detector Temperature: 250 0C

Injector Temperature: 150*C

Carrier/Flow Rate: He/40 mL/min

Detector: Alkali bead, flame ionization

2. Mustard Simulant-Chloroethyl Ethyl Sulfide

a) A 3 cm x 3 cm piece of grafted film was placedin a septum-sealed vial, 1.0 mL of a 100 pg/mL hexane solutionof CEES was added to the vial and the concentration of CEESwas monitored over time by GC.

44

TABLE XII

ABSORBANCE AT 273 NM OF 9.01 X 10- 5 MDNPP AFTER INCUBATION WITHGRAFTED DERIVATIZED FILMS

Time After Placement AbsorbanceFilm Designation of Film in Solution at 273 nm

(minutes)

BD-l-Q-OH 0 0.799 0.78

45 0.77133 0.76

1121 0.68

BD-3-Q-OH 0 0.8064 0.57

232 0.49342 0.26

BD-2-Q-S 2 0 3 0 0.80238 0.36347 0.14

Control, 0 0.77DNPP + OH- 18 0.72

28 0.6564 0.52

226 0.241440 0.15

Control, DNPP 1440 0.80Acidified c HCI C max 316;

45

be a) 4) 4w~ *4) w4 0 4) a)) I(01 1 w to) 0~ M-q a) 14) w

0) 4)4 C ~ c ~ rr) 14 r.4 0

z Q)

ci 4)3 -4)4) -

Ln I= 4) .S ý -4 b4 4) %4) Q 0x1 P )-4 0 1 4 ) w44 (D -- z .14 144) a .,41 4) 4) 430 ) 0 (0 14 W4

3 4 >4 m .~ 09 oua4 In l 0 U

144

(H 4) -4

u Q *) (1*) c0 04 w) C14) 4143 . 4 14) 01.14 enV m4 4o) 14) ) r.

w~- WU4 Q1 1U 9UQ 01 44 C1 M4 .q

14 H4 u14a . Q

H -4 00 A9ý

4) 0000wmu Q

04 Q) 1 4 >i4J 1 r. z) ý 4),a 4 0 0 4 r 0 0 -14 t r. 14 14:

41 4) U ) r= r. ,4 -44 4) -4 -'

14 0 0>4 00 O~N-40 W' E r 00 s En r:EC W 4 14 ~ =I La z (3 (o r_41~ 0 4) 04-1 .0 14 u 0 (n -4 z 41 .q1 U 4)m~ w4 .4 w144) -. 4 *N.-4 W M 0 0 a~u co U CZ2 0 o IT) a4 a4 a4

.1410

- .f~ . 0 V v 4-4 m' C ~ -4

46

A'

U14 m IWO ao 1, o

a) Z1 X) ,(a-u a) a) 4~a

U >4 >4 ou 0)a a

U)z a)

Ln4 -14 (1 ) a) ýx 0 a a aI3 w .j' 1-4 a) w () :3a

-43 (n) -a) 0 w tow 40 l u u, 0C u

41

00

1-4

H4 0

x 41

~ -4

E-4 '-4 (D 0 0))

47

C: 0 .~ )4J~cz0) 0)- O~5 0 ma~ ~4 a)~ -H r-4 "-4 -. ~4 -4

u It - ) '3 0 'J 0)1-0 4 0 w- Q) -4 -r ý 4 M (>f- 0 0)- >4 > ~ ) .-

(D -4 (Af$ J4

Of )~ -3.a 4* -4 *1 -40

Q) -4r4 (a- 0 '1 w IIq->

0

-~4

E4~

0) CA48

* /

I

b) A 3 cm x 3 cm piece of grafted film was placedin a septum-sealed vial; 1.0 pL of CEES was applied to the filmsurface using a Hamilton syringe. The reaction was quenchedby addition of 1.0 mL of diethyl ether to the vial and analysisof the solution by GC as above. Multiple samples were usedto observe the effect of time.

The GC analytical conditions for CEES from clays andfilms are given below:

Column: 6 ft x 4 mm I.D. 5% OV-I on Chromosorb W

Oven Temperature: 70'C isothermal

Detector Temperature: 250'C

Injector Temperature: 250 0C

Carrier/Flow Rate: He/40 mL/min

Detector: Flame Ionization

L. VX Simulant 0-Ethy]-S-ethylmethylphosphonothioate

A 9 cm' sample of film was wetted with 1 mL of a solutionof EEMPT in ether (2.43 x 10-3 M). EEMPT in the film was analyzedby GC over 3 hours. No 'loss of simulant was observed. Theevaluation was repeated with a 1:10 dilution of the solution;however, again no loss of EEMPT could be observed.

Samples of £ cm, of the phenolphthalein-treatedBD-5-Q-NMe 3 +OH- were treated with 1 pL samples of EEMPT. Thecolor of the film was lost at the contact points. The filmwas folded to increase the contact surface with the EEMPT.At specific times the reaction was quenched by adding 4 mL ofether. After thorough extraction of the film, the solutionwas analyzed by GC. The results showed 40% removed at 1.5 hours;52% removed at 2.5 hours; and 61% removed at 3.5 hours.

The results of the films are summarized in Table XI.

49

4,

III. RESULTS AND DISCUSSION

A. Introduction

The detoxification of chemical and biological agents canoccur by a number of mechanisms, by adsorption or by nucleophilicor oxidative reactions. This project was designed to test thepossibility of developing reactive films and powders with adsorp-tive or nucleophilic and/or oxidative activity by grafting reac-tive groups onto polyethylene films and then forming quaternarysalts from these functions. Nucleophilic and oxidative anionswere then added as counterions of the quaternary ammonium ion.Indicator dyes were added to some of these films to evaluatethem as detectors for chemical agents.

Native Irwin or bentonite clays were processed by Anglo-,American Clays Corporation by proprietary processes to benefic-iate the clays or to introduce reactive amines, large surfaceareas, or metal ions intercalated between the layers. Theseactivated clays had the potential for providing nucleophilicor oxidative reactivity by catalytic processes. The processingalso increased the adsorptive capability of the clays. Theproperties of these clays which are available to SwRI are givenin Table Ii.

The effectiveness of these decontaminating agents was evalua-ted by use of simulants with reactivities similar to those ofthe common chemical agents. The experimental methods for deter-mining the effectiveness of the films and clays in decomposingthe simulant3 proved to be difficult to develop; however, satis-factory methods were found. The simulants are given in TableI.

B. Choice of Simulants

Chloroethyl ethyl sulfide (CEES), the simulant for theblister agent mustard (HD), was selected for its similarityin reactivity to HD. CEES was obtained from Aldrich ChemicalCompany and used without any further purification. The a-chloro-ethyl group is subject to nucleophilic displacement and thesulfide group will undergo oxidation to the sulfoxide. In evalua-tion of the reactive films, diethylsulfide (DES) was used toevaluate the oxidizing power of the film to form diethyl sulfox-ide.

The simulants for the G-type nerve agents were also chosenfor their similarity in reactivity with the fluoromethylphosphon-ates. Diethyl chlorophosphate (DECP) and diethyl-p-nitrophenyl-phosphate (DNPP) were obtained from Aldrich Chemical Companyand were used without any further purification. They both arehighly reactive with 'nucleophiles.

50

p..

The simulant for VX nerve agent was O-ethyl-S-ethyl-5-rrethyl-phosphorothioate (EEMPT), which was expected to be less reactiveto hydrolysis or oxidation than the chemical agent. Thus thesuccessful reaction of EEMPT with a decontamination system wouldbe a strong indication that VX would be detoxified. The EEMPTwas obtained from the Contracting Officer's Technical Representa-tive. The properties of these simulants' are given in TableI.

C. Clays

1. Analytical Methods

The analytical procedures for evaluating the modifiedclays listed in Table II evolved during the contract period,so no single test procedure was used for all these samples.The effectiveness of the powders could depend on the propertiesof the powder to undergo reaction with the simulant to producenontoxic products. Alternatively, the powder could adsorb thesimulant so tightly as to reduce the vapor pressure of the simu-lant, and thus render it nontoxic. The initial test procedureswere designed to evaluate the combination of these two effects.This method was the Headspace analysis (Section II.A), whichmeasures the concentration of the simulant in the vapor stateabove the clay which had been treated with simulant. The resultsof this test were erratic and did not distinguish between reactionand adsorption. We felt that reaction or irreversible adsorptionwas more important to measure, since weak adsorption could leadto toxic effects by slow desorption of agent., Thus we evaluateda reactive analytical method called the kinetic analysis. Thisprocedure measured the disappearance of simulant oný treatmentwith the powder and thus measured the reactivity of the powderwith simulant.

a) Headspace analysis

The first procedure was to inject a known volumeof the simulant into a weighed amount of clay and then measurethe amount of simulant in the gaseous state in the headspace.Additional simulant was added until the amount of simulant inthe headspace approached that found above liquid simulant withno clay added. This method was used to evaluate the samplesA'-E'. This procedure should indicate a combined effect ofadsorption and chemical reactivity for decontamination and theresults are shown in Figures 1 through 5.

Samples of Irwin (A' ) and sodium bentonite (B'control clays, these clays grafted with 1% DETA (C' and D'),and sodium bentonite grafted to 10- DETA (E') were received

51

from Anglo-American Clays. Clays B' and E' were treated, withmethanolic sodium hydroxide in an attempt to increase the nucleo-philicity of the clay. The infrared spectrum of E', the graftedclay, after base treatment showed that the amine had been removedby the treatment.

All of the clays except El showed a slow increasein DECP in the headspace as additional simulant was added (Figures1 and 2), suggesting that the simulant was not reacting withthe clay but being adsorbed onto it. However, El required 500iL of DECP/g of clay to give a gas chromatography (GC) peakheight approaching that observed with neat simulant. This per-formance was clearly superior to that of clays A'-D'. Also,this clay after treatment'with hydroxide was still more effectivethan the untreated sodium bentonite.

The Surrey Finest Fuller's Earth performed similar-ly on treatment with DECP, but it was not as effective as claysA'-El. This was in agreement with the resultf presented inthe Battelle reportl which showed that clays from the UnitedStates were as effective as the Fuller's Earth from Britainin detoxification of chemical agents.

The evaluation with' the blister agent simulantCEES showed that, clays A'-E' has about the same effectiveness,with E' requiring slightly more simulant to give the GC signalcomparable to that of neat simulant'. None of the clays wereas effective in adsorbing CEES as they were with DECP (Figure3).

This headspace analytical procedure was usedto evaluate the next series of clays, which also had had amineor other nucleophilic moieties added to the structure. Theresults of the tests are given in Figures 4 and 5. These datasuggest that the modified clays were no better thain,' and perhapsinferior to, the control clay in decreasing the vapor pressureof the simulant DECP. The analytical results, however, showedconsiderable scatter and did not allow distinction between adsorp-tion and reaction with the simulants.

b) Kinetic analysis - 6 minute test (see SectionII.B and C)

A second procedure was developed for clays 1-6to detect reaction with simulant. A 200 mq sample of modifiedclay was treated with 40 iL of DECP and the mixttire stirredfor 1 minute. After 5 minutes 2 mL of ether wz-s added to thesample. A 1:100 dilution of the ether, solution was analyzedby GC using an alkali flame ionization detector tAFID). Thismethod showed that the modified clays 1, 2 and 3 were causing

52

greater decomposition of the DECP than was the control clay(see Figure 6).

The clays had been shown by Anglo-American Claysto complex with metal ions such as copper. Accordingly thebetter clay, 2, was treated with copper ion as a 1% solutionin ammonia (5 g of clay with 50 mL of solution). The mixturewas stirred for 5 minutes, filtered and washed with water andmethanol. After drying in an oven at 60'C, the clay was evaluatedas before. The results (Figure 7) showed that the copper treat-ment increased the decomposition of DECP by 100%.

Several experimental approaches were attemptedto determine a standard procedure for comparing the adsorptionand/or reactivity of these clays. A ratio of 40 pL of simulantsuch as DECP to 200 mg of clay resulted in the loss of about30% of the simulant after 5 minutes, regardless of which claywas used. The clay 4375Tw/Cu 2 + was extracted with ether andthe ethereal solution analyzed by GC. Using 1 pL of DECP gave41% recovery while 10 wL gave 61% recovery and 40 uL gave 66%recovered DECP. This suggests that 40 pL saturated the 200mg of clay.

c) Kinetic runs - Variable time (see Section II.D)

For comparative purposes a standard procedurewas chosen touprovide a crude kinetic measure of the effectivenessof clays 1-6, 4374, and 4375. Several 2J0 mg samples of thedifferent clays were treated with 5 .L of a 20% solution ofDECP in ether. After careful mixing, the samples were allowedto stand for a measured length of time, after which each samplewas extracted by addition of 2 mL of ethyl ether with stirringby a vortex mixer. After separation of the clay a 3 wL samplewas analyzed by GC using an AFID. The percentage loss of simulantwas determined at intervals during the kinetic run.

The results of these kinetic runs were expressedas plots of percent simulant remaining vs. time, which couldbe expressed linearly, in natural In/ln plots. In this formthe y-intercept indicates the amount of simulant destroyed Iminute after mixing; the smaller the intercept, the greaterthe decomposition of simulant. The slope of the curve indicatesthe rate of further decomposition of the simulant; the morenegative the slope, the faster the rate of decomposition. Thesedata are summarized in Tables IV and V.

These data indicate that chemical modificationof the clays decreases their effectivneess in adsorbing/reactingwith the simulant CEES, as the unmodified control clay (Number6) produced the greatest disappearance of the simulant. Also,

53

copper treatment of the clays decreased their effectivenessin destroying CEES. Thus the most effective clay with CEESwas the untreated control clay, 6.

With the nerve agent simulant DECP, the situationwas different. The untreated control clay was effective indestroying the simulant; however, the clays having ethylenediamine side chains were more effective. Treating the clayswith copper increased their effectiveness in each case. Themost dramatic effect was exhibited with clay 4375T, which hasthe ethylene diamine chains only on the surface.

d) OWasher test" - Estimation of adsorption anddecomposition

The evaluation using the kinetic analysis (SectionII.F) measured the deactivation of simulants by the treatedclays based on the amount of unchanged simulant that could beextracted from the clay sample after a given reaction time.This method of analysis measures the amount of simulant whichunderwent reaction with the clay sample. Any simulant thatwas ineffectual because of adsorption by the clay would beextracted. Thus the analytical method was not really the 'bestmethod for estimating the total decontamination potential ofthe clay by both adsorption and reaction. The analytical methodbased on evaluation of the simulant vapor in the headspace abovethe clay samples (Section II.A) did consider the amount ofadsorbed simulant; however, this method was difficult to reproduceand required a volatile simulant.

The Contract Officer's Technical Representative,Dr. Millard Mershon, suggested a method for evaluating the com-bined adsorption and decomposition by measuring the breakthroughtime of a plug of clay held in a washer using M8 paper as thedetector. This method involved packing the powder' or clay ina washer affixed to a sheet of M8 paper. Simulant was addedand the time to a color change in the M8 paper was recorded.The details of this procedure are described in ExperimentalMethods (Section II.E) and the results are summarized in TablesVI and VII. Clay 6 treated with copper sulfate was effectiveagainst the HD simulant.

The nerve agent simulant DECP was effectivelyblocked by clay 4375T, a sodium bentonite clay in which ethyleneliamine was attached only to the surface.

These two treated clays were not the best ofth•s eries in the extraction test. As might be expected, theAds, rotive power of the clay was not directly related to thereric,-.,ity of the clay with the simulant.

54

The reactivity of clays 4952-4955 and 5030 wasestimated by the kinetic experiments and the "washer test" (TableVIII). The results in the kinetic test showed that CEES wasrapidly destroyed by all of these clays. Only 28-45% of theagent remained after 1 minute and less than 20% at 4 minutes.The decomposition of CEES was probably catalytic with 4952,since doubling the amount of simulant to 10 pL/200 mg did notgreatly change the rate of disappearance of CEES: 42% of CEESremained after 1 minute and 18% at 4 minutes.

In another series of tests, 5030T was comparedwith Fuller's Earth and found to be only slightly more effective.These results also showed the Fuller's Earth to be approximatelyas effective as the treated clays in this kinetic test.

As was evident from the "washer test," the treatedclays had a greater decontamination potential than Fuller'sEarth. Probably the kinetic experiments were run with too higha ratio of clay to simulant. Using a larger amount of simulantdemonstrated the superiority of the treated clays over Fuller'sEarth.

e) Kinetic test - 4 minute test

The previous methods using 5 or 10 pL of simulant,and measuring the simulant remaining at 4 minutes and longertimes did not show much difference among the samples of treatedclays and Fuller's Earth. Most of these clays gave about 80%decomposition, which suggested that a better comparison couldbe obtained by increasing the amount of simulant and keepingthe time of reaction constant at 4 minutes. This approach wasused in evaluating the last series of clays,, RT001 through RT009,received from Anglo-American Clays Corporation.

A 200 mg sample of finely ground powder was treatedwith a measured (5-100 PL) amount of simulant and mixed in avortex mixer for 1 minute (Section II.F). The mixture was allowedto stand for 4 minutes. Then 2 mL of diethyl ether was addedand the powder extracted by vigorous stirring. The solid wasseparated by centrifugation and an aliquot of the ether analyzedby GC-AFID. A simulant solution of known concentration wasused as the standard arj the results were expressed as percentsimulant remaining after 4 minutes.

(1) CEES

Table IX and Figures 8 and 9 sh-w the resultsobtained with the samples RT001 through RT009 with CEES. Vhedecomposition of 20 pL of CEES was about 90% with sample RT004

,, i i i i i

and 75% with RT006. RT007 and RT009 were tested with 30 PLof CEES; however, the results suggest that they too would havegiven extensive decomposition with 20 PL. All of the clayswere greatly superior to Fuller's' Earth.

Longer reaction times gave more completedecompositicn of the simulant. That was demonstrated with the5 .L samples reported in Table III and was confirmed by observingthat of 40 PL of CEES, only 44% and 36% remained after 30 minuteexposure to RT004 and RT006, respectively. These clays gave50% and 52% at 4 minutes.

The addition of CEES or DECP to RT008 causedthe solid to develop a yellow color. Clay RT009 is grey andon addition of the simulants became deep black.' It was clearthat the two clays indicate the presence of these two simUlantsand could indicate the completion of the decomposition.

Both RT00'8 and PT009 have metal ions inter-calated into the clay crystal lattice. The color changes suggestthat metal-intercalated clays could be used for developing asolid which could detect contaminated areas and indicate whendecomposition was complete.

Late in the 'last quarter of this contract,the Contract Officer's Technical' Representative provided samplesof polymer resins from Rohm and Haas for comparison with theclays received from Anglo-American Clays. These polymers werelabeled A through F and were described as being mixtures ofa' strong acid resin, a strong base resin, and 'a carbonaceousadsorbent. These samples differed in the ratio of the threecomponents.

The polymers of Rohm and Haas were studiedin a similar manner to that for the clays. Except for SampleA, for which 64% of CEES remained, there was no, apparent decom-position of 5 jtL of CEES within 4 minutes. These results aregiven in Table XIV. It was clear that no appreciable reactionhad occurred between CEES and polymers ' B through F. PolymerA showed less reactivity than the worst of the modified clays.

(2) DECP

The application of the 4 minute kinetictest with DECP proved to have 'a pitfall. Some of the powderstreated with DECP were not wetted by the simulant; this factorprobably slowed the reaction. Other powders formed balls aroundthe simulant. Ball formation interfered with the ' interactionof more powder with the simulant and, of greater importance,interfered with extraction of the simulant by ether in the final

56

TABLE XIV

REACTIONS OF ROHM AND HAAS RESINS WITH CEES

Percentage of Agent Remaining After 4 MinutesResin Reaction with Increasing Volumes of Simulant

5 I.L 10 jjL 20 pL

A 64% 62% 75%

B 100

C 100

D 99

E 99

F 92

57

analysis step. For example, in the test with clay RT009 only4% of a 60 pL sample of DECP was extracted with ether. If thelumps were carefully broken during the extraction, about 80%of the DECP was recovered unchanged.

This observation may be highly significant,for if these powders were used as decontaminants, they mightencapsulate G agents as they did the DECP. If a G agent werenot extractable by ether in this condition, it would probablybe ineffectual against a barrier composed of the powder. Thus"a physical decontamination may have been achieved even though"a complete chemical decomposition did not occur. The same effectwas observed with RT004, RTO06, and RT008, although to a muchsmaller extent. This encapsulation of simulant by RT009 occurredeach time. The results of the 4 minute test with vigorous stir-ing of ether' during extraction are given in Table XV.

The results of the test with Rohm and Haaspolymers showed high percents of decomposition with samplesA-D (Table XVI). The results with 10 •L were probably rot dueto the encapsulation effect observed with RT009 for a rerunof resin A with careful mixing gave 21% recovered DECP.

Summary

With the HD simulant CEES the modified clays were clearlysuperior to the polymers. The better clays were:

-RT004-RT006-RT007-RT009

With the 'G-agent simulant DECP, RT009 showed an unusualproperty of encapsulating the simulant and preventing its extrac-tion by ethe . For- decomposition of the simulant, the polymerswere probably superior to the clays, although the data dc notrule out an encapsulation mechanism for polymer performance.

f) Modified washer test

In tests of a plug of powder as barrier, thebreakthrough time and the amount of simulant required for thebreakthrough should indicate the combined adsorptive and reactiveproperties of the powder. Since the density, particle size,and degree of packing could vary in preparing the plug, a numberof different techniques were used to develop a star.dard procedure.

In the first variation of the washer test (SectionII.E), washers with outer diameter of 22.5 (±0.02) mm, inner

58

TABLE XV

DECOMPOSITION OF DECP BY CLAYS

Percentage of DECP Remaining After 4 Minutes

Reaction with Increasing Volumes of Simulant

Clay 10 oL 20 30 40 60' 80 100

RT006 61 68 77 79

RT009 52 84 99 94

59

- / V

TABLE XVI

REACTION OF ROHM AND HAAS RESINS WITH DECP

Percentage of Agent Remaining After Reaction for

4 Minutes with Increasing Volumes of Simulant

Resin 10ijL 60 jiL 100 pL

A 2.8 21 43

B 8.1 52

C 4.6 30 67

D 0.9 61

E 44 79

71 85

60

diameter of 9.7 (±0.02) mm, and thickness of 2.3 (±0.1) mmn werelightly packed by pressing the clay gently with a microscopeslide. The excess clay was removed by scraping the top of thewasher with the slide. The results of this test method withRT001-RT006 and 4952-4955 are given in Tables XVII and XVIII.With CEES the results were consistent and reasonably reproducible.With DECP, however, some of the powders swelled and channelingoccurred. This led to erroneous results. Also the thicknessof the plugged washers gave long test times with the moreeffective powders.

In an attempt to improve the test, a 5 mLround-bottom flask was pressed into the clay plug to compressthe powder further and to indent the surface to form a wellfor the simulant. Results with this method were difficult toreproduce and erratic. With Fuller's Earth and some of theresins, DECP did not penetrate because of its viscosity andits failure to wet the powder. These results would be misleading,for an effective decontaminant must wet the powder.

The best reproducibility with CEES was obtainedby a further modification. The washer, attached to M8 paper,was packed with powder with pressure from a microscope slideuntil excess powder was squeezed out over the sides of the washer.The excess powder was removed by scraping the slide over thewasher. A 5 mL round-bottom flask was pressed tightly againstthe inside rim of the washer. The weight of powder in the washerwas determined. Simulant was added (10 pL CEES or 20 1.L DECP)every 4 minutes until breakthrough was observed by a color changein the M8 paper. The results obtained with CEES are given inTables X and XIX.

Since the amount of powder packed into the washerby this method varied widely, a factor (pL sirjulant X time/mgof powder) was calculated to normalize the data. Of the Anglo-American Clays samples, RT004, RT008, and RT009 gave promisingresults. With the Rohm anr' Haas polymers, Samples D and F gavepromising results. Unfor unately the highly promising SampleF was not available in rifficient quantity to repeat the tert.In view of the results of the 4 minute kinetic evaluation thisvalue seemed questiongie (see Tables XIV and XVI).

Values for DECP evaluation by the washer testhave not been included, since reproducible results were notobtained. For example, one run of Fuller's Earth did not showbreakthrough even though no additional DECP could be added withoutoverflowing the washer. Similarly some of the clays and resinsformed solid layers on the surface of the washer which did notseem to be penetrated by DECP. However, the color which developedon M8 paper after liquid broke through RT004 and RT006 was neither

61

TABLE XVII

RESULTS OF WASHER TESTS ON CLAYS RECEIVED 8/84USING CEES AS SIMULANT

InteractionClay Replicate Time (min:sec)

RT001, Acid Activated (a) 1 11:00-2 5:303 11:00

RT002, Acid Activated (b) 1 10:302 11:303 10:00

RT003, Beneficiated 1 22:002 17:003 14:30

RT004, Beneficiated-Calcined 1 36:002 39:003 28:00

RT005, Al+ 3/Cr+ 3 Intercalated 1 17:002 15:003 16:00

RT006, Cr+ 3 Intercalated 1 16:302 14:003 14:00

4953 1 32:302 28:003 18:00

62

S0 00 0 a,0 z - -4 0. r LfNE-4 5a) m C1 (N -4D

<V C

,- '" -4

Ow0) a' 0- 4

0 W 0 a)4J Wo 0w 0 0 m 0>.C > .. 0 M 0 U") 0

41 4n E-4- L W CD m

01) LW 0 E -40 0 a) - i 0 0

ý4 -4 41J

E-0 -

E-

"E- - (" )w 4.1

U -) tr 0- C)Z ra V) -4 C)

0 0 0 0 01 0

4.) -4H r n 0f C LC) (N 0D

M- E-44

0 Q)uC)

-W H

r4'))W (1)LI 41 "a-

E-4 -4ON

- U 4J u)Ua) ý4 -4C) a) w) mU(

(U) V4 41

0) w 41 0 0- ~ - m 4(

a) 41 ) ~-4 -411 C

-- 41 E- 0, C 4C

63

4f

TABLE XIX

EVALUATION OF ROHM AND HAAS RESINS WITH TIGHTLYPACKED WASHERS USING CEES

Breakthroughwt Time Total Simulant pL

Material (mg) min:sec Simulant 1L X sec/mg

A 76 0:51 10 984 1:30 10 11

B 80 1:50 10 1475 4:29 20 3683, 4:10 20 30

C 78 1:40 10 1385 1:16 10 9,

D 68 6:02 20 10666 5:40 20 104

E 95 1':20 10, 879 0:33 10 4

F ,52 12:35 40 580

64

that observed after breakthrough of the other clays and resinsnor the color caused by DECP. This suggested that DECP hadundergone reaction with these two clays and the material passingthrough was no longer DECP.

2. Attempts to Determine the Reaction Products

Acetone extraction RT004 after treatment with DECPshowed only a small amount of DECP in the extract. An extractof the clay with deuterate dimethyl sulfcxide (DMSO-D 6 ) showedtwo 3 1 p signals in the phosphorus nuclear magnetic resonance(PNMR). One of these was due to unreacted DECP and the othersignal was at a chemical shift expected for (EtO) 2 PO 2 H or itsanion.

A washer test was run with RT008 and DECP modifiedso that liquid on breakthrough was caught on a sodium chlorideplate and the infrared (IR) spectrum was measured using theFourier Transform Infrared. In the spectrum, shown in Figure10, DECP is indicated by the background absorption and a largepeak at 1035 cm- 1 , which may be due to phosphate ion. Trisodiumphosphate absorption is shown as a broad peak at 1015 cm- 1 . 5

Many other species absorb in this region of the spectrum, however.The spectrum of the DECP used is given in Figure 11, showingabsorption assignments.

Similar attempts were made to determine the materialremaining on the clay powders after treatment with CEES. Theratio of CEES to clay RT004 was based on the data in Table IXto ensure the presence of some unchanged CEES. The mixturewas extracted with ether to remove the excess CEES. After drying,the clay samples were extracted with D6 -DMSO. The proton nuclearmagnetic resonance (IHNMR) of the D6 -DMSO extract (Figure 12)shows water and ether as impurities in unchanged CEES (Figure13).

These experiments suggested that CEES and DECP reactedwith the oxygens of the clay by a nucleophilic mechanism. Theproducts of these reactions, hydroxyethyl ethyl sulfide fromCEES and diethylphosphate from DECP, were bound to the claysurface througih a bond with oxygen. Since these hydrolysisproducts were bound to the clay, they were not extracted bythe solvents used. With DEC? the reaction may have been cata-lytic, so water present in the cry'stal lattice released diethyl-phosphoric acid, which was the first breakthrough product.

65

(.) ulsgalzul gI;un XAVII~qiy uo~ss~msuvalI 2uaoa.Zd

00

0

e-

00

66.

00

CDU

In670

.d

4. 41

a -4

co

1 Z A.

X -4

68 U).'

4sf 4

j4

4 w

69

D. Grafted Films

1. Preparation

The irradiation of polyethylene films in the presenceof reactive vinyl groups was used to graft 2-vinylpyridine,4-vinylpyridine, and 4-chloromethylstyrene onto the inert polymerfilm. These reactive functions were converted to quaternarysalts by reaction of methyl iodide with pyridine and of trimethyl-amine with 4-chloromethylstyrene. The products of these reactionsare illustrated in Diagram 1.

Films prepared by gamma-irradiation grafting of poly-ethylene with 2- (see A) or 4-vinylpyridine (see B) were treatedwith hydrobromic acid and bromine to give the hydrcbromide per-bromide. The chlorochromate salt was also formed. The parentfilms were converted to the methyl quaternary salts with methyliodide or dimethyl sulfate (see C and D). By an ion exchangeprocess, the quaternary salts of the two series were convertedto the polyiodide, azide, hydroxide and thiosulfate.

Because of the problem of possible covalent bond forma-tion between the nucleophiles and the grafted polymers withquaternized pyridine rings and the oxidization of the ring byoxidizing anions, a reactive polyethylene film was preparedfrom 4-chloromethylstyrene (see E). This film was then treatedwith trimethylamine or triethanolamine. The quaternary trimethyl-amine salt was formed, as shown by infrared spectral analysis,but triethanolamine did not form a salt. The quaternized film(BD-5-Q) was converted to the hydroxide, azide, perchlorate,chlorate, permanganate, dichromate, persulfate, perborate, m-chloroperbenzoate, and hypochlorite salts (structures givenin Diagram 1 and Table XI). The structures of the films areshown in Diagram 1.

2. Evaluation

These films were evaluated using dibutyl sulfide orCEES as simulants for the blister agent HD. DNPP, paraoxon,'was used as a simulant for G nerve agents and EEMPT was usedas a simulant for the nerve agent VX. A summary of the resultsis given in Table XI.

The HD simulants were evaluated with representativeexamples of all of- the films, as described in ExperimentalMethods. Using this procedure, the quaternary periodides andazides gave no reaction after several hours. The periodidewas too poor an oxidizing agent to convert sulfides to sulfoxides.The azide was expected to cause the nucleophilic displacementof the halide of CEES. The lack of reactivity may have resultedfrom covalent bonding of the azide ion to pyridinium moiety,giving dihydropyridines which were no longer nucleophilic (seeEquation 1). The blue color of the film was also consistentwith this hypothesis.

70

DIAGRAM 1

STRUCTURE OF THE REACTIVE FILMS

A B

BD-1-HBr, X=Br BD-2-HBr3 X=Br3

BD-l-HCrO3C1, X=CrO3cl BD-2-HCrO 3 C, X=CrO3Cl

6 CH3 +

C D CN3

BD-l-Q13 1 X=1 3 BD-2-Q1 3 f X=13

BD-l-QOH, X=OH BD-2-QS 20 3 f x=S20 3BD-l-QS2 03 1 X=S2 0 3 BD-2-QN3, X=N 3BD-l-QN 3, X=N 3 copolymner with divinylbenzene

BD-3-QOH, X=OH

( CH2 CH)n & -C2-H

Me N ~ ~ 1IE CH 2 C CH2N(CH3)3 X

BD-5-QX, X

X=N3, OH, C10 4, C10 3

MnO4, Cr2O7, SO5,

71 m-C1C H Co-, C10-

Equation 1

--- (CH2-H)n - (CH2-CH)n

N3- + N3

3 CH3BD-2-Q-N 3 covalent

bond

The hydrobromide perbromide forms of the films, preparedfrom polyethylene grafted with 2- and 4-vinylpyridine, werea reddish-brown color. Any reaction which resulted in reductionof the perbromide was accompanied by loss of this color. Thereaction of either film with dibutyl sulfide or CEES causedan immediate color loss at the site of contact, and analysisof the product showed the formation of the sulfoxide correspondingto the sulfide used. With CEES nucleophilic displacement ofC1- by Br- also occurred (see Equation 2).

Equation 2

BD-l-HBr3_,,IC2HBR-S-CH2 CH 2 Cl D-IH R-T-CH2 CH 2 Cl + RS-Ci 2CH2 Br

R = CH 2 CH 3

Since the simulants were used neat, the reaction rapidlydepleted the available perbromide at the site of contact onthe film. Reaction continued at a rate controlled by the dif-fusion of the simulant through the film. The film became trans-parent and colorless as the perbromide was depleted. The rateof the reaction could be followed by measuring the area of thecolorless spot. The principal product of the reaction, evenafter several hours, was the sulfoxide corresponding to thesulfide used.

The perbromide films BD-I-HBr 3 and BD-2-HBr 3 gavea faint odor of bromine. The stability of the films was deter-mined by allowing them to stand and then reevaluating theirreactivity. Even after 7 days, BD-2-HBr 3 still converted sulfideto the sulfoxide. These films gave no reaction with the simulantsDECP, DNPP or'EEMPT.

72

Samples of the hydrobromide perbromide of the 2- and4-vinylpyridine-grafted films (BD-l-HBr 3 and BD-2-HBr 3 , respec-tively) were given to the COTR of this project for evaluationof detoxification of surety chemical agents.

The success with grafted oxidizing agent perbromidesuggested the testing of other oxidizing films. The chlorochrom-ate form was prepared from the 2- and 4-vinylpyridine filmsto give BD-1-HCrO3 HC1 and BD-2-HCrO3Cl. Neither of these filmsreacted with the sulfides.

The quaternary hydroxide BD-l-Q-OH gave a very slowreaction with DNPP (Table XII). BD-3-Q-OH, prepared from the4-vinylpyridine quaternary salt, reacted rapidly with DNPP,as did hydroxide ion.

The difference in reactivity of the quaternary hydrox-ides BD-l-Q-OH and BD-3-Q-OH (see Table XII) may reflect thedifference between covalent bond formation with 2- and 4-pyridin-ium hydroxide (Equation 3). In view of this possibility, thefilm RD-5-Q was converted to the hydroxide form, since the tri-methylammonium hydroxide could not form a covalent bond. Thisbenzyl film BD-5-Q-OH reacted with all the simulants (see TableXI).

Equation 3

0-(CH2CH)n (CH 2 CH)n -(CH2CH)n

HNCH3

+ OH-(CH23+J

I ICH3

C& (CH2CH)n

N H

A film with a good nucleophile was prepared by exchang-ing chloride ion with thiosulfate to give BD-2-Q-S 2 0 3 . On reac-tion with CEES, the peak in the GC for the HD simulant disappeared

73

and no new signal was evident. This observation was consistentwith a nucleophilic displacement of the chloride ion by thiosul-fate, the product of which would be non-volatile and associatedwith the film.

The most reactive film with a solution of DNPP wasthe thiosulfate form (BD-2-Q-S 2 0 3 ). The reaction was ev Wlatedby the change in ultraviolet absorption (Table XII). The 12tailsof the evaluation procedure are given in Experimental :iethods.The disappearance of ultraviolet absorption for the simulantwas very rapid; however, the nature of the product was inquestion. Hydrolysis of the simulant would give thep-nitrophenolate ion, which is a deep yellow. This was observedwith BD-3-Q-OH. However, addition of hydroxide to the BD-2-Q-S 20 3film after reaction with DNPP did not givu a yellow color.Thus another nerve agent simulant must be used with this thio-sulfate film. The V simulant EEMPT did not react with thisfilm.

A preliminary, qualitative test of BD-l-HBr 3 and BD-2-HBr 3 with EEMPT gave no color change witb, the films.

The BD-5-Q films with oxidizing anions were surprisinglyinert. The perchlorate, chlorate, permanganate, dichromate,peroxyacid, and hypochlorite did not react with EEMPT and reactedwith DECP only if the film was moist. The colored films frompermanganate and dichromate gave no change on treatment withCEES; however, there was a decrease in the GC peak for CEES,EEMPT, and DECP when the hypochlorite or peroxyacid films werewetted and then treated with CEES (Tables XX, XXI and XXII).

3. Detector Films

Several of the films were colorless because of theanion present. Thus the films containing periodide, perbromide,chlorochromate, permanganate, dichromate, and pyridinium azideeach had a detectable color which changed on reaction of theanion with a simulant or other agent. Unfortunately these filmsdid not have the best reactivity with the simulants of chemicalagents. The more reactive films were BD-5-Q-OH, BD-5-Q-N 3 ,and BD-2-Q-S 2 0 3 , all of which were lightly colored or colorless.

It was therefore desirable to add a dye or indicatorto the more reactive films to provide a mechanism for monitoringthe reaction by a color change. Since these films were allnucleophilic in reactivity, acid-base indicators were appropriatefor detecting reactivity with the nucleophilic anion. Afterwashing the grafted films in the appropriate solution of theindicator dye, reaction of the films with simulants gave colorchanges, as indicated in Table XXIII.

74

0l 0

10 a

c C

0. 0 0 0 - 0I I 'a i- 'a

1.40)

x 0)

04)

24 41

0 11 ya a -

.. >. 'a 4w f -l z

6. I 4d 1. 414 4). c

C, 'a 6. a 0 0 ~o1 ti ZZ t I, VI 'W 00 Z

0 11

oI 3 134

0 0

4JI

75

4.41

as4 0.4

-4 en".3C r.) C6 co w 4) 4 $

0% 0 0 0

R. r. 2 Cm 0) 0 0 014 4 4 .0 41 41

3 ~ E-4i ASo0 -. .i14 1.4 -4

-w 'a 0 0 0 0co (M Z r- co N N z 4

ok 00J 4

Cu '- "4 u CuCE-4

141

C 00-

-4 Z E-4 4

0 i I : co (

*j 01N 4 a 4

410U 0 0 4 0 U

C, - -o Z 04)

r *1-4C 44 4J)I 4

=1 Xna z m J, '4-- n * J P "414)01 u I 1. N: co .C QIu 1 0 z 0" )).0 Nn on~ 0 00 ~

4.O U 0 03(J764

E* - / -N, -4

00

E-4'-

04 IN N I- Io (N I 1-1

'0

0rI u E

I I(O 0 I I I I I I 0" 00

-4-

H 0E

E- 0

~-4 C:~

o• -4'4 • 4.)

•" r- -4 • •••

.•0 0 0. • o1 0EZ:

coV 4.4

/ 0L r- 0 Z)00

IH- ---4• .t • • ., r"

0 J4

a, 0

r- co .1 :

4 j -44 - 4 -

41 .04

(1 X4-* U-4 4.1 a) 'd. V W-4 Lfl4-I 0 - (N Q) 4U1

r ý4 -o 1 1 4.) 1-1 4)J2 (n

~C) 0 OJ U U3 0W41 0 41i ~ U 0 0 Lo w 0

0100 0 0 0' 0 0' C 041'- 0 01 4.) 4.) 4.) I l-H E-4 - 41)

1.4 1U u co m I I co a).~ W

77

w 0) ww w1

$44

C: ý4 ~-4 Xaci21 (00 .ý 4 w~ 0 ) m~ 0

w ~ Q0 -.4$. W~i a) IV w

040

w 0 0) w ) 4)

0 (1C ) 0 w4 0 w 4 0,04 mr 0 m4 a > 4 0 u 0 a Cu a

Co

04

~~~~- f.4 S4t- r'(1) 1.4 >1 4.) Ud r-4 41 0r.$04 (d x t( CA Q) (a

14 04 > r

-4 -4Q 3 N-4$

m 0 l - 04 - 4a

I 478

0C4 0 0

al - - 0 Q~) - avcU -4 to 1-4 0)- p

>4 0 e >4 > u

U) U

in E 1

o 14 -)O 3 0J -4 -14 w- -4 0-4 - 0 O0 C -4 ~

8 CJ- 4 u on

1.4

414

14 U

79 C

E. Polymeric DS-2 Analogs

1. Films with Grafted DS-2

The nucleophilic films evaluated previously showedlittle or no reactivity with the simulant for HD. Sin e DS-2(a sodium hydroxide-containing decontaminant) is highly e fectivefor deactivation of HD, the grafting of DETA onto the fil wouldbe expected to improve its reactivity.

Polyethylene film grafted with vinylbenzyllhlorideshould provide the starting material for introducing t e DETAmoiety. However, no reaction of neat or diluted (aqueous) aminewith the benzylchloride function occurred. Only after eatingthe benzylchloride film with neat DETA for 65 hours was a signifi-cant amount of the amine introduced onto the film. A n npolarsolvent such as toluene assisted the reaction; however, thefilm disintegrated at high temperature. At room temperatureDETA in toluene was introduced in only 24 hours.

2. Polymers with DETA Side Chains

Since reaction of the benzylchloride group Dn thefilms proved difficult, a better approach seemed to be to preparea poly DETA styrene to obtain the maximum density of reactivesites.

The preparation of the polymer has been at emptedby two different routes. 4-Chloromethylstyrene was reatedwith DETA to introduce the DETA side chain on the monomer whichwas then polymerized. The second approach involved the conversionof chloromethylated polystyrene, prepared by polymerL zationof 4-chloromethylstyrene or by chlormethylation of poly tyrene(Bio-Rad), to the DETA polystyrene by reaction with DETA.

The chloromethylated polystyrene from Bio-Rad hadabout 64% of the phenyl rings substituted. Reaction of thepolymer with DETA in toluene gave complete substitution of chlor-ine atoms. Thus this polymer had nearly the theoretical naximumnumber of DETA units.

The polymerization of 4-chloromethy1styrene was accom-plished by benzoyl peroxide catalysis to give 80% yiel~J of awhite solid polymer, which was soluble in methylene chlorideand insoluble in methanol. A solution of the polymer in me hylenechloride was treated with 2 molar equivalents of DETj. Themixture jelled within 15 minutes. More methylene dic loridewas added; after stirrinq for 3 days, water was added tc removethe. DETA hydrochloride. The remaining solid was the Dolymer

80

containing a DETA side chain. The polymer was insoluble inmethanol, DMSO, acetone and acid.

A similar polymer was prepared by treatment of 4-chloro-methylstyrene with DETA to form N-(4-vinylbenzyl)DETA. The1 HNMR spectrum showed that the central or secondary amino nitrogenhad undergone reaction, for the product seemed to h . symmetrical.

The polymerization of monomer with DETA side chainwas not easily accomplished. Radical polymerization using aperoxide catalyst was not successful, since the amine of theside chain acted as a chain stopper. Polymerization with hydro-chloric acid was slow and gave low molecular weight polymer.

The N-(4-vinylbenzyl)DETA was successfully polymerizedwith 2,2'-azobisl2-methylpropionitrilel as initiator. The polymerwas less rigid than that formed from chloromethylstyrene, indicat-ing a lower molecular weight. Otherwise the polymers were verysimilar.

The treatment of either of the polymers having DETAside chains with Methyl Cellosolve caused the polymer to swelland absorb about 3.5 times its weight of solvent. A 20 g sampleof polymer treated with a saturated solution of sodium hydroxidein Methyl Cellosolve (14.9%) gave 72 g of swollen polymer contain-ing about 1.1% cf sodium hydroxide, based on sodium ion analysis.

A series of solid mixtuies of DETA polymer MethylCellosolve and sodium (potassium) hydroxide were reacted withCEES in the kinetic 4 minute test. The results showed morethan 80% recovrwred simulant. With greater amounts of solid,CEES decompositi n was only slightly increased.

This lack of reactivity probably reflects protectionof the reactive amino side chain by the polymer chain. Theswelling of the polymer by the adsorbed Methyl Cellosolve shouldprovide a mechanism for the simulant and hydroxide to interact;however, because of the hiqh concentration of Methyl Cellosolve,the mixture may not function as does DS-2. We will attemptto remove some of the Methyl Cellosolve so that the relativeamounts of DETA side chain, hydroxide ion, and Methyl Cellosolvemore nearly resemble those in DS-2.

31

IV. FUTURE WORK

The evaluation of the modified clays using simulants ofnerve agents and mustards has been successful using the washertest and the 4 minute kinetic test. Similar investigationsusing the corresponding chemical surety agents would indicatewhether the simulants are satisfactory an. would help to determinethe validity of the test procedures.

The decomposition of simulants by the modified clays studiedappears to have a limit, which may suggest that the reactionis not catalytic. Improvement of the modified clays shouldbe attempted by intercalation of metal ions to produce catalyticdecomposition. New modifications should be evaluated when theyare prepared.

Other variations producing expansion of the clay crystallattice should also improve the reactivity of these clays.These modifications should also be studied in any future work.

It should be possible to duplicate the unique propertiesof DS-2 in a polymer system. The preliminary studies were notsuccessful because too much Methyl Cellosolve was absorbed bythe polymer. Future studies should attempt to match the propor-tions of DS-2 solution. This polymer, if shown to have highdecontamination potentials, should be prepared in the form ofnonwoven fabric, foam, and powder.

62

V. REFERENCES

1. Thomas B. Stanford, "Sorbant Powders for Chemical Warfare(CW) Decontamination," Battelle Memorial Institute, FinalReport, Contract No. DAAH01-81C-A277, Wright PattersonAFB, Ohio, April 1982, ADB064597.

2. R.E. Lyle, H.F. Hamil, E.P. McGovern and H.J. Schattenberg,"Decontamination and Detection by Grafted Polymer Filmsand Powdered Clays," ADPA Proceeding of Technology Applica-tions in Chemical Operations, Brooks Air Force Base, SanAntonio, TX, October 3-5, 1984, pp. 52-55.

3. H.F. Hamil, L.M. Adams, W.W. Harlowe and E.C. Martin, "Irrad-iation-Grafted Polymeric Films, I. Preparation and Proper-ties of Acrylic Acid-Grafted Polyethylene Films," J. PolymerSci. A-l, 9, 363 (1971).

4. G.W. Snedecor and W.G. Cochran, Statistical Methods 6thEdition, Iowa State Univesity Press, Ames, IA, 1967, pp.135-171.

5. R.T. Conley, Infrared Spectroscopy, 2nd Edition, Allynand Bacon, Boston, MA, 1972, pp. 198-200.

83

DISTRIBUTION LIST

1 Copy: CommanderU.S. Army Medical Research and Development CommandATTN: SGRD-EM1-5Fort DetrickFrederick, MD 21701-5012

5 Copies: CommanderU.S. Army Medical Research and Development CommandATTN:' SGRD-PLEFort DetrickFrederick, MD 21701-5012

*12 Copies: AdministratorDefense Technical Information CenterATTN: DTIC-DDACameron StationAlexandria, VA 22314

1 Copy: CommandantAcademy of Health Sciez.ces, U.S. ArmyATTN: AHS-CDMFort Sam Houston, TX 78234

1 Copy: Dean, School of MedicineUniformed Services University

of the Health Sciences4301 Jones Bridge RoadBethesda, MD 20014

84

Documents

TI. · 9E. LIST OF FIGURES ... Page Figure 1. The Evaluation of Irwin Clays in Adsorbing DECP Using the Headspace Method (Section II.A) 17 Figure 2. The Evaluation of Bentonite Clays