DTIC ADA538061: Simulated Rainfall-Driven Dissolution of TNT, Tritonal, Comp B and Octol Particles

0.41 MB·English

by  Defense Technical Information Center

Chemosphere75(2009)1074–1081
ContentslistsavailableatScienceDirect
Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere
Simulated rainfall-driven dissolution of TNT, Tritonal, Comp B and Octol particles
Susan Taylora,*, James H. Levera, Jennifer Faddena, Nancy Perrona, Bonnie Packerb
aColdRegionsResearchandEngineeringLaboratory,72LymeRoad,Hanover,NH03755-1290,UnitedStates
bUSArmyEnvironmentalCommand,HoadleyRoad,AberdeenProvingGround,MD21010,UnitedStates
a r t i c l e i n f o a b s t r a c t
Articlehistory: Live-firemilitarytrainingcandepositmillimeter-sizedparticlesofhighexplosives(HE)onsurfacesoils
Received7October2008 whenroundsdonotexplodeasintended.Rainfall-drivendissolutionoftheparticlesthenbeginsapro-
Receivedinrevisedform8January2009 cesswherebyaqueousHEsolutionscanenterthesoilandgroundwaterascontaminants.Wedripped
Accepted9January2009
waterontoindividualparticlesofTNT,Tritonal,CompBandOctoltosimulatehowsurface-deposited
Availableonline11February2009
HEparticlesmightdissolveundertheactionofrainfallandtousethedatatoverifyamodelthatpredicts
HEdissolutionasafunctionofparticlesize,particlecompositionandrainfallrate.Particlemassesranged
Keywords: from1.1to17mganddripratescorrespondedtonominalrainfallratesof6and12mmh(cid:2)1.FortheTNT
Dissolutionmodel
andTritonalparticles,TNTsolubilitygoverneddissolutiontimescales,whereasthelower-solubilityof
Highexplosives
RDXcontrolledthedissolutiontimeofbothRDXandTNTinCompB.Thelarge,low-solubilitycrystals
Laboratorydriptests
TNT ofHMXslowedbutdidnotcontrolthedissolutionofTNTinOctol.Predictionsfromadrop-impingement
CompB dissolutionmodelagreewellwithdissolved-masstimeseriesforTNT,TritonalandCompB,providing
someconfidencethatthemodelwillalsoworkwellwhenappliedtotherainfall-driven,outdoordisso-
lutionoftheseHEparticles.
PublishedbyElsevierLtd.
1.Introduction ofparticlesize,compositionandrainfallrateareneededtopredict
HEaqueousinfluxtosurfacesoilsattrainingranges.
Militaryrangesprovidesoldierstheopportunitytotrainusinga Lynch et al. (2002a,b) and Phelan et al. (2003) conducted,
varietyofmunitions.However,live-firetrainingcandeposithigh respectively, stirred-bath and glass-bead-column experiments to
explosives(HE)suchas2,4,6-trinitrotoluene(TNT),1,3,5-hexahy- measure the aqueous dissolution rates of collections of HE parti-
dro-1,3,5-trinitrotriazine (RDX) and octahydro-1,3,5,7-tetranitro- cles. We followed the method described in Lever et al. (2005)
1,3,5,7-tetrazocine (HMX) on the range. Detonations scatter HE and dripped water on individual, field-collected, mm-sized parti-
particlesbroadlyoversurfacesoils.High-orderdetonationsscatter cles of four common high-explosive formulations: TNT, Tritonal
lm-sizeHEparticlesandlow-order(LO)detonationsscattermm- (80:20mixofTNTandaluminumflakes),CompositionB(59:39:1
to-cm-sizedHEparticles(Tayloretal.,2006).Giventhelowdrink- mix of RDX:TNT:wax) and Octol (70:30 mix of HMX:TNT). These
ing water screening levels for TNT (2.2lgL(cid:2)1) and RDX areallmeltcastexplosiveswherealuminum,RDXandHMXcrys-
(0.6lgL(cid:2)1) (EPA, 2008) gram-to-kilogram quantities of these talsareaddedtomoltenTNT.Theexperimentssimulatedtherain-
explosivesinaqueoussolutionmaycontaminateanaquifer,asoc- fall-drivendissolutionofHEparticlesscatteredonsurfacesoils.We
curred at the Massachusetts Military Reservation (Clausen et al., then used the resulting dissolved-mass timeseries to validate a
2006).Thelowdrinkingwaterlevelsreflecttheirtoxicitytohuman drop-impingementdissolutionmodeldevelopedforCompB(Lever
health(ATSDR,1995). etal.,2005).ByexpandingthetypesofHEformulationstestedand
RangecharacterizationstudiessummarizedbyPenningtonetal. using two simulated rainfall rates, our intention was to improve
(2006)foundHEinsurfacesoilsatall27militaryinstallationssam- confidenceinthemodel’spredictionsofrainfall-driven,dissolution
pledandHEwasfoundingroundwateratFortLewisWA(Jenkins ofHEparticlesontrainingranges.
et al., 2001), and at three Canadian installations (Martel et al.,
1999;Penningtonetal.,2006).DissolutionofHEparticlesbypre-
cipitationhastooccurbeforeaqueoustransportorbio-degradation 2.Experimentalmethods
oftheHE.ThereforedissolutionratesofHEparticlesasafunction
2.1.HEParticlesamples
TNTisasingle-componentexplosive,asisTritonalalbeitwith
* Correspondingauthor.Tel.:[phone];fax:[phone].
20%aluminumflakes,whileCompBandOctolaretwo-component
E-mailaddress:[email protected](S.Taylor).
0045-6535/$-seefrontmatterPublishedbyElsevierLtd.
doi:10.1016/j.chemosphere.2009.01.031
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
FEB 2009 2. REPORT TYPE 00-00-2009 to 00-00-2009
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Simulated Rainfall-Driven Dissolution Of TNT, Tritonal, Comp B And
5b. GRANT NUMBER
Octol Particles
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
US Army Environmental Command,Hoadley Road,,Aberdeen Proving REPORT NUMBER
Ground,MD,21010
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
Chemosphere 75 (2009) 1074-1081
14. ABSTRACT
Live-fire military training can deposit millimeter-sized particles of high explosives (HE) on surface soils
when rounds do not explode as intended. Rainfall-driven dissolution of the particles then begins a process
whereby aqueous HE solutions can enter the soil and groundwater as contaminants. We dripped water
onto individual particles of TNT, Tritonal, Comp B and Octol to simulate how surface-deposited HE
particles might dissolve under the action of rainfall and to use the data to verify a model that predicts HE
dissolution as a function of particle size, particle composition and rainfall rate. Particle masses ranged
from 1.1 to 17 mg and drip rates corresponded to nominal rainfall rates of 6 and 12 mm h1. For the TNT
and Tritonal particles, TNT solubility governed dissolution time scales, whereas the lower-solubility of
RDX controlled the dissolution time of both RDX and TNT in Comp B. The large, low-solubility crystals of
HMX slowed but did not control the dissolution of TNT in Octol. Predictions from a drop-impingement
dissolution model agree well with dissolved-mass timeseries for TNT, Tritonal and Comp B, providing
some confidence that the model will also work well when applied to the rainfall-driven, outdoor dissolution
of these HE particles.
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 8
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
S.Tayloretal./Chemosphere75(2009)1074–1081 1075
explosives. We tested two particles for each HE formulation, all 2.3.Analyticalmethods
collected from training ranges. The TNT samples were collected
after a LO detonation of a 155-mm round, the Comp B samples Explosive concentrations were determined following SW-846
werefromtheLOdetonationofan81-mmround,andtheTritonal Method8330B(EPA, 2006). One mL ofthe effluentwas added to
sampleswerefromaLOdetonationofa230-kg(500-lb)bomb.We 2mLofde-ionizedwaterandto1mLofacetonitrileandthenfil-
obtainedasingleOctolparticlefromsoilssampledonananti-tank teredthrougha0.45lmMilliporesyringefilter.HighPerformance
firingrangeandsplitittoformthetwosamplestestedhere. Liquid Chromatography separated TNT, RDX, HMX, and their co-
Beforestartingthedripexperiments, weindividuallyweighed contaminantsandbreakdownproducts.WeusedaWatersNovaP-
thedryHEparticlesonaMettlerToledoMX5microbalance(uncer- akC8column(3.9mmby150mm)elutedat1.4mLmin(cid:2)1(28(cid:2)C)
tainty±0.01mg)andphotographedthemusingacameraattached with85:15water:isopropanolmixanddetectedbyUVat254nm.
toabinocularmicroscope.Partwaythroughthetests,were-pho- A commercially available standard (Restek) was used for calibra-
tographedtheparticlestodocumentchangestotheirappearance. tion.Weprepared1-ppmand10-ppmdilutionsofthesestandards.
Lever et al. (2005) showed that mm-sized Comp B detonation The1ppmstandardswereruneverytensamplesandblankswere
residues varied substantially in their RDX/TNT ratios, averaging run before and after each standard run. The 10ppm standards
1.74±0.28(±1standarddeviation)comparedwithanominalcom- wereinterspersedwiththesamplesasinternalchecks,withblanks
positionof60/39=1.54.Significantdeviationsincompositionare aftereachtopreventcarryover.Basedontheconcentrationsofthe
likelyforfield-recoveredHEparticleswithcomponentcrystalsthat standardsandthe precisionoftheeffluentvolumes,weestimate
arelargecomparedwithoverallparticlesize.Tomeasurethisvar- that the cumulative dissolved HE masses have uncertainties of
iabilityforTritonal,weweighed21mm-sizedparticlesonaSarto- about±1%.
rius electronic balance (uncertainty±0.1mg), dissolved them in
acetonitrileandanalyzedtheirTNTcompositions.Wedidnothave
3.Results
sufficient Octol to investigate variability in Octol particle
compositions.
3.1.Hemassrecovery
2.2.Laboratorysetup Weanalyzed21Tritonalparticlesforcompositionalvariability.
ThepercentageofTNTbyweightrangedfrom52%to84%ofparti-
Two Cole Palmer multi-syringe pumps allowed us to drip de- cle mass and averaged 77.5%±6.6%. Thus, the average of the 21
ionized water simultaneously onto eight separate particles, two particlesfellwellwithin1standarddeviationofthenominalcom-
particlesforeachofthefourHEformulationsattwodifferentdrip positionofTritonal(80%TNT).
rates.AneedleoneachsyringedrippedwaterontoanHEparticle Table1liststhedissolutiontestparametersandtherecovered
resting on a 10-mm-diameter glass frit at the base of a Buchner massfortheeightdriptestparticles.Thenominaldriprateswere
funnelthatdrainedintoa20-or40-mLscintillationvial.Thevials 0.5mLh(cid:2)1and1.0mLh(cid:2)1forthe‘‘1”and‘‘2”particledesignations,
werereplaced daily and the volume and HEconcentrationof the respectively. We recovered essentially 100% of the mass for the
effluent analyzed. These data provided dissolved HE mass as a two TNT particles and for Comp B 2. The recovered HE mass for
function oftime for each ofthe eight particles. At the end of the CompB1(96.5%)wasclosetotheaverage(96%±1%)foundbyLe-
tests,weextractedeachfritandfunnelwithacetonitrileandana- veretal.(2005)for30individualCompBparticlesobtainedfrom
lyzedthesolutionsforHEmass. the same low-order detonation. The two particles dissolved here
The programmed flow rates for each pump were 0.5 and hadRDX/TNTratiosof1.69and1.87,similartotheaverageratio
1.0mLh(cid:2)1. We measured the actual flow rates and number of obtained by Lever et al. (2005) of 1.74±0.28. We measured less
drops per hour for each syringe. These averaged 0.47± than1%HMX,animpurityinRDX,intheseCompBparticles.
0.06mLh(cid:2)1 with 26 drops h(cid:2)1 and 0.95±0.09mLh(cid:2)1 with 56 TherecoveredTNTmasswas67.8%and78.5%oftheinitialmass
dropsh(cid:2)1atthetwo,programmedrates.Forreference,theseflow for Tritonal 1 and Tritonal 2, respectively. This included TNT ex-
ratescorrespondtorainfallratesof5.9and12.0mmh(cid:2)1basedon tractedfromthefrits.Bothparticlesfellwithin2rofthenominal
the cross-sectional area of each glass frit. The average volume of Tritonalcomposition.Aluminumgrainsconstitutedtheremaining
the arriving drops was larger than that expected for the natural massfortheTritonalparticles.
rainfall at the same rates (Pruppacher and Klett, 1997) and each ForOctol1andOctol2,werecovered1.86mgand3.61mgof
arrivingdropwettedtheparticleonthefrit.Thede-ionizedwater TNT, representing 28.4% and 20.8% of the initial particle masses,
used for the tests was at room temperature and the latter was respectively.Octolnominallycontains30%TNT,butwewouldex-
fairly constant throughout the tests at 22±1(cid:2)C. The tests were pect millimeter-sized particles to vary significantly from this va-
normally halted over the weekends and the particles allowed to lue owing to the large ((cid:3)1mm) crystals of HMX present in the
dry. TNT matrix.Bothdissolutiontestscontinuedfor manydays with
Table1
DissolutiontestparametersandtherecoveredmassforHEparticles.
Particle Driprate(mLh(cid:2)1) Initialmass(mg) Dissolved-mass(mg) Time(days) %Massrecovered
TNT1 0.5 5.34 5.33 201 99.9
TNT2 1.0 9.59 9.70 98 101
Tritonal1 0.5 1.89 1.28 72 67.8*
Tritonal2 1.0 6.40 5.02 73 78.5*
CompB1 0.5 2.31 2.23 200 96.5
CompB2 1.0 9.09 9.03 141 99.4
Octol1 0.5 6.53 2.15 101 33.9*
Octol2 1.0 17.33 9.92 140 57.2*
* AlltheTNTmass.
1076 S.Tayloretal./Chemosphere75(2009)1074–1081
tests recovered all the TNT initially present in the two Octol
particles.
3.2.Appearanceoftestparticles
Fig.1showsphotographsofthe8particlestakenatthebegin-
ningandpartwaythroughthedissolutiontests,at30daysforthe
0.5mLh(cid:2)1 tests and at 21days for the 1.0mLh(cid:2)1 tests. The TNT
particlesbecamesmootherandsmallerbutretainedtheiroriginal
shapes.TheTritonalparticlesbecamesmallerandslightlybumpier
asTNTdissolvedexposingthealuminumgrains.TheCompBpar-
ticles,onthehand,becamenoticeablybumpierand‘sugary’look-
ing as dissolution of the surface TNT revealed the larger
((cid:3) 0.1mm), slower-dissolving RDX crystals. The Octol particles
disaggregatedcompletelywhentheirTNTmatrixdissolvedleaving
large((cid:3)1mm)HMXcrystals(seeFig.1,Octol2at21d).
3.3.Dissolutionmodelingandcomparisonwithdissolved-mass
timeseries
Lever et al. (2005) developed and validated a drop-impinge-
mentmodelfortherainfall-drivendissolutionofmm-sizedComp
Bparticles.Itpertainstothepracticalcasewherespatiallyisolated
HEparticlesresideonwell-drainingsurfacesoilsandthusareex-
posed to direct impingement by raindrops. The present experi-
ments simulated these circumstances. With rare exceptions,
water quickly drained through the glass frits so that impinging
dropsinteracteddirectlywiththeHEparticles.Thepresentexper-
imentsofferadditionaldatatovalidatethismodelforotherHEfor-
mulationsandfortwosimulatedrainfallrates.
The drop-impingement dissolution model assumes that rain-
dropshitandwetanHEparticle.Betweenraindrops,theparticle
holds a stagnant water-layer against its surface, which saturates,
via diffusion, with HE. Arrival of the next raindrop washes away
thedissolvedHEandrefreshesthestagnantlayer.Thedissolution
rate,m (gs(cid:2)1)ofHEspeciesj,averagedoveradrop-arrivalinterval,
j
t (s),isthus
d
SV
m ¼ j l ð1Þ
j t
d
whereS isthesolubilityofspeciesjinwater(gcm(cid:2)3)andV isthe
j l
water-layervolume(cm3).Forasphericalparticle,awater-layerof
thickness h (cm) will effectively saturate (average concentra-
tion>0.9S)provided
j
h2
t > (cid:4)t ð2Þ
d D s
j
whereD isthediffusioncoefficientofspeciesjintowater(cm2s(cid:2)1)
j
andt definesthelayersaturationtime.Additionally,forEq.(1)to
s
apply,thevolumeofanarrivingwaterdrop,V ,mustbelargerthan
d
thevolumeofthestagnantwater-layer,whichgenerallylimitsits
applicabilitytomm-sizedandsmallerparticles.
EachHEspeciesinaparticlecoulddissolveindependently,fol-
lowingEq.(1),ifitisincontactwiththewater-layerandhasasuf-
ficientlyhighdiffusioncoefficienttosatisfyEq.(2).However,Comp
B and Octol particles consist of low-solubility crystals (RDX and
HMX, respectively) embedded in higher-solubility TNT matrixes.
Initially,TNTonthesurfaceoftheseparticlescandissolveindepen-
Fig.1. Photosandinitialmassesofmm-sizedparticlesusedinthelaboratorytests dentlyintothewater-layer,leadingtohighinitialTNTdissolution
pairedwithphotosofthesameparticletakenpartwaythroughthetests. rates.However,thelower-solubilitycrystalseventuallyrestrictthe
pathways for TNT molecules to diffuse into the water-layer. TNT
nomeasurableTNTinthewatersamples,andwefoundnoaddi- diffusionintothelayerthenbecomesanalogoustomoleculardif-
tionalTNTinaseriesofacetonitrileextractionsofthefunnelsand fusion through a porous medium, where porosity and tortuosity
frits at the end of the tests. Thus, we think that the dissolution combine to reduce significantly the effective diffusion coefficient
S.Tayloretal./Chemosphere75(2009)1074–1081 1077
Table2
Modelparameters.ParticledensitiesarefromArmyMaterielCommand(1971).AqueoussolubilitiesarefromLynchetal.(2001)at22(cid:2)C,anddiffusioncoefficientsarefromLynch
etal.(2002b)at25(cid:2)C;listedareTNTvaluesforTNTandTritonal,RDXvaluesforCompB,andHMXvaluesforOctol.ThetwodensitiesforTritonalreflectthedifferentTNT/
aluminumratiosofthetwoparticles.
Particle q(gcm(cid:2)3) Sj(gcm(cid:2)3) Dj(cm2s(cid:2)1) td(s) h(mm) ts(s) Vd/Vl
TNT1 1.65 1.17E(cid:2)04 6.71E(cid:2)06 138 0.075 8.4 21
TNT2 1.65 1.17E(cid:2)04 6.71E(cid:2)06 64 0.095 13 10
Tritonal1 1.89 1.17E(cid:2)04 6.71E(cid:2)06 138 0.090 12 35
Tritonal2 1.80 1.17E(cid:2)04 6.71E(cid:2)06 64 0.082 10 17
CompB1 1.65 4.02E(cid:2)05 2.20E(cid:2)06 138 0.092 38 28
CompB2 1.65 4.02E(cid:2)05 2.20E(cid:2)06 64 0.110 55 9.1
Octol1 1.80 3.64E(cid:2)06 1.50E(cid:2)04 138 0.12 1.0 12
Octol2 1.80 3.64E(cid:2)06 1.50E(cid:2)04 64 0.19 2.4 3.5
6.0
5.0 TNT 1
g)
m 4.0
s (
s
a
M 3.0
d
e
v
ol
s 2.0
s
Di
1.0
0.0
0 50 100 150 200 250
Time (d)
12.0
10.0 TNT 2
g)
m 8.0
s (
s
a
M 6.0
d
e
v
ol
s 4.0
s
Di
2.0
0.0
0 20 40 60 80 100 120
Time (d)
Fig.2. Dissolved-massversustimefortheTNTtestparticles(symbolsaremeasuredvalues,smoothcurvesaremodeledvalues).
q
(Bear,1972;ChambreandPigford,1984).Consequently,TNTwill m ¼m TNT ð3Þ
TNT RDXq
notsaturatethewater-layerbetweendropsanditseffectivedisso- RDX
lutionratewilldecrease.
whereq andq arethemassdensitiesofeachspeciesinthe
Lever et al. (2005) observed this effect during laboratory drip TNT RDX
originalCompBparticle.Thisapproachensuredthatthetimescale
tests of Comp B particles. They implemented the limiting case
for RDX dissolution controlled the predicted TNT dissolution rate.
where RDX dissolution controls the dissolution of TNT by using
We applied this approach for both Comp B and Octol particles,
Eq. (1) for RDX and assuming that TNT dissolves at a rate that
substitutingq forq inEq.(3)forOctol.
maintainsthebulkRDX/TNTmassratio: HMX RDX
1078 S.Tayloretal./Chemosphere75(2009)1074–1081
We implemented the drop-impingement model for the eight Fig.4showsthemeasuredandpredicteddissolved-massesfor
test particles by treating each particle as a sphere of equivalent the two Comp B particles. The predicted RDX dissolution agrees
mass,sothatthewater-layervolumeissimply quite well with the measured results for both particles. In both
tests,TNTinitiallydissolvedquicklybuteventuallywascontrolled
4
Vl¼3p½ðaþhÞ3(cid:2)a3(cid:5) ð4Þ bytheslowerdissolutionofRDX.Themodelpredictsthisbehavior
reasonablywellusingthelimitingcaseofRDXcontrolofTNTdis-
whereaissphereradius(cm).Becausewecouldnotmeasurethe solution(Eq.(3)).Bycomparison,itpredictsmuchfasterdissolu-
water-layerthicknesses,wedeterminedhbyfittingthepredictions tion under the assumption that TNT dissolves independently of
tothemeasureddissolved-masstimeseriesfortheslower-dissolv- RDX.Interestingly,thefastestdissolutionratesobservednearthe
ingspeciesineachparticle.Effectsofnon-sphericalparticleshape beginningofthetestsdoapproximatethoseforindependentTNT
are thereby folded into h. Table 2 summarizes the parameters dissolution. The transition from independent TNT dissolution to
neededtorunthedrop-impingementmodelandtocheckitsvalid- RDX-controlleddissolutionprobablyreflectsrestrictionofTNTdif-
ity(td>ts,Vd>Vl). fusionintothewater-layerbytheporousmediumofRDXcrystals.
Figs. 2 and 3 show the measured and predicted dissolved- The fitted water-layer thicknesses average 0.101±0.013mm
massesfortheTNTandTritonalparticles,respectively.Thesepar- for the two Comp B particles (Table 2). This compares well with
ticlescontainonlyTNT,soEq.(1)appliesdirectly.Themodelpre- theaveragevalue(0.13±0.05mm)obtainedforthefourCompB
dicts the dissolution of all four particles quite well. The average particlestestedbyLeveretal.(2005).ForbothCompBparticles,
layer thicknesses for the TNT and Tritonal particles were essen- thedrop volumeswerelargerthanthe water-layervolumes,and
tiallythesameat0.085±0.014mmand0.086±0.006mm,respec- thedrop-arrivaltimeswerelongerthantheRDXsaturationtimes
tively. Note that for the particles, the drop-arrival times were (Table2).TNTsaturationtimesmustincreasetoaccountforeven-
longer than the water-layer saturation times, and the drop vol- tualRDXcontrolofTNTdissolution.
umes were larger than the water-layer volumes (Table 2) as re- Fig.5showsthemeasuredandpredicteddissolved-massesfor
quiredbyEq.(1). thetwoOctolparticles.AswithCompB,measuredTNTdissolution
1.4
1.2 Tritonal 1
g) 1.0
m
s (
s 0.8
a
M
d
ve 0.6
ol
s
s
Di 0.4
0.2
0.0
0 20 40 60 80 100
Time (d)
6.0
5.0 Tritonal 2
g)
m 4.0
s (
s
a
M
3.0
d
e
v
ol
s 2.0
s
Di
1.0
0.0
0 20 40 60 80 100
Time (d)
Fig.3. Dissolved-massversustimefortheTritonaltestparticles(symbolsaremeasuredvalues,smoothcurvesaremodeledvalues).
S.Tayloretal./Chemosphere75(2009)1074–1081 1079
1.6
RDX
1.4 CompB 1
1.2
g)
m
s ( 1.0 Independent
as TNT TNT
M
0.8
d
e
v
ol 0.6
s
s
Di
0.4
Controlled TNT
0.2
0.0
0 50 100 150 200 250
Time (d)
7.0
RDX
6.0 CompB 2
g) 5.0
m
s (
s 4.0
a
M Independent
d TNT TNT
e 3.0
v
ol
s
s
Di 2.0
Controlled TNT
1.0
0.0
0 20 40 60 80 100 120 140 160
Time (d)
Fig.4. Dissolved-massversustimefortheCompBtestparticles(symbolsaremeasuredvalues,smoothcurvesaremodeledvalues).DissolutionofRDXeventuallycontrols
dissolutionofTNT(modeledviaEq.(3)).IndependentTNTdissolutionwouldoccurmuchfaster.
initially proceeded quickly, with maximum dissolution rates andRDXmassrecoveriesfromallparticleswereverygood;HMX
approachingthepredictedratesforindependentTNTdissolution. massrecoverywascomplicatedbyitslow-solubilityanddisaggre-
UnlikeCompB,however,thelargerHMXcrystalsdidnoteventu- gationoftheOctolparticlesfollowingcompleteTNTdissolution.
ally controlTNT dissolution,and alltheTNT dissolved fromboth Changesintheappearanceoftheparticlesprovidedinsightinto
particles, leaving HMX crystals. We terminated both tests after their dissolution mechanics. The single HE-species particles, TNT
many days with no TNT in the water samples. The fitted water- and Tritonal, essentially retained their original morphologies as
layer thicknesses averaged 0.155±0.049mm, much larger than they dissolved, and their dissolution timeseries are fairly simple
thosefortheotherparticles(Table2).Nevertheless,thecalculated curves.TheconcentrationofaluminumgrainsinTritonalisappar-
saturation times (Eq. (2)) based on either TNT or HMX diffusion entlytoolowtodisruptdissolutionoftheTNT.Bycomparison,the
werelessthanthedrop-arrivaltimes,andthewater-filmvolumes CompBandOctolparticlesquicklybecamelumpyassurfaceTNT
werelessthanthedropvolumes,asrequiredbythemodel. dissolved to reveal underlying RDX and HMX crystals. Subse-
quently,theCompBparticlesbecamesmallerconglomerationsof
4.Discussion RDXcrystalsshieldinginternalTNT,whiletheOctolparticleslost
alltheirTNTanddisaggregatedintocollectionsofHMXcrystals.
The laboratory tests dripped water on individual HE particles We do not understand the reasons for discontinuities in the
collected from field detonations. They more closely simulate the measureddissolutionratesforsomeparticles(e.g.,TNT1,Tritonal
physicalcircumstancesofHEparticlesexposedtorainfallonsur- 2andCompB1).Thediscontinuitiesdonotcoincidewiththedates
facesoilsthandostirred-bath(Lynchetal.,2002a,b)orglass-bead ofparticleremovalforphotographingorobservedsplittingofthe
columnexperiments(Phelanetal.,2003).Theyyieldestimatesof particles, the good mass balances confirm that the particles did
aqueousHEinfluxtosurfacesoilsasfunctionsofparticlesize,par- notlosepiecesduringhandling,andseveralparticlesdidnotdis-
ticlecompositionandrainfallratewhileavoidingthecomplexityof playdiscontinuitiesbeyondrandomfluctuations.
aqueous-phaseHE-soilinteractionsthatoccurduringsoil-column Thedrop-impingementdissolutionmodelpredictsverywellthe
experiments (Pennington et al., 2006; Morley et al., 2006). TNT dissolved-mass timeseries of the TNT and Tritonal particles. The
1080 S.Tayloretal./Chemosphere75(2009)1074–1081
2.0
Independent
1.8
TNT TNT
1.6
g) 1.4
m
s ( 1.2
s
a Octol 1
M 1.0
d
e
v 0.8
ol
s
s 0.6
Di
0.4 HMX
0.2
Controlled TNT
0.0
0 20 40 60 80 100 120
Time (d)
4.0
Independent
3.5 TNT
TNT
3.0
g)
m
s ( 2.5
s
a
M 2.0 HMX
d
ve Octol 2
ol 1.5
s
s
Di 1.0
0.5
Controlled TNT
0.0
0 20 40 60 80 100 120 140 160
Time (d)
Fig.5. Dissolved-massversustimefortheOctoltestparticles(symbolsaremeasuredvalues,smoothcurvesaremodeledvalues).TNTdissolutionrateapparentlywasneither
controlledby,norindependentof,HMXdissolution.
singletuningparameter,water-layerthickness,h,wasreasonably Thedrop-impingementmodelaccountsfordifferencesinrain-
consistent among the four particles, averaging 0.086±0.009mm. fall rates via differences in average drop-arrival time, t , in Eq.
d
The model also predicts well the overall dissolution behavior of (1).Thegenerallygoodagreementbetweentheobservedandpre-
theCompBparticlesandprovidesinsightintothegoverningpro- dicteddissolutiondataattwosimulatedrainfallratessuggeststhat
cesses.InitiallyTNTdissolvesindependentlyofRDX,butincreasing t doesindeedscaledissolutionrate.
d
restrictionofTNTdiffusionbytheRDXcrystalseventuallycauses We have initiated a series of outdoor dissolution tests where
TNT dissolution to be controlled by RDX dissolution. With addi- cm-sizedchunksofHEareexposedtonaturalweatherconditions.
tionalwork,wehopetoquantifythiseffectbymodelingTNTdiffu- ThesetestsshouldcloselysimulatethedissolutionofisolatedHE
sionwithinachangingporousRDXmedium. pieces on range soils and provide a more realistic validation of
The model does not predict well the TNT dissolved-mass thedrop-impingementmodel.Wearedevelopinga‘‘large-particle”
timeseries for the Octol particles. The large, low-solubility HMX version of the model to circumvent the restriction that raindrop
crystals appeared to impede but not control TNT dissolution. volumebegreaterthanwater-filmvolume.Nevertheless,themod-
That is, neither limiting model case (independent or HMX-con- el predicts quite well the dissolution of millimeter-size TNT, Tri-
trolled dissolution) fits the observed TNT data. The reduced tonal and Comp B particles. This provides confidence it will also
influence of HMX on TNT dissolution, compared with the con- workwellwhenappliedtotherainfall-driven,outdoordissolution
trolling role of RDX in Comp B, is consistent with less constraint oftheseHEmaterials.
ofTNTdiffusionintothesurroundingwater-layer.Thehigherfit-
tedvaluesoflayerthickness(0.155±0.049mm)couldreflectthe Acknowledgements
creation of surface area as the Octol particles disaggregated fol-
lowing TNT loss. These values give reasonable agreement to the We thank Dr. Jeffrey Marqusee and Dr. Andrea Leeson of the
limitedHMXdissolutiondata,butmoreworkisneededtoquan- StrategicEnvironmentalResearchandDevelopmentProgram(SER-
tify the impeding effect of HMX crystals on TNT dissolution in DP)forfundingourproject(ER-1482)andCareyLiandPamelaCol-
Octol. linsforhelprunningthelaboratorytests.
S.Tayloretal./Chemosphere75(2009)1074–1081 1081
References Lynch, J.C., Brannon, J.M., Delfino, J.J., 2002a. Effect of component interactions
on the aqueous solubilities and dissolution rates of the explosive
formulations Octol, Composition B and LX-14. J. Chem. Eng. Data 47,
ArmyMaterielCommand,1971.EngineeringDesignHandbook:ExplosivesSeries,
542–549.
Properties of Explosives of Military Interest. Army Materiel Command,
Lynch, J.C., Brannon, J.M., Delfino, J.J., 2002b. Dissolution rates of three
Alexander,VA.
high explosive compounds: TNT, RDX and HMX. Chemosphere 47, 725–
ATSDR, 1995. Toxicological profile for 2,4,6-trinitrotoluene and RDX. US
734.
DepartmentofHealthandHumanServices,PublicHealthService,Agencyfor
Martel R., Mailloux, M., Lefebvre, R., Michaud, Y., Parent, M., Ampleman, G.,
ToxicSubstancesandDiseaseRegistry,Atlanta,GA.
Thiboutot,S.,JeanS.,RoyN.,1999.Energeticmaterialsbehavioringroundwater
Bear,J.,1972.DynamicsofFluidsinPorousMedia.DoverPublications,NewYork.
at the Arnhem Anti-Tank Range, CFB Valcartier, Quebec, Canada. INRS-
Chambre,P.L.,Pigford,T.H.,1984.Predictionofwasteperformanceinageological
GeoresourcesReport1999-02.
repository.Mat.Res.Soc.Symp.Proc.,ElsevierSciencePublishingCo.,vol.26,
Morley,M.C.,Yamamoto,H.,Speitel,G.E.,Clausen,J.,2006.Dissolutionkineticsof
985–1008.
highexplosivesparticlesinasaturatedsandysoil.J.Contam.Hydrol.85,141–
Clausen,J.L.,Korte,N.,Dodson,M.,Robb,J.,RievenS.,2006.ConceptualModelfor
158.
theTransportofEnergeticResiduesfromSurfaceSoiltoGroundwaterbyRange
Pennington, J.C., Jenkins, T.F., Ampleman, G., Thiboutot, S., Brannon, J.M.,
Activities.ERDC-CRRELTM-06-18.
Hewitt, A. D., Lewis, J., Brochu, S., Diaz, E., Walsh, M.R., Walsh, M.E.,
EnvironmentalProtectionAgency,2008.Region6HumanHealthMedium-Specific
Taylor, S., Lynch, J.C., Clausen, J., Ranney, T.A., Ramsey, C.A., Hayes, C.A.,
ScreeningLevels.www.epa.gov/Region6/6pd/rcra_c/pd-n/screen.htm.
Grant, C.L., Collins, C.M., Bigl, S.R., Yost, S., Dontsova K., 2006. Distribution
Environmental Protection Agency, 2006. Nitroaromatics, Nitramines and Nitrate
and fate of energetics on DoD test and training ranges: final report. ERDC
EstersbyHPLC,SW-846Method8330B.
TR-06-13.
Jenkins,T.J.,Pennington,J.C.,Ranney,T.A.,BerryJr.,T.E,Miyares,P.H.,Walsh,M.E.,
Phelan J.M., Webb, S.W. Romero, J.V. Barnett, Griffin F., Eliassi M., 2003.
Hewitt, A.D., Perron, N.M., Parker, L.V., Hayes C.A., Wahlgren E.G., 2001.
MeasurementandModelingofEnergeticMaterialMasstransfertoSoilPore
Characterizationofexplosivescontaminationatmilitaryfiringranges,ERDC/
Water-ProjectCP-1227.SandiaReport2003-0153.
CRRELTR-01-5.
Pruppacher,H.R.,Klett,J.D.,1997.MicrophysicsofCloudsandPrecipitation.Kluwer
Lever, J., Taylor, S., Perovich, L., Bjella, K., Packer, B., 2005. Dissolution of
AcademicPublishers,Boston.
CompositionBResiduals.Environ.Sci.Technol.39,8803–8811.
Taylor,S.,Campbell,E.,Perovich,L.,Lever,J.H.,Pennington,J.,2006.Characteristics
Lynch, J.C., Myers, K.F., Brannon, J.M., Delfino, J.J., 2001. Effects of pH and
ofCompositionBParticlesfromBlow-in-PlaceDetonations.Chemosphere65,
temperatureontheaqueoussolubilityanddissolutionrateofTNT,RDXand
1405–1413.
HMX.J.Chem.Eng.Data46,1549–1555.