1. Stabilisation of nitrogen triiodide.

1.1 Under standard conditions

Nitrogen triiodide, is under normal atmospheric pressures and temperatures, a contact explosive. When disturbed by air movement or vibration, but not excessive heat ( >75C), it will explode.

When dry (the ammonia having evaporated from the adduct structure), the compound is unstable. The following sets of experiments determined the stability in air, under water, under dilute ammonia, as an organic adduct and under 0.880 S.G. ammonia (31% weight by volume [w/v]).

The amount of adduct used was measured as the amount of iodine used rather than the mass of NI3 produced. The NI3 was wet when produced and the ammonia also evaporated. The result of this loss was that a constant mass of sample could not be formed, though it could be estimated by a graph of mass of sample vs. time. The result was compared with the mass of iodine originally used. This mass / time estimate was unreliable.

1.1.1 Methods Method in still air (enclosed environment)

5 g of iodine was ground into a fine powder in a mortar and pestle. This was transferred to a weighing bottle and sealed. Iodine at room temperature will sublime slowly, so keeping the bottle sealed will keep the maximum amount of iodine. The bottle had been pre-weighed and the amount of iodine held measured.

Into each of 5 100 cm3 conical flasks (previously washed with chromic acid, rinsed with deionised water several times, heat - dried to a constant weight), 50 cm3 of 0.880 S.G. ammonia was pipetted.

0.5 g of iodine was weighed into a preweighed weighing bottle and added to the first flask. This process was repeated for each of the other flasks. The flasks were then placed into a fume cupboard and left for fifteen minutes to enable the NI3 to form.

After fifteen minutes, the sample from the first flask was filtered under slightly reduced pressure with any remaining NI3 rinsed out with the mother liquor until the solution coming through the filter was clear. This sample was left on the paper at the back of an unused cupboard, free from external draughts and vibrations. The ambient temperature and pressure in the cupboard were considered constant.

The filtering of the other samples was carried out in the same way and the products placed into flasks containing water, pyridine, 0.880 S.G. ammonia and 2 mol dm-3 ammonia, each with a volume of 50cm3 liquid. These samples were kept in a fume cupboard which had been switched off (to avoid unnecessary vibrations).

After twenty minutes, the samples under each liquid and the dry sample were reweighed to find the amount decomposed. Twenty minutes was used, since after this period of time, any gas produced should have dispersed. The samples in liquid were filtered under normal pressure using pre-weighed filter papers. The Vibration method

The same method for preparation of NI3 and filtering as in was used for this set of experiments.

The dry sample was filtered and left to dry in a still environment for ten minutes.

The method of testing used was to place the solid and the solution flasks onto a vibrating surface (in this case, in a sonic bath was used). The time taken for decomposition was recorded for each sample or recorded as "after twenty minutes" which ever came sooner.

The energy created in the flask and liquid experiments was found by placing a flask containing the same volume of liquid as used in the experiment into the sonic bath for the same period of time used in the experiment. Any temperature rise was noted. This was subtracted from the initial figure. Thermal method

The method used for was used again for a single sample. This was placed in a beaker on a hotplate with a variable temperature control. At five minute intervals, the temperature was increased up to the next setting (0 - 10), left to equilibrate, left for a further five minutes and then the temperature increased again.

The experiment was over when the adduct detonated, The time and temperature was recorded. Direct detonation.

The NI3 sample was made as in and allowed to dry. A cone speaker connected to a signal generator was placed above the sample and switched on. A range of frequencies were used until the correct one for detonation found.

1.2 Agar gel times

Agar gels form a matrix with any aqueous liquid by interaction between the aqueous medium and the sulphur - nitrogen links in the agar. The gel is strong but liable to attack by bacteria and also by the evaporation of water from the matrix which makes the gel break down into a brittle rusk.

1.2.1 Gel times using varied concentrations of ammonia

Ammonia (0.880 S.G.) contains about 31% w/v ammonia. In other words, 100 cm3 of 0.880 S.G. ammonia will contain 31 g ammonia. The rest is water.

A series of solutions containing varying amounts of ammonia, each of a selected volume, was boiled with constant stirring. When boiling, 3.12 g of high strength agar powder was added until fully mixed. This mass was used as a stable gel forms with this mass of the high strength agar powder.

Each agar liquid was poured into a test tube held in a thermostated water bath set at room temperature. Into each liquid was inserted a pipette connected to an oxygen cylinder which had a known gas flow rate.

The gel time was determined by noting when the final bubble was so deformed that it was unable to move through the matrix and break the surface of the gel.

1.3 Determination of minimum ammonia required for stabilisation

As seen in 1.1, NI3 is best kept in 0.880 S.G. ammonia as here it is stable at all temperatures. This set of experiments was designed to see if a lower concentration of ammonia could be used and from this, the best (with respects to ammonia content and stability) ammonia / water agar gel be found.

1.3.1 Method

This test was an adaptation of with the exception that the ammonia was in varied concentrations. The solutions were placed on the vibration bed. The time required for each detonation was recorded. These will vary between that for 100% water and the 0.880 S.G. ammonia.

1.4 Infrared scans

The infra red scan was performed using the nitrogen triiodide trapped in an agar made up with 0.880 S.G. ammonia and left for 24 hours to set. This was then ground and a Nujol mull prepared and scanned using NaCl or KBr plates as appropriate.

1.4.1 Method

3.12 g of high strength agar is added to boiling 0.880 S.G. ammonia with constant stirring. After ten minutes, a small amount of nitrogen triiodide (wet) was mixed in and allowed to set for a day.

The gel was then ground into a powder and made into a mull with Nujol and smeared between the i.r. plates.

An i.r. scan was taken with Nujol is taken as the background between 4000 and 300 cm-1. Scans of agar and the gel were then taken over the same wavelength range.

As the lower end of the scale is below that of sodium chloride (the cut off point being around 700cm-1), potassium bromide plates were used.

1.5 Ultraviolet scanning

Nitrogen triiodide has a dark colour, so there should be a wavelength for maximum absorbance.

U.V. does depend on the sample being in solution. NI3 is insoluble in inorganic solvents (such as water and aqueous ammonia), but will dissolve in some aprotic solvents (such as the pyrrole/pyridines)1.

Samples of NI3 were taken and dissolved in pyridine, piperidine and pyrrole. These are all nitrogen - containing solvents, but with varying amounts of saturation. The absorbance readings were taken over a range of 190 to 820 nm using a quartz cell. The concentrations used were not important due to the purpose being just to determine the lmax. for NI3.

1.6 X-Ray diffraction

The object of this experiment is to determine the structure of nitrogen triiodide.

A series of gels containing agar (water made), agar (ammonical), NI3 - agar and agar powder was placed onto sample trays and using a Siemens D5000 X-ray diffraction apparatus and the intensity vs. angle measured.

1.7 Bomb calorimetry

The bomb calorimeter provides a very simple method for the direct determination of the heat of detonation.

The bomb is calibrated using a known mass tablet of A.R. grade benzoic acid in tablet form.

The bomb was set up as described in the theory section. The water bath was set at 25C. Some benzoic acid was pressed into a tablet using a suitable press and the tablet weighed. The tablet was then tied to a known length of cotton and the cotton tied to the bare fuse wire. This was placed into the bomb. The bomb was sealed and pressurised to 25 atmospheres with oxygen. This was then released and re-pressurised to 25 atmospheres to ensure that the bomb was free of air and moisture and had an excess of oxygen in it.

The Beckman thermometer reading was taken for five minutes prior to detonation. The sample was detonated and the temperature taken every thirty seconds. The temperature was taken for four minutes past the final temperature (the levelling off period).

1.7.2 Calorimetric determination for the sample

Two different measurements were taken for this experiment.

The first was for the agar itself. The second was for the NI3 / agar - ammonical gel itself.

The first reading was needed for subtraction from the second to give the true thermal value for NI3. Method

Some of the high strength agar was pressed into a tablet, weighed, and placed into the bomb calorimeter as in 1.7.1. This was then detonated and the temperature changes taken.

Some of NI3-agar was pressed into a tablet, weighed, and placed in the bomb. Again, it was detonated and the temperatures taken.

1.8 Glass bomb calorimetry

The use of the glass in any sort of pressure measurements was determined as "ill advised" in the mid 1960's after a spate of accidents using glass tubes during heating pressure tests. The glass could shatter creating a potential danger to other workers.

A glass calorimeter was constructed using thick walled glass (approximately 5mm) of a known composition glass (Pyrex). The internal volume was small. Large pressures were not used. The sample size was a maximum of 0.4g, and the expansion of the gas was minimal.

1.8.1 Calibration

The calibration was performed as a two - part experiment. The first was to determine the thermal capacity of the glass, the second was a standard calibration using benzoic acid. Thermal capacity

The bomb was filled with water at 50C and placed in the polystyrene cup which was full of water. The water in the cup had been allowed to stand until the temperature registered by the bead thermistor was constant.

The temperature changes (after the addition of the bomb) were recorded on the computer until they reached a constant value. The thermal capacity will simply be represented by the change of temperature in the water when the mass of water is known. Calibration of the glass calorimeter

Using an oxygen air bag, a known mass of high grade benzoic acid was placed into the the volcano. The calorimeter was sealed, removed from the bag and placed into the polystyrene cup. The cup was two thirds full of water which had been allowed to reach thermal equilibrium. The computer was already taking readings for the cup.

The cup was sealed by placing the lid on. The sample was detonated and the temperature changes recorded by the computer.

1.8.2 Direct determination of NI3

As the amount of NI3 used for this experiment was very small (less than 0.4g), the overall expansion of gases would be less than that judged to break the glass.

The same procedure was used as for with the following changes :

1. The calorimeter was cleaned and heat dried to a constant mass.

2. NI3 was used in place of benzoic acid and that the sample was allowed to stand for twenty minutes before sealing. This was to ensure the ammonia had evaporated from the adduct.

The mass of NI3 in the calorimeter was the difference between the respective masses.

1.9 Rate of ammonia evaporation from agar and NI3 adduct.

The rate at which ammonia is lost from both the ammonical-agar and the NI3 adduct is important. The reason for this is two fold; firstly, to see how much ammonia is held within the structure after a given time and secondly, as NI3 is here an adduct with ammonia, the time required for dryness is important. When the sample is wet, the explosive capacity is greatly diminished. When dry, the amount of time before the solid explodes will be short. In effect, the ammonia desorption is a fuse for the explosive.

1.9.1 Method

The same method was adopted for both the sample (NI3 adduct in the ammonical agar) and the ammonical agar as it gave a good result and was easy to perform (though not quick). For arguments sake, I will describe both as the sample.

A known weight of the sample was placed into a weighing bottle and placed onto a four point digital balance. The mass was recorded initially every five minutes for the first hour, every fifteen for the second hour, then every half hour for the next three hours. A final reading was taken at the end of the sixth hour.

The agar sample was left until twenty four hours had elapsed and the final mass recorded.

1.10 Theoretical estimation of dHdet NI3.

The theoretical determination of dHdet is not simple. This is due to there being no recorded value for the N-I bond dissociation energy. A Born - Haber diagram cannot be constructed for this process.

Other bond energies are well known for the likes of =N-H and H-I (as well as H-H). By using the equation for the reaction

N-H + H-I - (H-H) -> N-I

and then multiplying the result by 3 (to give NI3), the approximate value for the heat of dissociation can be obtained.

The value will be a first approximation, but it will not be completely accurate as there will be a difference in the bond energies after each successive removal of an iodine. There will also be a difference due to the free energies of the nitrogen and iodine when they combine. It is probable that there will be some form of sp hybridisation and this will create another error.

1.11 Determination of ammonia in the sample

The gas chamber had had the carrier gas, helium, pumped through for thirty minutes before the experiment commenced to ensure that the chamber was dry, to reach thermal equilibrium with the gas and to equalise the pressures between the cells and the incoming gas. The gas flow rate was measured. A time period of thirty minutes was required as the glass would reach a the temperature of the gas faster on the side it was being pumped into than the other side. The temperatures of the two sides of the vessel were taken at 3 minute intervals and after around 30 minutes, they were at the same temperature.

A sample of known weight was placed into one side of the gas chamber with 50cm3 of standardised hydrochloric acid (0.1 mol dm-3) in the other side of the chamber. The chamber was then left for 5 hours in a vibration free environment (in this case, on a thick polythene mat) in a switched off fume cupboard. The gas was vented off into a neighbouring fume cupboard which was switched on. After this time, all of the ammonia will have left the sample and the sample will have detonated, so releasing any ammonia which may have been "trapped" within the centre of the adduct structure.

The acid was titrated after this time against freshly prepared sodium carbonate solution (0.05 mol dm-3). Using the stoichiometry CO3-2 : H+ the concentration of the acid could be determined. As the stoichiometry of ammonia to hydrochloric acid is 1 : 1, the difference in the acid concentration would equal the concentration of ammonia absorbed from the adduct. This can then be related to a molar volume and the final composition of the adduct calculated.