The analysis of unfired propellant particles by gas chromatography mass spectrometry a forensic approach

This technique focuses on the organic constituent make up of the propellant paying particular attention to diphenylamine, ethyl centralite and dibutyl phthalate.. Within the field of unf

in fulfilment of the requirements for the degree of

Masters of Applied Science (Research)

by Shiona Croft Bachelor of Applied Science

Under the Supervision of:

Dr John Bartley

School of Physical and Chemical Sciences Queensland University of Technology

April 2008

In Australia, the 0.22 calibre ammunition is the most encountered ammunition type found at a crime scene [1] Previous analysis of gun shot residue (GSR) and unfired propellant has involved studying the inorganic constituents by Scanning Electron Microscopy or similar technique However, due to the heavy metal build up that comes with some ammunition types, manufacturing companies are now making propellant that is safer to use Therefore, it has become appropriate to study and analyse unfired propellant by other means One such technique is unfired propellant analysis by gas chromatography – mass spectrometry (GC-MS) This technique focuses on the organic constituent make up of the propellant paying particular attention to diphenylamine, ethyl centralite and dibutyl phthalate It was proposed that different batches of ammunition could be discriminated or matched to each other

by using this technique However, since the main constituents of unfired propellant are highly reactive, it was not possible to accomplish batch determination of ammunition However, by improving extraction techniques and by removing oxygen (a catalyst for the degradation of diphenylamine) a superior method was established

to help in the analysis of unfired propellant Furthermore, it was shown that whilst differentiating batches of the same ammunition was not possible, the improved methods have helped identify different types of the same brand of ammunition With the aid of future studies to fully explore this avenue, the analysis of unfired propellant could one day become an integral part of forensic science

The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made

To the Queensland Police Service (Mr Gary Asmussen and the members of the Analytical Services Unit) for allowing me to take up this research but for also giving

me the freedom to explore this project in the direction I thought most appropriate Thank you

To my colleague and friend Dr Helen Panayiotou, thank you for your words of wisdom Your encouragement and valuable direction when I felt lost was appreciated greatly

To my mum and dad who has been supportive from day one Your support, enthusiasm and confidence in my abilities allowed me to have courage in my work Thank you for never allowing me to give up – although I am too stubborn to do so!

To my dear Chris, who everyday told me how proud of me he was Thank you for putting up with the late nights and the stress For your love, friendship and strength –

I honestly could not have done this without you You mean everything to me

To my brother Kevin, who I know is very proud of me Thanks for your support Kev!

To my Ouma and Grandad, Connie and Gerald Campbell, I wish you could be here but you are always in my thoughts Thank you for your support and interest in my thesis It means so much to me that even though you are far away your love and encouragement is not forgotten I miss you

To my very much loved group of friends; Scott, Niki, Amy, Mick, Nikki and everyone else who has been there for me Some of you have been around for more than a decade and your love, encouragement and support is never forgotten You all mean the world to me and thank you for giving me the strength to go on

Finally, to all the post graduate students whom I may not have seen as much as I would have liked (since being off campus) but to my friend Dr Sarah Ede in particular, who constantly inspired me and who I always knew would do great things Thank you

$%675$&7 II 67$7(0(172)25,*,1$/$87+256+,3 III

$&.12:/('*(0(176 IV 7$%/(2)&217(176 VI /,672)),*85(67$%/(6 VIII

$%%5(9,$7,216 XI

 ,1752'8&7,21 1

1.1 B ACKGROUND 1

1.2 T HE 0.22 CALIBRE AMMUNITION 1

1.2.1 The Cartridge 2

1.2.2 The Projectile 2

1.2.3 The Propellant 3

1.2.4 The Primer 3

1.3 P REVIOUS WORK RELATED TO O RGANIC G UN SHOT RESIDUE OR UNFIRED PROPELLANT ANALYSIS 4

 (;3(5,0(17$/ 18

2.1 M ATERIALS 18

2.2 I NSTRUMENTATION 18

2.3 S TANDARD PREPARATION 19

2.4 E THYL ACETATE ALONE PROCEDURE 19

2.5 E THYL ACETATE / DICHLOROMETHANE PROCEDURE 19

2.6 C ONSISTENCY OF PROPELLANT COMPOSITION EXPERIMENT 20

2.7 E XCLUSION OF OXYGEN EXPERIMENT 20

2.8 T YPE DETERMINATION OF W INCHESTER AMMUNITION 20

 5(68/76$1'',6&866,21 22

3.1 M ASS SPECTRA OF UNFIRED PROPELLANT COMPONENTS 22

3.1.1 Diphenylamine (C 12 H 11 N) 22

3.1.2 Ethyl centralite (C 17 H 20 N 2 O) 24

3.1.3 Dibutyl phthalate (C 16 H 22 O 4 ) 26

3.2 C ONTROLLED S TANDARDS 28

3.2.1 Diphenylamine, ethyl centralite and dibutyl phthalate variation 29

3.3 T HE ANALYSIS OF PROPELLANT USING E THYL A CETATE ALONE 30

3.4 R EMOVAL OF THE NITROCELLULOSE COMPONENT OF PROPELLANT USING ETHYL ACETATE AND DICHLOROMETHANE 44

3.5 C ONSISTENCY OF PROPELLANT COMPOSITION FROM A SINGLE BOX / BATCH OF AMMUNITION 53

3.6 T HE EFFECTS OF EXCLUDING OXYGEN 58

3.7 T YPE DETERMINATION OF W INCHESTER A MMUNITION 65

3.7.1 Winchester Laser LR HP 2DRM41 65

3.7.2 Winchester Expert 23DLH02 66

3.7.3 Winchester Winner IDKE52 66

3.7.4 Winchester Subsonic LR Rim fire AED1FH31 67

3.7.5 Winchester Superspeed LR HV solid SDSB51 68

3.7.6 Winchester Superspeed LR HV hollow point 2DRL62 69

 &21&/86,216$1')8785(:25 72

4.1 C ONCLUSIONS 72

4.2 F UTURE W ORK 73

$33(1',; 74

Ethyl Centralite standard 74

Dibutyl phthalate standard 75

Diphenylamine standard 76

Winchester Laser Long Rifle Hollow point 2DRM41 77

Winchester Expert 23DLH02 78

Winchester Winner IDKE52 79

Winchester Subsonic Long rifle Rim fire AED1FH31 80

Winchester Superspeed Long Rifle High velocity solid 2DSB51 80

Winchester Superspeed long rifle high velocity hollow point 2DRL62 81

5()(5(1&(6 82

Figure 1.1: Dissected view of the case [8] 2

Table 1.1: Elution order for constituents using HPLC-MS[29] 6

Figure 1.2 Degradation of DPA [10,11] 7

Table 1.2: Results from Northrop[46,47] 15

Table 3.1: Selected target compounds (from NIST library) 22

Figure 3.1 Mass spectrum of diphenylamine 24

Figure 3.2 Chemical structure of diphenylamine and m/z = 77 fragment ion (C6H5) 24

Figure 3.3: Ethyl Centralite 25

Figure 3.4: Mass spectrum of ethyl centralite 25

Figure 3.5: Fragmentation ions of ethyl centralite 25

Figure 3.6: Fragment A = m/z 120 and Fragment B = m/z 148 26

Figure 3.7: General structure of phthalate esters where R, R’ = CnH2n+1; n=4-15[50] 26

Figure 3.8: Characteristic fragmentation ions of butyl phthalate esters [51] 27

Figure 3.9: Mass spectrum and chemical structure of dibutyl phthalate 28

Table 3.2 DPA, EC and DBPH standard variation 29

Table 3.3 Variation in peak area of DPA between three random propellant samples and variation observed between the three samples 31

Figure 3.10: Diphenylamine degradation over time 31

Figure 3.11: Sample 1 - diphenylamine degradation 32

Figure 3.12: Sample 2 - diphenylamine degradation 33

Figure 3.13: Sample 3 - diphenylamine degradation 33

Figure 3.14: 2-nitro-diphenylamine amounts detected (three random samples) – Extrapolated to time zero to give appreciation of initial amounts of 2-nitro-DPA in each sample 34

Figure 3.15: Mechanisms of N-Nitroso-DPA in the presence of NO2 and O2 [16] 36

Figure 3.16: Lussier and Gagnon [14]: Concentration of DPA and its derivatives as a function of added nitrogen dioxide 37

Figure 3.17: Effect of storage on diphenylamine concentration (sample one and two)

38 Figure 3.18: Sample 1 - Effect of storage on diphenylamine (triplicate) 39 Figure 3.19: Sample 1 - The effect of storage on 2- nitro-diphenylamine (analysed

three times) 40 Figure 3.20: Sample one and two analysed each three times (average): comparison

between stored samples and samples left in solution for one week 41 Figure 3.21: The effect of leaving propellant in solution over one week (2-nitro-DPA

average – each sample analysed three times) 42 Figure 3.22: Diphenylamine response (EtAc/Ch2Cl2 procedure) 45 Figure 3.23: Comparison between EtAc alone and EtAc/CH2Cl2 on DPA (average)

46 Figure 3.24: Comparison of EtAc alone and EtAc/CH2Cl2 procedures - 2-nitro-dpa

(average: each sample analysed three times) 47 Figure 3.25: 2-nitro-dpa levels - sample 1: comparison between EtAc alone and

EtAc/CH2Cl2 procedures 48 Figure 3.26: Comparison between EtAc alone and EtAc/CH2Cl2 procedures (ethyl

centralite average) 50 Figure 3.27: Dibutyl phthalate – comparison between EtAc alone and EtAc/CH2Cl2

procedures 51 Table 3.4: Relationship between sample size and population size 53 Table 3.5: Masses from ten random samples from one box of ammunition 53 Figure 3.28: Levels of diphenylamine of ten (10) random samples of propellant from

the one box of ammunition 55 Figure 3.29: Levels of dibutyl phthalate detected of ten (10) random samples of

propellant from the one box of ammunition 55 Table 3.6: Ratios (peak area) of main constituents from one box of ammunition 56 Figure 3.30 Inert gas procedure consequences on the main constituents of unfired

propellant particles 58 Figure 3.31: Inert gas procedure (dibutyl phthalate) 60 Figure 3.32: The effects of leaving propellant in solution under an inert atmosphere

61

Figures 3.33: Re-analysis of samples A-D 24 hours later (individually separated for

visual clarity) 63

Figure 3.34: Winchester Laser LR HP 2DRM41 65

Figure 3.35: Winchester Expert 23DLH02 66

Figure 3.36: Winchester Winner IDKE52 67

Figure 3.37: Winchester Subsonic LR Rim fire AED1FH31 68

Figure 3.38: Winchester Superspeed LR HV solid 2DSB51 69

Figure 3.39: Winchester Superspeed LR HV hollow point 2DRL62 70

Table 3.7: Type determination of Winchester ammunition 71

GC-MS: Gas chromatography – mass spectrometry

OGSR: Organic gun shot residue

SEM: Scanning electron microscopy

THC: Tetrahydrocannabinol

THC acid: Tetrahydrocannabinol acid

TLC: Thin layer chromatography

2-nitro-DPA: 2-nitro diphenylamine

4-nitro-DPA: 4-nitro diphenylamine

One area of research that has been explored is investigating the organic constituents

of unfired propellant to identify their key role in ammunition make up and functions

It is widely known that the inorganic components of gunshot residue are readily analysed by the scanning electron microscope to identify lead, barium and antimony and other key ingredients[2-5] However, the potential health risks associated with heavy metal build up, have led to heavy metal free ammunition being introduced [6]

In this situation, organic analysis of the residue or propellant is required and subsequently, analysts have demonstrated that this is possible There are some discrepancies though with which constituents are unique to smokeless powders, which will aid in the identification of the ammunition

In Australia, the use of 0.22 calibre ammunition is wide spread[1] and so, investigative forces are mostly concerned with this type of ammunition To fully appreciate the diversity of the 0.22 calibre projectile some background information will be given about the physical make up of ammunition itself, including: the cartridge and cartridge case, projectile, propellant; and primer and their roles in the organic make-up of the ammunition[1,7]

1.2.1 The Cartridge

The calibre of the ammunition refers to the diameter of the bore inside the firearm A typical cartridge will contain the case, primer, propellant and projectile The ammunition can either be rim fire or centre fire In rim fire ammunition, the priming materials are concentrated around the outer edge of the base of the cartridge making the rim the most susceptible to detonation Conversely, centre fire concentrates the priming material into the centre of the base of the cartridge leading to this centre being the most susceptible to ignition

Figure 1.1: Dissected view of the case [8]

Lands and grooves are often marked onto fired projectiles as they spiral out of the muzzle of the firearm These physical striations can help in the physical characterisation of the projectile through the use of a comparison microscope The shape, cannelures, dimensions of the hollow point and coating can all be discriminating identifiers of the projectile origin It is known that the inorganic composition of the projectile is usually lead, or lead with antimony added as a hardener[2,4,5,9]

This figure is not available online

Please consult the hardcopy thesis available from the QUT Library

1.2.3 The Propellant

The most important part of the ammunition in relation to analysis is the propellant The propellant occupies considerable space inside the cartridge case and contains various compounds For 0.22 calibre ammunitions, smokeless powders are the propellant of choice[1] These powders come in two varieties, single and double based Single based powders are those that contain nitrocellulose (NC) as the main explosive material whereas double based powders contain both NC and nitro-glycerine (NG) The addition of NG increases the hydrophobic tendencies of the powder, raises the energy content and softens the powder The stabilizers ethyl and methyl centralite behave in a chemically similar way however, only one compound is used in the ammunition make up, never both The role of these compounds is to remove oxides formed by the decomposition of NG and NC If these oxides are not removed, they behave as catalysts and are involved in further decomposition, which will shorten the shelf life of the ammunition It has also been stated that self-ignition may occur from the increased degradation due to the auto-catalysis[10-17] of the ammunition Ethyl centralite (EC) removes oxides by acting as a weak base and reacts with the decomposition products to form nitro and nitroso derivates Diphenylamine (DPA) is another stabilizer commonly seen in single based powders Its usual concentration is only 1% and it is common to see both EC and DPA in ammunition types Its function is similar to that of EC, as its main function is to absorb the free nitrates that have derived from the nitrocellulose Its degradation and the subsequent derivatives that are formed will be discussed later in the chapter

Plasticizers are also found in propellants and work to convert NC from its natural fibrous state into a gel state Dibutyl phthalate, diethyl phthalate, dioctyl phthalate, and glycerol triacetate are common plasticizers found in ammunition today[9,18] These compounds can also function as burning modifiers which reduce the initial burning rate of the propellant grains and increase pressure and efficiency[19]

Most primers have explosive and oxidizing properties Many of the compounds found in primers can be multi or mono functional Organic compounds with nitro functional groups are often used as primers Such compounds include Trinitrotoluene (TNT) or derivatives thereof In addition to these organic compounds mentioned, primers are also comprised of inorganic elements such as barium, lead and antimony These elements play key roles; for instance, an initiator (lead styphnate), an oxidiser (barium nitrate) and a fuel source (antimony sulphide)[2,5,9,18,20]

The primer is ignited when hit by the firing pin of the weapon which results in hot gases and temperatures being created The ignited primer now decomposes and the enormous pressure and energy build up consequently causes the projectile to be expelled from the chamber of the firearm Lead, Antimony and barium are converted into a gas during this process which in turn condenses into tiny spheres or droplets of residue These spheres can range in size from 0.5 to 10 microns, making them a valuable tool in forensic science

drug analysis, which suggests it could be extended to GSR research Aebi et al[21]

has stated that GC coupled with dual MS and a Nitrogen-Phosphorous Detector (NPD) is a powerful tool for forensic analysis Their study concentrated on identifying masked pharmacologically active compounds via this method and noting its advantages over other detection systems However, they observed that not all pharmacologically active compounds contain nitrogen and phosphorous Their LQYHVWLJDWLRQRI – tetrahydrocannabinol and its metabolites (THC acid) show this This compound is the active component in marijuana or cannabis, is non volatile and therefore Liquid Chromatography (LC) is often used to establish the concentration of THC in a particular sample The reason for this is because with higher temperatures often used within a GC system, THC acid is decarboxylated to THC This subsequently results in an inaccurate estimation of the total concentration of THC in

the sample This is of significance as a small amount of THC acid decarboxylating to THC in this process has a dramatic effect on the amount of THC detected In their study, while not covering a large number of compounds, did show the possible advantages of using GC-MS in organic studies of this nature GC allows for rapid and sensitive separation of compounds while MS provides identification of the resulting peaks from the chromatogram By using mass spectral libraries, unknown substances can usually be identified This in turn has allowed GC-MS to be a valuable tool in forensic analysis and will continue to do so in the future, with many studies utilising this analytical technique[22-25]

Within the field of unfired propellant analysis, the organic compounds used for identification have not changed significantly over the years, however, major discrepancies exist over which of these organic compounds are totally ‘unique’ to smokeless powders

Mach undertook two studies in 1977[26] and 1978[27] of 33 different kinds of ammunition to determine the feasibility of identifying gunshot residue via its organic constituents using GC-MS From their earlier study[26] they were not able to predict the composition of the powder just by its brand and type Their later study[27] agreed

with Wu et al[28,29] that the main characteristic compound is ethyl centralite but it

is present in comparatively small amounts Mach’s work did not extend to discriminating different batches of ammunition Their main focus was on comparing the unburnt particles to burnt ones The results indicated the components detected were of varying concentrations It could have been useful, therefore, to establish a database of the concentrations of the various components detected Unknown samples could have been compared to this database to establish whether it had similar properties to the samples of known origin in the database (i.e a profiling method) Interestingly, their results showed that diphenylamine was the most common additive, and while this is not a new discovery, the lack of investigation into other organic compounds in the samples would indicate their method is in need of further exploration Dibutyl phthalate was seen in about half of their samples but nitro-glycerine was seen in a substantial number of samples, which suggests that their work was paving the way for further investigations into this area

Wu et al[29] used a tandem mass spectrometry system (MS-MS) to show that the

organic components of GSR are more characteristic than their inorganic counterparts MS-MS has the added advantage of being more selective in that molecular ions are separated by the first mass spectrometer, and these subsequent selected ions are reanalysed by a second mass spectrometer to give a more specific pattern It was stated that the main components of the gunshot residue are NG, trinitrotoluene (TNT), 2, 4-dinitrotoluene (DNT), DPA and methyl centralite (MC) In their research they were able to separate the main components of the gunshot residue by High Performance Liquid Chromatography (HPLC) and observe the molecular ions of the main stabilizer methyl centralite, which are said to be the most characteristic compounds of gunshot residue[28] The components were analysed by HPLC-MS and the compounds eluted in the following order (table 1.1):

Table 1.1: Elution order for constituents using HPLC-MS[29]

(OXWLRQRUGHUIRUFRQVWLWXHQWV

When the compounds were then subjected to MS, characteristic ions were seen This however, was done using negative ion mass spectrometry which is not the usual methodology The Molecular ion (M-H)- at m/z 226 was seen with fragmentation

ions: (M-HNO3)- at m/z 163 and (NO3)- at m/z 62 These ions are said to be

characteristic of NG[29] The characteristic molecular ion of MC was seen at m/z

241 and the fragmentation ions at m/z 134, 106, 77, 51 While it has been suggested that the various components of the smokeless powder do have other origins, MC and

EC are thought to be characteristic of smokeless powders especially when seen with NG[28] These two compounds are uncommon in other industries NC, on the other hand, is used in lacquers, varnishes, printing and in the pharmaceutical industries while DPA is used in rubber preparations and in the food industry NG has uses in the explosive and pharmaceutical industries[28]

This table is not available online

Please consult the hardcopy thesis available from the QUT Library

Wu et al [29] did however suggest that by focusing their studies on methyl centralite

and using concentrated sample volumes, the level of contamination of the injector contributed to lower than expected detection levels Several cleaning methods did not help the situation and so, they adopted a 100 fold dilution of the samples with methanol to avoid further contamination Once this method was adopted, they were able to successfully discriminate 10 persons who had recently fired a revolver, from

10 persons who had not fired a gun Further experiments could adequately identify

MC on a person who had fired a pistol eight hours earlier and on the hand of an individual who had washed their hands after firing a pistol Both tests showed a positive correlation and they argued that their work could be used in criminal investigations when an offender managed to escape after firing a firearm, but was caught several hours later They proposed that even if the offender washed their hands with water, detectable amounts of MC could be identified using this method This method was found to be sensitive and selective when using Multiple Reaction Monitoring (MRM) mode, which used ions of m/z 241 as precursor ions and ions of m/z 134 as product ions, and while their method mainly focused on methyl centralite,

it could be adapted to identify other components, and therefore is a promising technique

While Wu et al [29] concentrated on MC as the most characteristic compound for the

identification of smokeless powders; DPA and its derivates have received wide

attention It is known to decompose substantially[10-17,30-35] and Tong et al[17]

suggested that the ‘detection of DPA may be taken as evidence’ as many gun powders do not contain the stabilizers methyl or ethyl centralite However, it should

be noted that DPA is also found in the rubber and food industry

Bergens and Danielsson[10,11] investigated the degradation mechanism of DPA and their results are shown above in figure 1.2, however, this is in contradiction to the report of Lussier[14] who suggested other reaction mechanisms Bergens and Danielsson concentrated on monitoring the DPA at a temperature of 85ÛC to note the breakdown products achieved The conclusion was after analysis by LC, only a small percentage of the DPA is converted to 2-nitro-DPA and 4-nitro-DPA Conversely, N-nitroso DPA corresponds to the largest breakdown product produced, which in turn is responsible for further interaction with the NC degradation products After 15 days

of storage at this temperature, the concentration of the 4-nitro-DPA is about twice as much as 2-nitro-DPA which is supported by the work by Lussier and Gagnon[14] The concentration-time curves further support the idea that the DPA is rapidly degraded into these products when introduced to a NC matrix

Concentration anomalies of DPA were observed in another study[14] which encountered difficulties during extraction stages; however in the study, critical evaluation of their extraction and solvent techniques have rendered the paper quite useful The cause of the significant DPA loss was not fully explained, and while it is

of extreme importance to understand the chemistry of the reactions, the authors have suggested that the interaction between the NC matrix and DPA has produced a compound not amenable to extraction techniques Methylene chloride was used suggesting that most of the NC matrix should have been removed during the extraction stage It does then beg the question of what DPA is reacting with to create degradation products The authors gave many possible explanations including the interaction with nitrate esters, causing cleavage to leave alkoxy radicals, which are said to be responsible for rupture of the cellulose chain Furthermore, the NC-DPA structure that is then formed could react with nitrous oxides and N-nitroso-DPA, with

2 and 4-nitro-DPA possibly being by-products of this interaction In part 2 of the study[10] it was suggested that spectroscopic techniques be employed to determine whether or not the DPA has been incorporated into the NC matrix While more complex reasons were suggested for the DPA degradation, it is important to note the possible interaction with oxygen as a degradation source[11] The authors did not give reasons behind the reaction between DPA and oxygen nor its interaction with

light It may be worth investigating these two physical properties on the decomposition of DPA

Mathis et al [36] further highlighted the importance of developing an analytical tool

for the characterisation of GSR Their study involved seven compounds; DPA, nitroso-DPA, MC, EC, dimethylphalate, diethylphthalate and dibutylphthalate

N-(DBPH) Mathis et al understood the necessity of producing a chemical profile that

allows for determination of distinguishable characteristics While they acknowledged the use of GC as a tool for the analysis of GSR, the study involved the use of gradient reversed phase liquid chromatography as it was felt that the nitro compounds were susceptible to thermal degradation by GC-MS This contrasts with

the work of Burns et al[22] who used GC-MS to analyse and characterise NG based explosives Burns et al successfully identified nitro esters and nitro aromatic

compounds from the samples and undertook the important task of examining any batch-to-batch differences between commercial explosives Determination of the sample’s total DNT content allowed for further discrimination of the ammunition type This information was compared to the manufacturer’s specifications, which enabled the explosive batch number to be established

Despite the contradiction, Mathis et al were able to quantify several organic

constituents such as centralites, phthalates and diphenylamine by gradient reversed phase LC This phase was chosen due to the spread in polarities of the main constituents Coupled with MS, this technique is very powerful in both separating and identifying capabilities The separation method was further developed by

Wissinger et al in 2002[37] however slight changes were made to accommodate

Mathis’ study Ammonium acetate (CH3COONH4) was used to assist the ionisation process Methylene chloride as a an extraction solvent was also used as its application is well documented in OGSR[10,11,25,37] Its role is to keep the NG insoluble, which enables ease of removal upon reconstitution The LC conditions were modified compared to those used by Wissinger by using a lower flow rate, smaller column and by adding ammonium acetate to the mobile phase to help in the ionization stage The study was significant to GSR analysis as it successfully discriminated characteristic organic compounds; however, the ambiguity of the

conclusions reduces the overall value of the work Wissinger did comment on the power of GC-MS, particularly with respect to the analysis of phthalates which may have been missed by LC-MS

In 1989, Keto et al[18] compared smokeless powders using Pyrolysis Capillary Gas

Chromatography This method was chosen since there is no sample preparation involved and consequently any error associated is eliminated The authors also argued that by using the capillary column, the better separation would increase the

analytical benefit Meng et al [20]agreed with this statement by saying that the

‘separating power of the capillary column in the GC is unparalleled’ However, by using statistical measurements, the amount of detectable difference between manufacturers was shown to be very small, rendering this method limited in source identification This is a situation where the use of MS could be of significance to properly identify the peaks in the chromatogram While no real differences seen between different manufacturers, close examination of the results showed different levels of peak intensity, which was not appropriately recognised in the paper The size of the study was limited to only three samples: Hercules Red Dot; Winchester-Western 540; and Dupont Hi-Skor 700x brands This factor could be the reason for their poor results Many brands and types of ammunition should have been included, especially when their statistical method of validation is a chemometric method, which requires a large sample population

In a survey of GSR analysis by Meng et al[9], efforts were made to compare the

various techniques available for the study of GSR They incorporated both inorganic and organic analysis From this report, it is clear that there is a variety of techniques available for analysis of the organic components, whereas in inorganic analysis there are only four main instruments of choice: Atomic Absorption Spectrometry (AAS); Anodic Stripping Voltammetry (ASV); Neutron Activation analysis (NAA); and Scanning Electron Microscopy coupled with the Energy Dispersive X-Ray detector (SEM/EDX) When analysing the organic components, there is a greater variety of methods to choose from Such techniques include Thin Layer Chromatography (TLC); Gas Chromatography (GC); Infrared Spectroscopy (IR); High Performance Liquid Chromatography (HPLC); Mass Spectrometry (MS); Fluorometric detection;

Supercritical Fluid Chromatography (SFC) and Capillary Electrophoresis (CE) The use of Raman Microscopy has also been noted in Organic GSR analysis[38] More recently, the use of time-of-flight secondary ion mass spectrometry (TOF-SIMS) has been noted in the study and characterisation of propellant samples[39,40] This analytical technique has the advantage of obtaining elemental and molecular information from surface samples with a particularly low level of detection It can also capture molecular images which may be useful for investigating distribution of additives and explosives constituents of the powder

TLC has the advantage of being simple, rapid, moderately sensitive and relatively cheap However, it has poor quantification capabilities and requires a large amount

of sample Infrared Spectroscopy (IR) has been successfully used to determine the presence of nitrocellulose in samples[41] However, it was not as successful in determining the other minor components present at low concentration, but which are equally important in function Such components include the stabilizers in the propellant grains, which may be used as characteristic compounds of GSR Furthermore, the fact that only NC (not being totally unique to GSR) was accurately detected in this study may render this technique unsuitable for the analysis of GSR,

as there are many other vital components, which have been overlooked It may be useful for the confirmation of substances when a positive result using another method such as HPLC, has been obtained

HPLC can be used for the analysis of organic compounds and has the added advantage of being able to separate ionic compounds, polymeric materials and poly-functional compounds of high molecular weight Also, HPLC is not limited by thermal stability of the compounds Due to the higher temperatures in the GC instrument, some compounds cannot be adequately analysed as they may undergo decomposition during injection

The MS system is a highly specific and sensitive method and has previously been shown to be a powerful tool in analysing explosive compound[28,29] Mass Spectrometry is an excellent analytical system which is best coupled with the GC or HPLC systems to give both separation and identification of the compounds in a

mixture It has been proposed that the coupling of MS with GC will render the best results since both operate in the gaseous phase and the separated components can sequentially flow to the MS detector where the individual mass spectra can be obtained[21] Most investigations into the organic compounds of GSR have coupled the MS with HPLC but the use of GC is equally as valid

Fluorometric detection has been noted as a sensitive and selective mode of detecting

organic components when using HPLC Meng et al[20] in 1996 investigated the

detection of ethyl centralite and 2-4 DNT in GSR after derivatisation with fluorenylmethylchlorformate Their method used a fluorometric detection which they claimed to be a more sensitive and selective method for the detection of organic GSR EC was first hydrolysed to N-ethylaniline (NEA), which was further derivatised with dansyl chloride (DNS-Cl) The end product was a fluorescent compound that could be separated using TLC or reversed phase HPLC While the authors claim that fluorometric detection is sensitive and selective, their analysis revealed that only three out of eleven kinds of ammunition contained EC They analysed three other compounds recommended as characteristic compounds[35] These were 2,4 DNT, 2–nitro diphenylamine, and 4–nitro diphenylamine It is suggested that they can be reduced to their aromatic amines which then allows for derivatisation using the labelling agents such as 9-fluorenylmethylchloroformate (FMOC) Interestingly, the use of diphenylamine derivatives is in complete contradiction to their previous statement in 1994[42] where they stated that, ‘DPA has been regarded as an evidentially irrelevant compound’ The authors were able to achieve good detection levels (1 pg levels) which is lower than their previous detection limits of 60pg using DNS-Cl[42] However, only six out of eleven samples were shown to contain EC or 2, 4-DNT or a combination thereof Their method failed to detect the other two compounds previously mentioned to be more characteristic than EC and DNT The authors suggested that while the method may

9-be sensitive, further analysis of the reduction, hydrolysis, derivatisation and fluorescent steps could ultimately improve the technique

The use of Supercritical Fluid Chromatography (SFC) has not been extensive in the area of organic gunshot residue analysis; however it has been reported that its

application is suitable for thermally labile or non-volatile substances An advantage

of SFC is that it is compatible with virtually all detectors used in established

chromatographic techniques such as GC and HPLC In a study by Munder et al,

smokeless powders were analysed using this technique and while they were able to detect several minor constituents such as ethyl centralite, diphenylamine, dinitrotoluenes and dibutyl phthalate, they were not able to distinguish between different brands The possibility then of classifying smokeless powders by manufacturer using this technique is low

An interesting example of characterising black powders is the study by MacCrehan

et al in 2002[43] who looked at associating gun powders and residues by compositional analysis They identified the main constituents of the powder to be nitrocellulose, nitro-glycerine, diphenylamine and ethyl centralite Characterisation

of the ammunition type was achieved by calculating a propellant to stabiliser ratio (P/S ratio) developed by Reardon et al which is a simple method to use and interpret The study collected 7 brands with only 50 rounds in each Each of the cartridges were assigned numbers from 1 – 50 and the, numbers 1, 10, 20, 30, 40, 50 were assigned codes The powder was removed from the cartridge and placed in a vial which could only be traced back to this code This method then allowed for a random choice for the order of analysis Ultrasonic Solvent Extraction/Capillary Electrophoresis was their method of choice and while it was adequate for detection

of certain compounds, some brands were shown to contain diphenylamine only This observation is unusual as powders are comprised of various substances such as propellants, hardeners, stabilisers and so on and therefore detection of these would be expected To fully appreciate the characteristic associations, a more sensitive and selective analytical procedure should be adopted so that a characteristic profile for

each cartridge could have been developed Furthermore, as suggested by Smith et

al[44], electrophoresis should only be a complementary tool as results can vary according to different parameters The study however shows promise in selectivity and sensitivity but compared to other analytical techniques, it appears that this method is somewhat lacking MacCrehan’s study is an example however, of how important it is to characterise and discriminate powders from each other This can

only be of benefit for investigating authorities who rely on validated databases with which to compare their unknown samples

Furthermore, in an important recent study by MacCrehan[45] an attempt was made to create a smokeless powder reference material for laboratories undertaking explosive and propellant analysis The aim was to characterise the organic additives commonly found in smokeless powders such as nitroglycerin, ethyl centralite, diphenylamine and the nitration product N-nitrosodiphenylamine This would be achieved using the analytical techniques Micellar capillary electrophoresis (CE) and Liquid Chromatography (LC) Three candidate smokeless powders were studied Powder 1 was Hi Skor 700X (an NG containing powder stabilized with EC) Powder 2 was Winchester 231 (an NG powder stabilized with DPA) and handgun reloading smokeless powders purchased from a local gun shop Powders 1 and 2 became the inter laboratory comparisons of smokeless powders analysis and measurement and interestingly, MacCrehan noted a number of inconsistencies from the 20 participants Many laboratories reported surprising results with some identifying EC in powder 2 which was stabilized by DPA They report that manufacturers often use recycled surplus of materials and this material may have different compositions than intended These trace compounds and inhomogeneity of the propellant lead to uncertainty of the true make up and chemical composition of the smokeless powder For this reason, the commercial powders were deemed unsuitable for reference materials However, the use of Powder 3 (the handgun reloading smokeless powder) was explored and through a series of experiments and involving thermal stability and homogeneity, it was deemed that Powder 3 was suitable for reference materials with the use of LC The testing revealed that storage at room temperature or below was acceptable and sampling sizes should be 20 mg or greater These conclusions were supported by the fact that relative uncertainties for the additives were much smaller with room temperature stored samples and of sizes 20mg or greater The study was able to provide a reference material from Powder 3 with a level of uncertainty <5%

As this is a relatively new study, further follow up inter laboratory analysis and testing should be carried out to support the long term use of such reference material

Electrophoresis also has been applied to organic gunshot residue analysis [44-47] Northrop completed a two part study[46,47] which concentrated on 13 compounds including derivatives of the main organic constituents This work was an extension of previous work undertaken in 1992[48] which examined sampling protocol for Micellar Electrokinetic Capillary Electrophoresis (MECE) Results are shown in Table 1.2

Table 1.2: Results from Northrop[46,47]

Compound Detection Levels (pg)

2,4 DNT 1.1 2,6-DNT 1.2 3,4-DNT 1 2,3-DNT 1.3 Diphenylamine 0.9 2-Nitrodiphenylamine 1.9 4-Nitrodiphenylamine 2.1 N-nitroso-diphenylamine 1 Dibutyl Phthalate 2.6 Diethyl Phthalate 2.2 Ethyl centralite 1.8 Methyl centralite 1.1 Nitro-glycerine 3.8

Zeichner et al[25] argued that MECE with diode array UV detector is not sensitive

enough for crime scene comparison however Northrop used this tool to develop a method for separating organic constituents of gunpowder Northrop argued that the phthalate esters commonly seen in OGSR analysis are not unique to gunpowder and have not been included in his study While this may be the case, the detection of phthalate esters is a confirmatory tool in the identification of smokeless powders and their inclusion is both important and significant to the ammunition makeup Part two

of the study[46] concentrated more on the direct examination and characterisation of the main organic compounds of GSR 16 out of 26 samples taken from the back of the right hand after firing under controlled conditions, showed no detectable GSR The other samples were found to contain mostly nitro-glycerine, diphenylamine, and N-nitroso diphenylamine Two samples were shown to contain ethyl centralite and only one sample was shown to contain all components (nitro-glycerine, diphenylamine, N-nitroso diphenylamine, 2-nitrodiphenylamine, 4-nitrodiphenylamine and ethyl centralite) When a 0.22 calibre weapon was used, no

detectable OGSR was found It was suggested that re-testing of other 0.22 weapons would be of use to determine whether the lack of OGSR is due to the size of the weapon or ammunition, or because of the configuration of the weapon This assumption is flawed as no attempt was made to discover if the analytical techniques and recovery procedures played a significant role here Furthermore, it was suggested that it may be possible for OGSR not to be deposited on skin surfaces, or that the weapon characteristics prevent detectable OGSR from being deposited after each firing Further investigations into more reliable techniques could dramatically improve the research However, this study has demonstrated its limited use in forensic casework which relies on validated analytical procedures for reliable conclusions

The important paper by Wrobel et al[1] emphasizes the necessity for a method that

combines all information pertaining to the organic make up of smokeless powders Their work indicated that 0.22 calibre ammunition is the most commonly encountered projectile in Australia and it appears that this trend has not altered The identification of the ammunition was carried out using both physical and chemical characteristics, which generally were from an inorganic source A database was created that consisted mainly of SEM/EDX spectra, which can be compared directly from materials gathered from a crime scene The database was useful in identifying the cartridges from crime scenes, since physical characteristics were being included into the database However, the conclusions were limited due to the project consisting mainly of inorganic substituents which unfortunately do not vary substantially between types

A database consisting of organic substituents could in fact be beneficial to investigating personnel By combining the powerful separating and identification qualities that GC-MS possess, a database of considerable importance could be created which ultimately should be able to differentiate the type of ammunition by the type and quantity of organic matter present It is vital that even though a database could be implemented, it must be constantly reviewed with the changing times and

possible new ingredients This idea is supported in the study by Collins et al[49]

where recent research has shown that glass containing particles in gunshot residue

can be analysed and characterised Even though their study concentrated on the inorganic components of GSR, it does raise the important point that new discoveries are constant and analysts need to be aware of this

Organic constituents can be more characteristic and by using GC-MS, reproducible and repeatable results can be obtained Due to the lack of validated GC-MS methodology, this work will focus on establishing a tool for the forensic investigation of gunshot residue Previous studies have shown that GC will not be the ideal instrumentation to analyse non-volatile substances However, this can be overcome by ignoring the major nitrocellulose matrix and maintaining lower port temperatures By doing this, it can determine the presence of the major organic compounds by a sensitive and selective analytical tool An instrument such as GC-

MS readily fits this description It is important to distinguish which compounds are unique to smokeless powders and by intercalating the literature assessed in this survey, decisions about these compounds can be made

As previously mentioned, the 0.22 calibre ammunition is the most commonly encountered ammunition type recovered at a crime scene This study will therefore focus on using this type of ammunition Furthermore, investigations into proper solvent choice, developing an extraction technique to remove the nitrocellulose matrix, will be established To fully appreciate the possibility of matching or discriminating propellant to a box or batch of ammunition, it is imperative that the propellant is examined for homogeneity Also, as previous papers have indicated, oxygen plays a significant role in the degradation of diphenylamine, it may therefore

be important to investigate the effects of removing oxygen from the experiments The ultimate aim will be to determine whether it is possible to match a propellant sample back to its batch of manufacture

 ([SHULPHQWDO

The standard compounds DPA (99% purity), EC (98% purity) and DBPH (99% purity) were obtained from Sigma-Aldrich (USA) HPLC-grade ethyl acetate and

dichloromethane (DCM) were obtained from Selby (Brisbane, Australia) The

propellant samples were obtained from the Ballistics section of the Queensland Police Service (Brisbane, Australia) The GC vials and 300µL glass inserts were obtained from Biolab (Brisbane, Australia)

The GC-MS consisted of a Hewlett Packard HP6890 Series GC system and 5973 Series Mass Selective Detector (Agilent Technologies) equipped with an auto sampler and injector Chromatography was achieved using a DB-1 capillary column (0.25mm x 30m x 0.25µm) using a pulsed splitless injection technique at one ml/min The oven initial temperature was set at 50 degrees for one minute with an equilibration time of 0.2 minutes A rate of six degrees per minute was selected to reach a maximum of 270 degrees The front inlet had a temperature of 250 degrees with a pulse pressure of 40 psi The MSD transfer line heater was set at 280 degrees 1µL of sample was injected into the GC with a total run time for each sample being 45.7 minutes EI was used over a mass range of 29-300 amu and an operating voltage

of 70eV MS Chemstation software (Hewlett Packard) was used for automation and data analysis Each sample was analysed under these conditions and was used for the detection of all compounds except nitro-glycerine Unless stated otherwise, NIST library was used for compound identification

To analyse nitro-glycerine, another pulsed split-less technique was used The same operating parameters as previously described were used; however, the initial front inlet temperature was set at 150 degrees and a sampling volume of 2µL was used As

all samples analysed in this study were double based in nature (i.e it contained both

NG and NC), it was decided that analysis of NG would not be efficient

To gain an understanding of the instrumental variation, standard solutions for DPA,

EC and DBPH were made This was done by dissolving in volumetric flasks 5mg in

50 ml ethyl acetate (0.01% w/v) for DPA and 100mg in 100ml ethyl acetate (0.1% w/v) for EC and DBPH Three stock solutions containing DPA was made and ten stock solutions for EC and DBPH were made An aliquot from each stock solution was taken and analysed Due to limited availability of DPA, three stock solutions could only be made

Three random rounds of ammunition were obtained from Winchester 0.22 Expediter with batch number ADD1BGH2 50mg of the propellant from each projectile was accurately weighed and dissolved into three separate volumetric flasks with 5ml of ethyl acetate in each An aliquot from each volumetric flask was taken and placed in

a 2ml GC vial labelled sample 1, 2 and 3 Each of the vials was analysed three times

to gain an understanding of the variation caused by this dissolution method

a 300µL glass inserts Each of the vials were analysed three times to gain an understanding of the variation caused by the extraction technique

Six different types of Winchester brand 0.22 ammunition were selected; (1) Winchester laser Long Rifle High Performance with batch number 2DRM41, (2) Winchester Expert with batch number 23DLH02, (3) Winchester winner with batch number IDKE52, (4) Winchester subsonic long rifle rim fire with batch number AED1FH31, (5) Winchester super speed long rifle high velocity solid with batch number 2DSB51 and (6) Winchester super speed long rifle high velocity high performance with batch number 2DLRL62 From each type of ammunition listed two projectiles were randomly selected from each and analysed as per conditions described in the extraction procedure

The selected target compounds are listed in table 3.1, with their major ions These compounds were diphenylamine (DPA), N,N-Diethylcarbanilide (Ethyl centralite, EC) and dibutyl phthalate (DBPH) Due to the fact some components are rare, and/or are explosive in nature with associated health and safety issues, nitro-glycerine and nitrated derivatives of DPA could not be obtained commercially As a consequence, analysis was concentrated on DPA, EC and DBPH

Table 3.1: Selected target compounds (from NIST library)

Major ion data for standard compounds

organic radicals Amines, like ammonia, are weak bases because the unshared electron pair of the nitrogen atom can form a coordinate bond with a proton Amides can be produced by reacting amines with acids anhydrides or esters Furthermore, amines react with acids to give salts or reaction with halogenated alkanes can occur

to form longer chains

Many amines are not only bases but also nucleophiles that form a variety of substituted compounds Due to their basic chemical functionality, they are important intermediates for chemical syntheses with substitution occurring at the nitrogen atom

Aromatic amines also exist, such as phenylamine and benzylamine, which dissociate

in water (some very weakly) Aromatic amines are much weaker bases than aliphatics The term benzyl describes the radical, ion or functional group C6H5CH2-, derived by the methyl group of toluene losing one hydrogen atom Phenyl on the other hand is the term for the monovalent radical C6H5-, derived by the removal of a hydrogen atom from a benzene ring One of the most important aromatic amines is aniline, a primary aromatic amine replaced by one hydrogen atom of a benzene molecule with an amino group

Diphenylamine or N-phenyl aniline is a highly reactive secondary aromatic amine which undergoes electrophilic aromatic substitution (for example in the presence of nitrogen oxides) which is heavily activated by the amino function (which is an otho/para director) and the electron rich benzene ring

The mass spectrum for commercially available diphenylamine is shown in figure 3.1

Figure 3.1 Mass spectrum of diphenylamine

The mass spectrum of DPA is identifiable by the characteristic m/z 169 ion which is the molecular weight of DPA The next major ion is m/z 77 which relates to the aromatic residue found in DPA (C6H5 = m/z 77)

Figure 3.3: Ethyl Centralite

Figure 3.4: Mass spectrum of ethyl centralite

The mass spectrum (figure 3.4) of ethyl centralite shows a characteristic ion at m/z

268 which corresponds to the molecular ion Again, m/z 77 ion is observed which is indicative of the phenyl ring The m/z ions of 148 and 120 indicate a cleaved bond between carbon and nitrogen to form radical ions This is illustrated in figures 3.5 and 3.6

N

C2H5C

O N

C2H5

Figure 3.5: Fragmentation ions of ethyl centralite

C2H5

N

C2H5C

Figure 3.7: General structure of phthalate esters where R, R’ = CnH2n+1; n=4-15[50]

Phthalate esters are a group of plasticizers which are made by the reaction of alcohol with phthalic anhydride which results in a stable compound Its use in propellant is primarily to convert NC from its natural fibrous state into a gel (treated with solvent first which is later evaporated) Furthermore it provides flexibility to the powder grains

Phthalate esters not only exhibit similar chemical properties, the mass spectra for this group of compounds is remarkably similar, with the indicative ion m/z 149 displayed Diagnostic ions are shown in figure 3.8 which show the chemical fragmentation of butyl phthalate esters

C

OH O

C

C O

O

OC 4 H 9

m/z = 149 m/z = 205 m/z = 222 (where R and R‘ = butyl) m/z = 223

Figure 3.8: Characteristic fragmentation ions of butyl phthalate esters [51]

As previously stated, the characterization of phthalate esters is usually relied upon the identification of m/z 149 However, for this reason it is imperative that correct identification of each phthalate ester occurs In the instance of dibutyl phthalate, it is important to consider the higher mass range ions which are usually low in abundance and readily overlooked Dibutyl phthalate has two indicative ions in the higher mass region; being m/z 205 and m/z 223[50] Figure 3.9 shows the mass spectrum of dibutyl phthalate and its chemical structure

3.2.1 Diphenylamine, ethyl centralite

and dibutyl phthalate variation

Table 3.2 DPA, EC and DBPH standard variation