Destruction or decomposition ofhypergolic chemicals in a liquidpropellant testing laboratory

Hypergols are chemicals that ignite when combined, and are often used in spacecraft propulsionand power systems. The White Sands Test Facility laboratory performs specification analyses,materials compatibility testing, and hazards assessments with a variety of hypergolic spacecraftpropellants. The hypergols are typically highly reactive, toxic and are some of the most severeexamples of incompatible chemicals encountered anywhere. For safety and environmentalreasons, the laboratory has procedures for decomposing laboratory-scale quantities of hypergolsinto less reactive or more benign, well-characterized products. Additionally, some analyticalprotocols require the decomposition of a hypergol as part of a sample preparatory procedure, andthis must also be done in a safe, controlled manner. The purpose of this work is to present a reviewand discussion of the laboratory decomposition of propellant hydrazines (hydrazine, methylhy-drazine, and 1,1-dimethylhydrazine), dimethyl-2-azidoethylamine (DMAZ), dinitrogen tetroxideoxidizers, and concentrated (>70%) hydrogen peroxide into known and/or less hazardousproducts. Hydrazines and common contaminants or partial oxidation products includingN-nitrosodimethylamine and N-nitrodimethylamine are reduced to ammonia and/or aliphaticamines. DMAZ is reduced to N,N-dimethlyethylenediamine. Dinitrogen tetroxide oxidizers arehydrolyzed to nitrate and nitrite solutions. Finally, hydrogen peroxide is decomposed to water.

By Benjamin Greene, Mark B.McClure, and Harry T. Johnson

INTRODUCTION

The NASA Johnson Space Center(JSC) White Sands Test Facility(WSTF) liquid propellant testinglaboratory performs a wide range oftesting and evaluation of hypergolicfluids that are typically used in aero-space.1 These testing and evaluationactivities involve fluid specificationanalyses, materials and fluids compat-ibility testing, thermal and ignitionhazards analyses, stability testing, reac-tion product characterizations, andtoxic vapor monitor and personal pro-tective equipment (PPE) evaluations.

Hypergols are chemicals that spon-taneously ignite when combined, acharacteristic that makes them particu-larly attractive in spacecraft propulsionand power system. A hypergolic mix-ture consists of a fuel and an oxidizer;there is no ignition source requiredto ignite them. Hypergols are some ofthe most dramatically incompatiblechemical combinations that exist. Thefuels routinely tested at WSTF are thehydrazines:

Although hydrazine is typically usedas a monopropellant (only a catalyst isneeded for its energetic decomposi-tion), it is hypergolic with a number ofoxidizers.1,2-Dimethylhydrazine(sym-metrical dimethylhydrazine, SDMH),(CH3)HN–NH(CH3), is not used inaerospace due to its toxicity. The hydra-zines are hypergolic fuels due to theirN–N bond energies and amine func-tional groups.2–4 The hydrazines arealso flammable, corrosive, and toxic.

Incomplete oxidization of methy-lated hydrazines (in particular,UDMH), readily form other hazardousproducts such as N-nitrosodimethyla-mine (NDMA) and N-nitrodimethyla-mine (DMNM) (Equation 1)2,5

ðCH3Þ2N�NH2UDMH

!½O�ðCH3Þ2N�NONDMA

!½O�ðCH3Þ2N�NO2DMNM

(1)

Although NDMA has been a com-pound of great concern because of itstoxicity, it is easily oxidized to DMNM,a compound that is also highly toxicandshould be presumed present whenUDMH or NDMA have been exposedto air. Because of this, decomposition ofthese compounds was also investigated.

The oxidizers that are hypergolicwith the hydrazine fuels are primarilydinitrogen tetroxide (NTO).6 NTO-based oxidizers are hypergolic with

FEATURE

6 � Division of Chemical Health and Safety of the American Chemical Society 1074-9098/$30.00

Elsevier Inc. All rights reserved. doi:10.1016/j.chs.2003.09.017

Hydrazine H2N–NH2

Methylhydrazine (MMH) (CH3)HN–NH2

1,1-Dimethylhydrazine (unsymmetricaldimethylhydrazine, UDMH)

(CH3)2N–NH2

Benjamin Greene and Mark B.McClure are affiliated with HoneywellTechnology Solutions Inc., NASAJohnson Space Center, White SandsTest Facility, P.O. Box 20, Las Cruces,NM 88004 and Harry T. Johnson isaffiliated with NASA Johnson SpaceCenter, White Sands Test Facility, P.O.Box 20, Las Cruces, NM 88004 (Tel.:(505) 525-761; fax: (505) 525-597;e-mail: bgreene@wstf.nasa.gov).

hydrazines due to their strong oxidiz-ing characteristics, and include mixedor modified oxides of nitrogen (MON),which are mixtures of NTO and nitricoxide (NO). Nitric oxide is added toNTO to serve as an oxygen scavengerto prevent stress corrosion cracking oftitanium alloy tankage and piping inaddition to lowering the freezing pointof the fluid.

The oxidizers routinely tested atWSTF are the NTO-based:

Dinitrogentetroxide(NTO)

100% N2O4

MON-3 97% N2O4 þ 3% NOMON-10 90% N2O4 þ 10% NOMON-25 (75% N2O4 þ 25% NO)

is occasionally tested

In addition to the hydrazine-NTOhypergolic combinations, some testingand evaluations have been perfor-medwithdimethyl-2-azidoethylamine(DMAZ; (CH3)2NCH2CH2N3) whichis a relatively new and experimentaltertiary amine azide fuel that has beenproposed as a ‘‘green’’ alternative tosome of the hydrazines owing to itsreduced toxicity. DMAZ is hypergolicwith inhibited red fuming nitric acid(IRFNA), due to its energetic azide andamine functional groups.

Testing and evaluation of high-testhydrogen peroxide (HTP) at concen-trations typically greater than 70% hasalso been performed at WSTF due toan increased interest in its use forrocket propulsion and power.7 Propul-sion and power systems may use HTPas a monopropellant (catalytic decom-position produces high-energy steamand oxygen), or as a bipropellant typi-cally in combination with hydrocar-bon-based fuels.

For safety and environmental rea-sons, it may be desirable to decomposeor destroy the hypergols after testing inorder to form less reactive and/or morebenign products. In other cases, such aswith the NTO oxidizers and HTP, spe-cification analyses require the fluids bedecomposed under controlled condi-tions while maintaining their analyticalintegrity.

Methods used for decomposinghypergols herein are available fromthe literature or developed and vali-

dated in our laboratory.8–12 We reviewour evaluation of the laboratory decom-position or destruction of hydrazinesand some of their oxidation products,DMAZ, NTO oxidizers, and hydrogenperoxide. This evaluation includesidentification of the major decomposi-tionproducts, anassessment of the ratesand completeness of the reactions, anda description of the practical laboratoryprocedures used in our laboratory. Thedevelopment of a novel spill pillow forcontrol of UDMH spills based on thiswork is also described.

We wish to note that Lunn and San-sone earlier described the proceduresfor destruction of the hydrazines andNDMA.8 These authors emphasize theadvantages of reduction rather thanoxidation for UDMH destruction toavoid producing undesirable oxidationproducts. No procedures were found inthe literature for the destruction ofDMAZ, but general procedures fordecomposition of organic azides wereapplied. Procedures for destruction ordecomposition of NTO oxidizers andhydrogen peroxide were taken usinggeneral chemistry principles or proce-dures adapted from military specifica-tions.

EXPERIMENTAL

Safety Note: The following procedureswith these highly hazardous chemicalswere performed by trained and experi-enced personnel in designated labora-tories at the White Sands Test Facility.The required engineering safety con-trols were operational and the appro-priate personal protective equipmentwas worn at all times. These precau-tions are not described here and thefollowing should not be attemptedwithout performing thorough hazardsanalyses.

Reagents and Materials for HydrazineReduction

Hydrazine, MMH, UDMH, and MON-3 conforming to their respective mili-tary specifications, were obtainedlocally.13–15 NDMA (Supelco) wasobtained as a 5,000 mg/L solution inmethanol. DMNM (Picatinny Arsenaland Los Alamos National Laboratory)was obtained as a white, crystallinesolid. All dilutions were performed

using deionized water. Other reagentswere American Chemical Society(ACS) grade unless otherwise specified.

A 1 M sodium hydroxide solutionwas prepared with ACS grade crystal( J.T. Baker) and deionized water. Fiftymilliliters of hydrazine, MMH, UDMH,NDMA, or DMNM solution werecombined with 50 mL of 1 M sodiumhydroxide immediately prior to theaddition of 5 g of 50:50 weight percentNi–Al alloy (Aldrich) or 2.5 g Al pow-der ( J.T. Baker).

Determination of the Products of theReduction of Hydrazines and TheirOxidation Products

A 5.0-g aliquot of Ni–Al alloy and analkaline solution of hydrazine, MMH,UDMH, NDMA, or DMNM werecombined in an open beaker equippedwith a magnetic stirrer. Aliquots of thereaction mixture were withdrawn atvarious intervals for analysis. Eachalkaline solution was tested in thismanner, then the same set of experi-ments was repeated using the equiva-lent amount of Al powder (2.5 g) ratherthan Ni–Al alloy.

Small volumes of reactant mixturesremoved at various intervals from thereaction flask were rapidly filteredthrough a 0.45-micron filter to quenchthe reaction by removing suspendedmetal. These aliquots were analyzedfor hydrazines by high performanceliquid chromatography (HPLC) usinga Hewlett-Packard model 1,100 seriesHPLC with amperometric detection.16

NDMA and DMNM were analyzed bygas chromatography (GC) with a flameionization detector (FID) using a Hew-lett-Packard model 5890 GC. Contentsof the acid-filled gas washing bottleswere analyzed for ammonia, methyla-mine, and dimethylamine using capil-lary electrophoresis (CE) with indirectultraviolet (UV) detection using aHewlett-Packard model HD3D CE.

A 5.0-g aliquot of Ni–Al alloy andone of the above alkaline solutionswere combined in a two-necked roundbottom flask equipped with a magneticstirrer, a heating mantle, and a gaseousnitrogen purge over the surface of thereactant solution. A gas outlet from thetwo-necked flask was connected to agas-washing bottle containing ACSgrade hydrochloric acid ( J.T. Baker)

Chemical Health & Safety, January/February 2004 7

diluted to 0.001 M with deionizedwater. The reactions were allowed toproceed without heating for 1 hour,then the two-necked round bottomflask was rapidly heated to boiling for1 hour while maintaining the nitrogenpurge to sweep remaining volatile pro-ducts into the acid-filled gas washingbottle. At the completion of each test, analiquot of the gas washing bottle con-tents was withdrawn for analysis.

UDMH Spill Pillow Tests

Spill pillows that absorb and decom-pose UDMH were prepared by adding5.0 g of Ni–Al Powder and 20 g of solidsodium hydroxide to the interior ofpads. For spill pillow tests, 10 mL of1,000 mg/L UDMH were poured ontoa pad; samples were taken from liquidsqueezed from the pad at varioustimes. Aliquots of this liquid were ana-lyzed for UDMH by HPLC.

Reagents and Materials for DMAZReduction

DMAZ was obtained as an experimen-tal chemical from 3 M. N,N-Dimethy-lethylenediamine (95%), granular tin,and Ni–Al alloy were obtained fromAldrich. ACS grade stannous chloridedihydrate, hydrochloric acid, sodiumhydroxide, Al powder, dichlorome-thane, nitric acid, and methanol wereobtained fromJ.T.Baker.Aqueous solu-tions of DMAZ were prepared by dilu-tion with deionized water. A 1 Msodium hydroxide solution was pre-pared with deionized water.

DMAZ Reduction and ProductsDetermination

A 5.0-g aliquot of Ni–Al alloy or a 2.5-galiquot of Al and 100 mL of a5,000 mg/L DMAZ solution in 0.5 Msodium hydroxide were combined inan open beaker equipped with a mag-netic stirrer. Aliquots of the reactionmixtures were withdrawn at variousintervals and were passed through a0.45-micron filter prior to analysis.These aliquots were analyzed byGC–FID and GC–MS.

Raney nickel or Al

A 5.0-g aliquot of Ni–Al alloy or a 2.5-galiquot of Al and 100 mL of a5,000 mg/L DMAZ solution in 0.5 Msodium hydroxide were combined in

an open beaker equipped with a mag-netic stirrer. Aliquots of the reactionmixtures were withdrawn at variousintervals for analysis by GC–FID orGC–MS. The Al reaction mixturewas extracted with dichloromethaneafter 30 minutes. The dichloromethanewas subsequently concentrated by eva-poration prior to analysis by GC andGC–MS.

Stannous chloride in methanol

A 16-g aliquot of stannous chloridedihydrate was suspended with stirringin 40 mL of methanol. 3.9 g of DMAZor 2.6 g of N,N-dimethylethylenedia-mine was added slowly to the mixturewhile stirring. Aliquots of the reactionmixtures were withdrawn at variousintervals for analysis by GC–FID.The precipitates formed were isolatedby filtration, washed with methanoland then dried before analysis by FTIRusing a Spectra-Tech IR-Plan infraredmicroscope interfaced to a NicoletMagna-IR 750 infrared spectrometer.An Agilent Model 4500 inductivelycoupled plasma mass spectrometer(ICP-MS) was used for the analysisof tin in solutions of the precipitatesprepared in 5% nitric acid.

Tin in hydrochloric acid

A 6.0-g aliquot of granular tin wasadded to 100 mL of concentratedhydrochloric acid 1.1 g of DMAZwas added slowly to the mixture whilestirring. Aliquots of the reaction mix-tures were withdrawn at various inter-vals for analysis by GC–FID.

Bulk NTO Decomposition Reagents andMaterials

MON-3 NTO conforming to its respec-tive military specification was obtainedlocally.17 MON-3 was bubbled into tapwater by routing the NTO from a stain-less steel Department of Transportation(DOT) rated cylinder through a PTFEtube to the bottom the 4-L beaker con-taining the water. If the NTO wasalready in an open vessel, it was veryslowly poured into the beaker with stir-ring. A few drops of phenolphthaleinsolution were added and a solution of50% weight/volume technical gradesodium hydroxide (E.M. Science) wasadded while stirring until the solutionwas slightly basic.

Analytical NTO Decomposition Reagentsand Materials

NTO was transferred to a tared glassampoule, and the ampoule was flame-sealed. The ampoule was weighedagain and the mass of NTO was deter-mined by difference. A 250 mL heavy-walled borosilicate biological oxygendemand (BOD) bottle and stopperwas thoroughly rinsed with deionizedwater, and a water-rinsed PTFE-coated stir bar was placed therein.On hundred milliliters of deionizedwater and 20 mL of 30% ACS reagentgrade hydrogen peroxide (Mallinck-rodt) was added to the BOD bottle.A sealed ampoule of NTO was rinsedwith deionized water, placed in theBOD bottle, and the BOD bottle wasstoppered. The BOD bottle containingthe ampoule was chilled in an ice bathfor at least 15 minutes. The BOD bottleand stopper were held securely andshaken vigorously to break theampoule. Shaking was continued forapproximately 15 seconds until brownnitrogen dioxide fumes were no longervisible. The BOD bottle was returnedto the ice bath, and was removed andshaken for several minutes during thenext 15 minutes.

The stopper from each BOD bottlewas removed and rinsed with deio-nized water back into the BOD bottlewith deionized water. If necessary,any large pieces of the ampoule thatmight entrap liquid were broken upwith a clean, heavy glass rod. TheBOD bottle was fitted with a Vigreuxcolumn, placed on a hot plate, andheated for 45 minutes while control-ling the heat so the solution refluxedonly in the lower third of the Vigreuxcolumn. The BOD bottle was cooledto room temperature, then the inter-ior of the Vigreux column was rinsedwith deionized water and removed.Three drops of methyl red indicatorwere added to the BOD bottle, thentitrated with standardized sodiumhydroxide solution to a yellow-orangeend point.

Analytical HTP Decomposition Reagentsand Materials

HTP (70 or 90%) was obtained fromFMC. A 5-m length of 0.127 mm(0.005-inch) outer diameter (OD) pla-tinum wire was centered inside a 6-m

8 Chemical Health & Safety, January/February 2004

length of 1.59 mm (1/16 inch) innerdiameter (ID) 3.18 mm OD (1/8 inch)poly-tetrafluoroethylene (PTFE) tub-ing. The center 5-m portion of thetubing was wound into a coil arounda length of poly-vinylchloride (PVC)pipe and secured with some stainlesssteel safety wire to keep the tubing inplace. In use the coil is immersed in icewith the HTP inlet at the bottom of thecoil form. Fifty grams of HTP wastransferred to the receiving vessel (cen-trifuge tube) and a peristaltic pump setto deliver 0.5 mL per minute was usedto recirculate the HTP until oxygenevolution ceased. The entire length oftubing was then flushed with deionizedwater to bring the sample volume to50 mL. A separate deionized waterflush was then collected through theapparatus as an analytical blank. Theschematic drawing of the apparatus isshown in Figure 1.

RESULTS AND DISCUSSION

Raney Nickel Formation

Raney nickel is formed in situ by thedissolution of aluminum from theNi–Al alloy by hydroxide and the sub-sequent absorption of hydrogen in thepores of the finely divided nickel(Equation 2).

2NiðAlÞðsÞ þ 6H2O þ 2OH�

! 2AlðOHÞ4� þ 3H2 þ 2NiðsÞ (2)

The control experiments with alumi-num powder form hydrogen by the

action of hydroxide on aluminum inthe absence of nickel. This reactionproduces only hydrogen gas and alu-minate ion (Equation 3).

2ðAlÞðsÞ þ 6H2O þ 2OH�

! 2AlðOHÞ4� þ 3H2 (3)

The reactions of both Ni–Al alloyand aluminum with the 0.5 M sodiumhydroxide solutions produced obser-vable hydrogen gas evolution and heat.Excessive bumping and frothing wasavoided by the use of the magneticstirrer. Because isolation of the Raneynickel was not necessary, incorpora-tion of subsequent handling proce-

dures for pyrophoric materials notnecessary. Hydrochloric acid addedto each reaction mixture at the conclu-sion of each test dissolved the Ni andeliminated the possibility that Raneynickel could dry and ignite sponta-neously in air.

Hydrazines and Oxidation ProductsDestruction

Greater than 99% of the hydrazine,MMH, and UDMH were decomposedby Raney nickel in about 30 minutes(Figure 2).

The major products were ammoniaand the aliphatic amines that werepredicted on the basis of heterolyticcleavage of the N–N bond (Equations4–6).

H2N�NH2 þ H2!Ni

NH3 (4)

ðCH3ÞHN�NH2 þ H2

!NiðCH3ÞNH2 þ NH3 (5)

ðCH3Þ2N�NH2 þ H2

!NiðCH3Þ2NH þ NH3 (6)

The control experiments showedaluminum powder had no effect onthe starting concentrations of hydra-zine, MMH, and UDMH, suggestingthat reductive hydrogenation of thesecompounds must be catalyzed.

Figure 1. HTP decomposition apparatus.

Figure 2. Decomposition of hydrazine, MMH and UDMH by Raney nickel.

Chemical Health & Safety, January/February 2004 9

All detectable (>99%) NDMA wasdecomposed within 1 minute by Ni–Alalloy and its decomposition resultedin the formation of UDMH, an inter-mediate decomposition product thatfurther decomposed at a rate similarto that observed in the UDMH reduc-tion experiments with Raney nickel(Figure 3).

Overall, these results suggest the fol-lowing pathway for reductive decom-position of NDMA (Equation 7):

ðCH3Þ2N�NO !½H=Ni�ðCH3Þ2N�NH2

!½H=Ni�ðCH3Þ2NH þ NH3 (7)

DMNM was also decomposedwithin a minute by Ni–Al alloy, andits decomposition resulted in the briefformation and subsequent decomposi-tion of a small but detectable amountof NDMA (Figure 4). In addition,UDMH was formed and subsequentlydecomposed at rate similar to that asdescribed above.

Overall, these results suggest the fol-lowing pathway for reductive decom-position of DMNM (Equation 8):

ðCH3Þ2N�NO2 !½H=Ni�ðCH3Þ2N�NO

!½H=Ni�ðCH3Þ2N�NH2

!½H=Ni�ðCH3Þ2NH þ NH3 (8)

The molar yields of UDMH fromNDMA and DMNM were calculatedto be 103 and 80%, respectively. Thecontrol experiments with aluminumpowder alone rapidly decomposedNDMA and DMNM, but the decom-position did not progress past the for-mation of UDMH.

Raney Nickel Spill Pillow

Results of a spill pillow test with100 mL of 1,000 mg/L UDMH showedthe solution of UDMH was quantita-tively (>99%) decomposed within

30 minutes (Figure 5). These resultswere extremely promising, becausethe major product amines, ammoniaand dimethylamine are substantiallyless toxic than UDMH. Accordingly,spill pillows containing sodium hydro-xide and Ni–Al alloy could be used toabsorb and detoxify UDMH solutions.Special handling precautions, how-ever, would be required of such pillowsbecause the contents generate heatwhen exposed to water (exothermicreaction with sodium hydroxide) andNi–Al alloy is a carcinogen due to thepresence of nickel. Spill pillows, how-ever, could be kept hermetically sealedin a nitrogen atmosphere until theiruse with appropriate personal protec-tive equipment. Additionally, since dryRaney nickel is pyrophoric, used spillpillows could be placed in dilutehydrochloric acid to oxidize the nickeland stabilize the amine products. Theuse of a less toxic or hazardous catalystis another possibility.

DMAZ Destruction

Raney nickel or Al

DMAZ reacted rapidly with Raneynickel. Greater than 99.9% of DMAZfrom the initial 5,000 ppm solutionwas decomposed 2 minutes after thereaction was initiated (3 ppm DMAZremaining). The major reaction pro-duct was identified as N,N-dimethyle-nediamine with a molar yield of 94%.

Figure 3. Decomposition of NDMA by Raney nickel.

Figure 4. Decomposition of DMNM by Raney nickel.

10 Chemical Health & Safety, January/February 2004

Once a trace (about 5 ppm) of ammo-nia was found. These results suggestthe following reaction (Equation 9):

ðCH3Þ2�NCH2DMAZ

�CH2�N3 þ H2

!½Ni�ðCH3Þ2�NCH2�CH2�NH2 þ N2N;N-dimethylethylenediamine

(9)

The control experiments with alumi-num powder alone had little or noeffect on the starting DMAZ concen-tration. Ninety-six percent of theDMAZ from the initial 5,000 ppmsolution was unreacted after 30 min-utes of reaction time.

Stannous chloride in methanol

Addition of DMAZ to stannous chlor-ide dihydrate in methanol producedvigorous gas evolution and a yellow-white precipitate. DMAZ reactedrapidly in this reaction. Greater than99.9% of the DMAZ was decomposed(5 ppm remaining from the original97,500-ppm concentration of DMAZ)after 30 minutes of reaction time whendetermined by analysis of the reactionmixture filtrate. No N,N-dimethylethy-lenediamine was detected by GC,although a major, unidentified chro-matographic peak was observed. Themajor reaction product identified inthe filtrate was 2-chloroethyldimethy-lamine, which gave a 91% match withthe National Institute of StandardsTechnology (NIST) mass spectral

library. No hydrazoic acid wasdetected in an extract of the reactionmixture by GC–MS. These results andobservations suggest the followingreaction (Equation 10):

2ðCH3Þ2�N�CH2�CH2�N3DMAZ

þ SnCl2

!2ðCH3Þ2�N�CH2�CH2�Clþ3N2þSn2-chloroethyldimethylamine

(10)

The product was somewhat unex-pected and the mechanism for its for-mation is not known. Additionally,there was no available source of2-chloroethyldimethylamine to con-firm and quantitate the GC results orto confirm the GC–MS identification.

Analysis of the yellow-white crystal-line precipitate in the reaction mixtureby ICP-MS and FTIR indicated a mate-rial that was approximately 30% byweight tin, and whose FTIR spectrumindicated the presence of amines andwater. No further attempts to confirmthis reaction and products were made.

In an attempt to determine whetherN,N-dimethylethylenediamine was anintermediate in the reaction of DMAZwith stannous chloride in methanoland capable of further reaction, anequivalent amount of N,N-dimethy-

lethylenediamine was added to stan-nous chloride in methanol. A whiteprecipitate immediately formed. Ana-lysis of the precipitate in the reactionmixture by ICP-MS and FTIR indi-cated a material that was approxi-mately 30% by weight tin, and whoseFTIR spectrum indicated the presenceof an aliphatic amine salt.

Tin in hydrochloric acid

Addition of DMAZ to tin in hydro-chloric acid produced gas evolution,heat and a white, flocculent precipitateimmediately on addition of DMAZ.DMAZ reacted rapidly decomposedin this reaction. Greater than 99.9%of DMAZ (3 ppm remaining fromthe original 11,000-ppm concentrationof DMAZ) was decomposed after30 minutes of reaction time. The majorreaction product was identified asN,N-dimethylethylenediamine. Quan-tification of the N,N-dimethylethylene-diamine was not reproducible inmultiple attempts by gas chromato-graphic analyses, so a reliable molaryield cannot be reported. The results,however, suggest the following reac-tion, which is consistent with thatreported in the literature (Equation11)18

Analytical Decomposition of NTO

The analytical decomposition of NTOinvolves an initial hydrolysis and oxi-dation step as the NTO ampoule isbroken in the BOD bottle containinghydrogen peroxide, the net reaction isbelow (Equation 12). Because the ACSreagent grade hydrogen peroxide isstabilized with a small amount of phos-phoric acid, it is important to prepareand analyze a blank with the hydrogenperoxide and the deionized water.

N2O4 þ H2O2 ! 2HNO3 (12)

The resulting solution is titrated tothe methyl red endpoint. Assumingthat the acidity (minus the blank value)is derived from the NTO, the assay tobe calculated is based on the weightof the material in the ampoule.

Figure 5. UDMH decomposition by a Raney nickel spill pillow.

ðCH3Þ2�N�CH2�CH2�N3 þ Sn þ 2HClDMAZ

! ðCH3Þ2�N�CH2�CH2�NH2 þ SnCl2 þ N2N;N-dimethylethylenediamine

(11)

Chemical Health & Safety, January/February 2004 11

The resulting nitric acid is titrated bythe following reaction (Equation 13):

HNO3 þ NaOH ! NaNO3 þ H2O

(13)

Bulk Decomposition of NTO

The following reactions describe themajor NTO and water equilibria(Equations 14–19):

N2O4 fi 2NO2 (14)

N2O4 þ H2O ! HNO3 þ HONO

(15)

2NO2 þ H2O ! HNO3 þ HONO

(16)

3NO2 þ H2O ! NO þ 2HNO3 (17)

3HONO ! 2NO þ H2O þ HNO3

(18)

NO þ NO2 fi N2O3 (19)

The neutralization with base per-formed is the titration reaction (Equa-tion 20):

HNO3 þ HONO þ 2NaOH

! NaNO3 þ NaNO2 (20)

Quantities of up to 300 g of NTO inwater volumes of 8–12 L can easily behydrolyzed. The primary products ofinterest are nitric acid, nitrous acid andnitric oxide. The nitric oxide bubblesout rapidly as a gas, so the resultingsolution is a mix of the two acids. Thenitrous acid is continually decompos-ing to nitric acid and nitric oxide. Themixture of acids is neutralized by thesodium hydroxide, which yields thesodium salts and halts the nitric oxideevolution.

Catalytic Decomposition of HydrogenPeroxide

The decomposition of HTP as catalyzedby platinum is a very rapid reaction thatproduces large amounts of gas. It wasobserved that the liquid HTP reactedimmediatelywith the tip of theplatinumwire and the resulting gas evolvedtended to sweep the liquid throughthe decomposition coil rapidly, thusminimizing further reaction. As the

concentration of the HTP dropped withtime, it was observed that the liquid thatwas left was allowed to remain in con-tact with the platinum wire for a longerperiod as it traveled through the decom-position coil. The end of reaction waseasily observed when liquid flowingthrough the decomposition coil wasfree of oxygen bubbles. The conversionof the HTP (90%) to water usually wascomplete after allowing the system torun overnight. The resulting water pro-duct was then able to be analyzed foranions and metals as appropriate. Thecatalytic decomposition of HTP isdescribed by the following reaction(Equation 21):

2H2O2!½Pt�

2H2O þ O2 (21)

CONCLUSIONS

Reduction of hydrazine, MMH,UDMH, and UDMH oxidation pro-ducts (NDMA and DMNM) withRaney nickel was complete and wellcharacterized. The decomposition pro-ducts were aliphatic amines andammonia, which are substantially lesstoxic than the starting products.19 Thepathway of DMNM decompositionwas shown to proceed through NDMAand UDMH intermediates. Use of aRaney nickel-containing spill pillowto remediate a simulated UDMH spillwas successfully demonstrated.

Reduction of DMAZ with Raneynickel was complete and the reac-tion product N,N-dimethylethylene-diamine was identified. The reductionof DMAZ with tin in hydrochloric acidyielded tin- and amine-containing pre-cipitates, and similar products werefound when the predicted productamine N,N-dimethylethylenediaminewas reacted with tin in hydrochloricacid under the same conditions. Thereduction of DMAZ with tin inhydrochloric acid was complete andthe major reaction product was N,N-dimethylethylenediamine, but theproduct yield was inconclusive dueto analytical inference. However,there was no clear advantage in ourlaboratory to decompose DMAZto N,N-dimethylethylenediamine be-cause the product is similarly flam-mable and corrosive as the startingmaterial.

Quantities of up to 300 mL of dini-trogen tetroxide were hydrolyzed, thenthe acidic nitrate/nitrite solution wasneutralized. This can be of use in redu-cing nitrogen dioxide emissions,although the reaction does not stoi-chiometrically trap nitric oxide. Ana-lytical samples were oxidized to nitricacid with hydrogen peroxide to pre-serve the analytical integrity of thefluid, then titrated with sodium hydro-xide to produce a sodium nitrate andsodium nitrite solution.

Hydrogen peroxide was catalyticallydecomposed with platinum. This reac-tion can be violent unless controlled,and a novel method for unattendeddecomposition was demonstrated.The primary equipment necessary toperform the laboratory destruction ofthe liquid propellant hypergols wasordinary laboratory equipment includ-ing a beaker and a stirrer.

References1. http://www.wstf.nasa.gov2. Schmidt, E. W. Hydrazine and Its

Derivatives: Preparation, Properties,Applications, 2nd ed.; Wiley/Inter-science, New York, 2001.

3. Fire, Explosion, Compatibility, andSafety Hazards of Hypergols-Hydra-zine. American Institute of Aeronau-tics and Astronautics Special ProjectReport, AIAA SP-084-1999.

4. Fire, Explosion, Compatibility, andSafety Hazards of Hypergols-Mono-methylhydrazine. American Instituteof Aeronautics and Astronautics Spe-cial Project Report, AIAA SP-085-1999.

5. Tuazon, E. C.; Carter, W. P. L.; Atkin-son, R.; Winer, A. M.; Pilts Jr., J.N.Atmospheric reactions of N-nitrosodi-methylamine and dimethylnitramine.Environ. Sci. Technol. 1984, 18, 49–54.

6. Fire, Explosion, Compatibility, andSafety Hazards of Nitrogen Tetroxide.American Institute of Aeronautics andAstronautics Special Project Report,AIAA SP-086-2001.

7. Proceedings of the 5th InternationalHydrogen Peroxide Propulsion Con-ference, West Lafayette, IN, Septem-ber 2002; Purdue University.

8. Lunn, G.; Sansone, E. B. Destructionof Hazardous Chemicals in the La-boratory, 2nd ed.; Wiley/Interscience,New York, 1994.

9. Armour, M.-A. Hazardous LaboratoryChemicals Disposal Guide, 2nd ed.;Lewis Publishers, New York, 1996.

12 Chemical Health & Safety, January/February 2004

10. Greene, B.; Johnson, H. T. CatalyticDecomposition of Propellant Hydra-zines, N-nitrosodimethylamine, andN-nitrodimethylamine; in Proceedingsof JANNAF Propulsion and Subcom-mittee Meeting, Florida, 2000.

11. Greene, B.; Haney, W. A.; Johnson, H.T. Laboratory Treatment of Dimethyl-2-azidoethylamine (DMAZ) Waste; inProceedings of JANNAF Propulsionand Subcommittee Meeting, Colorado,2002.

12. Greene, B.; McClure, M. B.; Johnson,H. T. Destruction of hypergolic che-micals in a liquid propellant testinglaboratory. Division of ChemicalHealth & Safety (Cosponsored withDivision of Environmental Chemis-

try); in Proceedings of 224th AmericanChemical Society National Meeting,Boston, MA, August 2002.

13. MIL-PRF-28536E. MonopropellantGrade, Hydrazine, Military Specifica-tion. United States Department ofDefense, Washington, DC, 1997.

14. MIL-PRF-27404C. Methylhydrazine,Military Specification. United StatesDepartment of Defense, Washington,DC, 1997.

15. MIL-PRF-25604E. Propellant, uns-Di-methylhydrazine, Military Specifica-tion. United States Department ofDefense, Washington, DC, 1997.

16. Johnson, D. L.; Baker D. L. Determina-tion of Hydrazine, MMH, and UDMHin Aqueous Solution Using Cation

Exchange HPLC with AmperometricDetection; Application to AirbornePropellant. Chemical Propulsion In-formation Agency, CPIA-PUB-588,1992.

17. MIL-PRF-26539E. Propellant MixedOxides of Nitrogen, Military Specifica-tion. United States Department ofDefense, Washington, DC, 1997.

18. Augustine, R. L. Heterogeneous Cata-lysis for the Synthetic Organic Che-mist; Marcel Dekker, Inc., New York,1996.

19. ACGIH. Threshold Limit Values forChemical Substances and PhysicalAgents. American Conference of Gov-ernmental Industrial Hygienists, OH,1999.

Chemical Health & Safety, January/February 2004 13