The purpose of this study was to characterize the nonvolatile, water soluble, bis(hydroxyalkyl)N-nitrosamines produced from the reaction of spermidine and nitrite. The biologically occurring polyamine spermidine was reacted with nitrite in the presence of acid. The reaction products were isolated and converted to trimethylsilyl (TMS) derivatives. The derivatized products were separated by gas-liquid chromatography using an all glass, wall coated, capillary column. Mass spectral data were collected on the chromatographic effluent. Four isomers of bis(hydroxyalkyl)N-nitrosamine were synthesized by combining the appropriate aminopropanol with the appropriate chlorobutanol. Infrared, nuclear magnetic resonance, mass spectral, and thin-layer chromatographic data were collected on the synthetic compounds to assure their structure. Kovat's indices were collected on the TMS derivative of each authentic compound and compared to the retention data collected on the TMS derivatives of the reaction products of spermidine and nitrite. A comparison of the mass spectra of the TMS derivatives of the authentic compounds to the spectra of the reaction products having the same retention time allowed structural assignments to four compounds. The compounds identified, in order of decreasing amounts were: 4-hydroxybutyl(3-hydroxypropyl)N-nitrosamine, 3-hydroxybutyl(3-hydroxypropyl)N-nitrosamine, 4-hydroxybutyl(2- hydroxypropylJN-nitrosamine, and 3-hydroxybutyl(2-hydroxypropyl)N-nitrosamine. / Graduation date: 1977

Graduation date: 1979

No description available.

Nitroguanidine (NQ) is a nitramine used in many propellant formulations. The purpose of this study was to develop a method for treating nitroguanidine process wastewaters. The major wastewater contaminants were found to be nitroguanidine and guanidine nitrate (GN) (an intermediate product). Pilot tests were conducted for the removal of nitroguanidine using granulated carbon adsorption (GAC) and ion exchange for the removal of GN. The nitroguanidine concentration in demonstration plant wastewater was reduced from 15.5 mg/L to less than 1 mg/L. The carbon capacity was found to be 0.045 g NQ/g carbon. Strong acid resin was found to reduce guanidinium ion Gu⁺ concentration from 14.5 mg/L to 1 mg/L with a resin capacity of 0.13 eq Gu⁺ /L resin. No significant reduction in resin capacity for Gu⁺ was experienced after five regenerant cycles. A unique two-step regeneration was used. This included a 5% calcium nitrate solution to produce a spent regenerant that could be recycled to the production process. This was followed by regeneration with a 15% sodium chloride solution to place the resin in the sodium form. Anion exchange was used for the removal of nitrate ions. / Master of Science

Mesoscale models are needed to predict the effect of changes to the microstructure of plastic-bonded explosives on their shock initiation and detonation behaviour. This thesis describes the considerable progress that has been made towards a mesoscale model for two HMX-based explosives PBX9501 and EDC37. In common with previous work in the literature, the model is implemented in hydrocodes that have been designed for shock physics and detonation modelling. Two relevant physics effects, heat conduction and Arrhenius chemistry, are added to a one-dimensional Lagrangian hydrocode and correction factors are identified to improve total energy conservation. Material models are constructed for the HMX crystals and polymer binders in the explosives, and are validated by comparison to Hugoniot data, Pop-plot data and detonation wave profiles. One and two-dimensional simulations of PBX9501 and EDC37 microstructures are used to investigate the response of the bulk explosive to shock loading. The sensitivity of calculated temperature distributions to uncertainties in the material properties data is determined, and a thermodynamic explanation is given for time-independent features in temperature profiles. Hotspots are widely accepted as being responsible for shock initiation in plastic-bonded explosives. It is demonstrated that, although shock heating of crystals and binder is responsible for temperature localisation, it is not a feasible hotspot mechanism in PBX9501 and EDC37 because the temperatures generated are too low to cause significant chemical reaction in the required timescales. Critical hotspot criteria derived for HMX and the binders compare favourably to earlier studies. The speed of reaction propagation from hotspots into the surrounding explosive is validated by comparison to flame propagation data, and the temperature of the gaseous reaction products is identified as being responsible for negative pressure dependence. Hotspot size, separation and temperature requirements are identified which can be used to eliminate candidate mechanisms in future.