. . s . .. .. A z.- V .2 3...?! .. .ro: . 14.4 3.. .... An? ‘ "“unq" ‘Jqu... w a a v ‘ :vvh ,qn A . ., ... , 3.22;? Lutumwfi “rm ~ \ if , LIBRARY 2 ‘ Michigan State University This is to certify that the thesis entitled ELECTROSPRAY IONIZATION MASS SPECTROMETRY FOR THE DETECTION AND CHARACTERIZATION OF SMOKELESS POWDER presented by Gwynyth Scherperel has been accepted towards fulfillment of the requirements for the MS. degree in Criminal Justice MM Major Profess‘oF’s'Signature o’l AUGUJT 2907 A Date MSU is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/07 p:/ClRC/DateDue.indd-p.1 ELECTROSPRAY IONIZATION TANDEM MASS SPECTROMETRY FOR THE DETECTION AND CHARACTERIZATION OF SMOKELESS POWDERS BY GWYNYTH SCHERPEREL A THESIS Submitted to Michigan State University In partial fulfillment of the requirements For the degree of MASTER OF SCIENCE Department of Criminal Justice 2007 ABSTRACT ELECTROSPRAY IONIZATION TANDEM MASS SPECTROMETRY FOR THE DETECTION AND CHARACTERIZATION OF SMOKELESS POWDERS BY GWYNYTH SCHERPEREL Smokeless powder is one of the most common types of explosives used today in civilian ammunition and because of this it is also commonly found in improvised explosive devices. Thus, the detection of smokeless powder in the form of unexploded powder or residue from a gunshot or post-blast explosion can be of great forensic value. Tandem mass spectrometry (MS/MS and MS") in a quadrupole ion trap with electrospray ionization was examined in order to develop a method for the routine, rapid identification and comparison of smokeless powder. This method was optimized for the simultaneous detection of the smokeless powder stabilizers methyl centralite, ethyl centralite, and diphenylaminc and it was found that good limits of detection could be achieved for all three stabilizers. Seven smokeless powder samples and their residues were then analyzed as ‘real world’ samples. The use of tandem mass spectrometry and selected reaction monitoring increased the sensitivity and selectivity, respectively, of this technique for both the unburned and burned powders, allowing for the identification of trace components. Mass spectrometry was able to identify all of the unburned powder samples as being smokeless and it was able to identify gunshot residue at firing distances of 3" and 12" on unwashed, hand washed (in deionized water), and machine washed cloth. All but two of the unburned powders could be differentiated from each other by combining the mass spectrometry data with the physical characteristics and extraction tendencies of the smokeless powders. C0pyright by GWYNYTH SCHERPEREL 2007 ACIGVOWLEDGMENTS This project would not have been possible without the support of many people. First, I would like to thank my research advisor, Dr. Ruth Waddell, for her guidance, encouragement, and support. I would also like to thank my committee members, Dr. Gavin Reid and Dr. Vince Hoffman. A special thanks goes to Dr. Reid for allowing me to use his lab equipment, chemicals, and space, without which this project would not have been possible, and for all of his guidance and support throughout my graduate studies. This project could also not have been completed without the help of the certified firearm instructor employed by the Michigan State University Police Department who provided all of the ammunition and firearms used and who allowed us access to the Michigan State Police firing range where he was gracious enough to fire the bullets, providing us with an unending supply of gunshot residue. I would also like to thank previous and current group members of the Reid research group and the forensic chemistry graduate students for their friendship, help, and motivation. Finally, I would like to thank my parents and friends who endured this long process with me, always offering support and encouragement. iv TABLE OF CONTENTS LIST OF FIGURES ................................................................................................... LIST OF TABLES ............................................................................ LIST OF SCHEMES ......................................................................... 1. CHAPTER ONE: Introduction ........................................................... 1.1 Smokeless Powder .................................................................. 1.2 Current Applications of Smokeless Powder .................................... 1.2.1 Ammunition .................................................................... 1.2.2 Improvised Explosive Devices (IEDs) ....................................... 1.3 Manufacturing and Components of Smokeless Powder ........................ 1.4 Current Methods for the Forensic Detection and Analysis of Smokeless Powder .............................................................................. 1.4.1 Gunshot Residue (GSR) ..................................................... 1.4.1.1 Color Tests ........................................................... 1.4.1.2 Scanning Electron Microscopy/Energy Dispersive X-ray Analysis (SEM/EDX) .............................................. 1.4.2 Improved Explosive Device (IED) Post-blast Residue .................. 1.5 Developing Methods for the Detection and Analysis of Smokeless Powder by Identifying Organic Components .................................. 1.6 Aims of this Thesis .................................................................. 2. CHAPTER TWO: Mass Spectrometry ................................................... 2.1 Mass Spectrometry .................................................................. 2.1.1 Ionization ........................................................................ 2.1.1.1 Electrospray Ionization (ESI) ........................................ 2.1.1.2 Matrix-assisted Laser Desorption Ionization (MALDI)... . 2.1.2 Mass Analyzers ................................................................. 2.1.2.1 The Quadrupole Ion Trap Mass Analyzer ......................... 2.1.3 Detectors ....................................................................... 3. CHAPTER THREE: Experimental ....................................................... 3.1 Materials ............................................................................... 3.2 Standard Solutions of the Smokeless Powder Stabilizers MC, EC, and DPA for Analysis by Mass Spectrometry ................................................... 3.3 Smokeless Powder Samples ........................................................ 3.3.1 Stereomicroscopy of Smokeless Powder Samples ........................ vii ix 13 13 14 14 16 16 17 25 27 27 28 28 28 3.3.2 Extraction of Stabilizers from Smokeless Powder Samples for Analysis by Mass Spectrometry ......................................................... 3.4 Gunshot Residue (GSR) ............................................................ 3.4.1 Collection of GSR ............................................................. 3.4.2 Extraction of GSR from Cloth for Analysis by Mass Spectrometry... 3.5 Mass Spectrometry .................................................................. 4. CHAPTER FOUR: Nanoelectrospray Ionization Tandem Mass Spectrometry for the Analysis and Identification of Ethyl Centralite, Methyl Centralite, and Diphenylamine .............................................................................. 4.1 Introduction ........................................................................... 4. 2 Development and Optimization of Conditions for the Detection of MC, EC, and DPA.. 4. 3 Analysis of the Mass Spectra Obtained for MC, EC, and DPA Standards 4.4.3 DPA ............................................................................. 4.4 Limits of Detection and Calibration Curves ...................................... 4.5 Summary ............................................................................... 5 CHAPTER FIVE: Nanoelectrospray Ionization Tandem Mass Spectrometry for the Analysis of Smokeless Powder and its Residue... . 5.1 Introduction ........................................................................... 5.2 Physical Characteristic of Smokeless Powder Samples. .............................. 5.3 Extraction Efficiency of Smokeless Powder Samples ........................... 5.4 Analysis of Smokeless Powder Samples by nESI-MS .......................... 5.5 Analysis of Gunshot Residue by nESI-MS ....................................... 5.6 Summary ............................................................................... 6. CHAPTER SIX: Conclusions and Future Directions ................................... 6.1 Conclusions ........................................................................... 6.2 Future D1rect1ons APPENDIX ..................................................................................... REFERENCES ................................................................................ vi 29 3O 3O 30 31 32 32 34 35 38 41 46 48 51 53 53 54 56 59 73 8O 82 82 85 87 99 LIST OF FIGURES FIGURE PAGE 1.1 Components of a modern cartridge ............................................................... 2 2.1 Components of a mass spectrometer.. ........................................................... 13 2.2 Schematic of electrospray ionization (ESI) ................................................... 15 2.3 Diagram of a quadrupole ion trap. . . . .................................................... 18 2.4 Typical Mathieu stability diagram for quadrupole ion trap. The larger balls represent high mass ions whereas smaller balls represent low mass ions.. .. 21 2.5 Dehmelt pseudo well potential D (eV) as a function qz. Higher mass ions for a given V value have lower Dehmelt trapping energies ..................... 24 2.6 Diagram of a continuous-dynode electron multiplier with a conversion dynode .................................................................................. 25 4.1 Mass spectra under optimized nESI-MS conditions for standard solutions of (A) MC at 10 pmol/pL, (B) EC at 10 pmol/pL, and (C) DPA at 100 pmol/uL ................................................................................ 37 4.2 Calibration curve for EC in methanol (0.001 pmol/uL to 1 pmol/pL) using selected reaction monitoring of m/z 269 fragmenting to m/z 148 ...................................................................................... 50 5.1 Stereomicroscopy photos of smokeless powder samples ...................... 55 5.2 Mass spectra of smokeless powder 2 (9mm +P) in methanol at concentrations of (A) 0.002 mg/mL, (B) 0.02 mg/mL, (C) 0.05 mg/mL, and (D) 0.1 mg/mL extracted material ............................................. 61 5.3 Mass spectra of gunpowders (A) #1, (B) #6, (C) #3, (D) #5, (E) #2, (F) #4, and (G) #7 at concentrations of 0.1 mg/mL extracted material in methanol. 63 vii 5.4 Mass spectra for gunshot residue from smokeless powder 2 (9mm+P): (A) firing distance of 3" on unwashed cloth, (B) firing distance of 12" on unwashed cloth, (C) firing distance of 3" on cloth washed with deionized water, (D) firing distance of 12" on cloth washed with deionized water, (E) firing distance of 3” on machine washed cloth, and (F) firing distance of 12" on machine washed cloth. The inset to Figure 5.4F shows the region of m/z from 100 to 170 acquired by performing MS/MS on the m/z 269 protonated precursor ion ....................................................... 5.5 Comparison of mass spectra for unwashed cloth with and without gunshot residue. (A) Mass spectra of unwashed cloth (blank). (B) MS/MS product ion spectrum of the ion at m/z 269 in Figure 5.5A. (C) Mass spectra of gunshot residue from smokeless powder 1 (9 mm) on unwashed cloth (firing distance of 12"). (D) MS/MS product ion spectrum of the ion at m/z 269 ion in Figure 5.5C ........................................................ viii 74 78 LIST OF TABLES TABLE 4.1 4.2 4.3 5.1 5.2 5.3 5.4 MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 10 pmol/pL standard solution of MC in methanol .............................................................................. MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 10 pmol/pL standard solution of EC in methanol .............................................................................. MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 100 pmol/ttL standard solution of DPA in methanol .............................................................................. Extraction data for smokeless powder samples in methanol and in water. MS and MSn data for identified peaks in the smokeless powder samples. Stabilizers and other components detected in smokeless powder samples by nESI quadrupole ion trap mass spectrometry ............................... Stabilizers and other components detected in smokeless powder gunshot residue (firing distances: 3" and 12”) by nESI quadrupole ion trap mass spectrometry ........................................................................ A.1 Complete MS and MS" data for smokeless powder sample 1 ................ A.2 Complete MS and MSn data for smokeless powder sample 2 ................ A.3 Complete MS and MSn data for smokeless powder sample 3 ................ A.4 Complete MS and MSn data for smokeless powder sample 4 ................ A.5 Complete MS and MSn data for smokeless powder sample 5 ................ A.6 Complete MS and MSn data for smokeless powder sample 6 ................ A.7 Complete MS and MSn data for smokeless powder sample 7 ................ ix PAGE 40 45 48 58 66 72 79 88 90 91 93 94 96 98 LIST OF SCHEMES SCHEME PAGE 4.1 Possible mechanism for the fragmentation of protonated MC (m/z 241) to form product ions at m/z 134 and 106 ......................................... 39 4.2 Possible mechanism for the fragmentation of protonated EC (m/z 269) to form product ions at m/z 148, 120, and 92 ....................................... 42 4.3 Possible mechanism for the fragmentation of protonated DPA (m/z 170) to form a product ion at m/z 92 ..................................................... 47 5.1 Possible mechanism for the fragmentation of protonated N-NO-DPA to form a product ion at m/z 169 ....................................................... 69 CHAPTER ONE Introduction 1.1 Smokeless Powder One of the most common types of propellant used in civilian ammunition today is smokeless powder. Smokeless powder, as are all gunpowders, is a low explosive, meaning that it deflagrates rather than detonates [1, 2]. Deflagration is a subsonic combustion (less than 1100 fi/s) that spreads by thermal conductivity, 1'.e., burning material heats and subsequently ignites surrounding cold material. Detonation, on the other hand, is a supersonic combustion (greater than 1100 Ws) that creates shock waves that compress cold material causing its temperature to rise to the ignition point (shock compression) [3, 4]. Since smokeless powder is used for civilian ammunition and can be legally purchased for reloading, this propellant is one of the most common low explosives used to make improvised explosive devices (IEDs), such as pipe bombs [3-6]. The majority of homicides, aggravated assaults, and robberies in the United States today are committed using handguns, which consist of pistols and revolvers [7]. Thus, the detection of smokeless powder in the form of unexploded powder or in residue form from a gunshot or post-blast explosion can have significant forensic value. 1.2 Current Applications of Smokeless Powder 1.2.1 Ammunition A firearm cartridge, as illustrated in Figure 1.1 below, consists of a projectile (the bullet), a primer (explosive cap), a propellant (gunpowder), and a casing that holds everything together [8]. Bullet Bullet Case Propellant Figure 1.1 Components of a modern cartridge. (Adapted from “How Revolvers Work”. www.howstuffworks.com). Bullet designs and compositions vary by function. For maximum efficiency, 1'.e., to cause the greatest damage, a bullet should sufficiently penetrate a target to immobilize it without completely passing through it. Most bullets consist of a relatively sofi lead or lead alloy core and can be jacketed, semi-jacketed, or non-jacketed [8, 9]. A jacketed bullet has a hard metal coating, such as copper, over the lead core. This gives the bullet strength allowing it to deeply penetrate targets more. making it an ideal bullet for such applications as armor piercing. On the other hand, a non-jacketed bullet consists of only the soft lead core making the bullet structurally weaker than the jacketed bullets and therefore less likely to pass through a target such as small game [8, 9]. The caliber of a bullet refers its diameter in either mm or inches. The shape of a bullet dictates its aerodynamic and impact characteristics. The three primary shapes are solid-nosed, hollow point, and soft—nosed (or soft point) [8]. The end of a solid-nosed bullet is round and blunted and is therefore likely to stay in its original shape even after striking a target. As was the case for jacketed bullets, this is ideal for such purposes as armor piercing, but not for small game hunting where this type of bullet is apt to completely pass through a target. Hollow point bullets have a hollow cavity in the nose of the bullet that causes the bullet to expand (‘mushroom’) when a target is struck. This expansion causes the bullet to slow down inside the target. Sofi- nosed bullets fall in between these two as this type of bullet expands slower than the hollow point bullets do upon impact with a target. This causes the bullet to penetrate the target further than hollow point bullets, but less than the solid-nosed bullets [8]. The primer consists of a stable but shock-sensitive mixture that ignites when struck. The primary components are typically lead (Pb) styphnate (initiating explosive), barium (Ba) nitrate (oxidizing agent), and antimony (Sb) sulfide (fuel) [8]. When the trigger of a firearm is pulled, a series of actions takes place within the firearm that results in a firing pin hitting the back of the cartridge, igniting the primer. This in turn deflagrates the gunpowder, which expands and pushes the bullet out the path of least resistance (i.e., the barrel of the firearm). Smokeless powder has widely replaced black powder as the primary propellant (i. e., gunpowder) used in civilian ammunition due to its higher efficiency and smokeless nature, meaning that the primary combustion products are gaseous, leaving minimal residue in the barrel of the gun [3]. Thus, from a forensic perspective, one is more likely to encounter smokeless powder than black powder when dealing with firearm related cases. 1.2.2 Improvised Explosive Devices (IEDs) A pipe bomb is simply a pipe that contains an explosive material. It is a common form of improvised explosive device (IED) because of the relative ease in which the supplies can be obtained and assembled. Pipe bombs have become more prominent in the public eye lately as they have been used in many high profile cases, such as the terrorist attacks on US. soldiers in Iraq [10]. The United States Bureau of Alcohol, Tobacco, Firearms, and Explosives reported more than half of their explosive cases involved pipe bombs and about 50% of those used smokeless powder as the explosive [6]. Even though smokeless powder is a low explosive, when it is contained in a closed container, the large volume of gases created upon ignition will cause the container to burst in a relatively large explosion and release fragmented debris (shrapnel). Thus, while simple, pipe bombs can cause considerable damage, as exemplified in the Centennial Olympic Park bombing, which involved pipe bombs made with smokeless powder. Sharp objects, such as nails or screws, can also be incorporated into a pipe bomb in order to increase the amount of shrapnel, making the pipe bomb even more dangerous. 1.3 Manufacturing and Components of Smokeless Powder Smokeless powder can be widely classified as single-based, double-based, or triple-based according to the major component(s) of the powder. Single-based powder consists of nitrocellulose, while double-based powder contains nitrocellulose and nitroglycerine, which is added as a plasticizer in order to soften the propellant and raise the energy content. Triple-based powder contains nitroguanadine in addition to nitrocellulose and nitroglycerine. Triple-based smokeless powders, however, are used primarily in rockets and large caliber military grade weapons and are therefore difficult to obtain on the open market [1, 11]. Smokeless powders also contain many minor ingredients, such as flash suppressants, deterrents, opacifiers, plasticizers, and stabilizers [2]. Flash suppressants, which are commonly alkali or alkaline earth salts, help prevent secondary flash at the muzzle of the firearm [8]. Deterrents coat the exterior of the propellant granules to increase efficiency. Deterrents include such compounds as dioctyl phthalate and dinitrotoluene. Opacifiers, the most conunon of which is carbon black, enhance reproducibility of the burn rate. Plasticizers soften the propellant and reduce hygroscopicity. Examples of plasticizers include nitroglycerine, dibutyl phthalate, and dinitrotoluene [12, 13]. Stabilizers are added to slow down the decomposition of nitrocellulose by removing the nitrous and nitric acids that are produced. These decomposition products catalyze further decomposition of nitrocellulose, but if they are removed the decomposition is relatively slow [14]. Common stabilizers used today are diphenylamine (DPA), methyl centralite (MC), and ethyl centralite (EC) [12]. It is not possible to predict the composition of a powder based on the caliber or bullet type. This is because sub-batches of commercial powders are blended during the manufacturing of the powder in order to achieve the desired propellant performance and stability [3, l 1, 15]. Each batch of powder may also consist of ‘rework’; powder that did not meet specifications or was recycled [16, 17]. This blending of batches combined with the addition of ‘rework’ can lead to compositional heterogeneity (qualitative and quantitative) between commercially available cartridges [6, 18, 19]. This, however, allows for the development of distinct chemical profiles for different powders [5, 6, 20]. 1.4 Current Methods for the Forensic Detection and Analysis of Smokeless Powder Unbumed smokeless powders are initially characterized by their morphology, but powders are also distinguishable based on the presence or absence of certain compounds [21]. Thus, the physical and chemical characteristics of smokeless powder can help associate or differentiate unknown powders with residue (gunshot or post-blast) or with known powder samples [6, 11, 18, 19, 22, 23]. The agreement in composition between burned and unburned powder is generally very good [11, 18, 22], but the level of association will depend on the rarity of the composition of the powder [13]. Chemical characteristics can also be used to identify gunshot residue or to identify the type of powder used in an explosive. There have been a number of methods for the analysis of smokeless powders that have been examined. These procedures have been extensively reviewed [1 , 3] and only a brief discussion follows. 1.4.1 Gunshot Residue (GSR) Gunshot residue (GSR) is primarily a mixture of burned, partially burned, and unburned primer and propellant particles, but it can also include particles from the bullet, casing, and firearm [12, 24]. The exact composition will depend on a number of factors, including the firearm and ammunition used (type, caliber, age, etc). GSR can be deposited on any object or person that is near the gun when it is fired [8, 12]. The persistence of propellant residue on clothing, however, is considerably longer than it is on skin or hair, resulting in a higher probability of its detection and identification when it is on cloth [8, 13, 25]. The detection of GSR can help determine if a suspect has recently fired a gun, help discriminate between homicide and suicide by the location of GSR on the victim and/or suspect, induce a confession or admission, and assist in determining cause of death [12]. GSR can also help identify a shooter by comparing the composition of the GSR with that of unfired propellant found in the suspect’s possession. This is especially useful if the weapon, bullet, or both are not found or are too badly damaged for analysis [12]. The two primary approaches used today for the analysis of gunshot residue are color tests (i. 6., modified Griess test and sodium rhodizonate test) and scanning electron microscopy coupled with energy dispersive x-ray analysis (SEM/EDX) [24]. 1.4.1.1 Color Tests These chromophoric techniques aid in the visualization of gunshot residue by producing a visible color reaction with components in GSR [8, 24]. The modified Griess test (MGT) is a test for the detection of nitrite compounds [26]. Nitrite compounds are formed when smokeless powder is burned and thus can appear in gunshot residue. This test will yield a positive result if any nitrite is present. Thus, the possibility for a false positive is significant since nitrites exist is some everyday items such as deodorizers and disinfectants [13]. The sodium rhodizonate test (SRT) is used to determine the presence of lead, which can be present in GSR from the primer, from a lead bullet/barrel interaction, or from surface erosion of the bullet’s base [13]. As was the case for the MGT, the SRT will give a positive result for the presence of any lead compounds, not just those from GSR. Thus, these color tests lack the specificity desired for definitive forensic identification of GSR [15]. 1.4.1.2 Scanning Electron Microscopy/Energy Dispersive X-ray Analysis (SEM/EDX) SEM/EDX can be used to simultaneously determine a particle’s morphology (SEM) and elemental composition (EDX). GSR particles produced from the primer have a distinct spherical shape making them detectable under the high magnification of SEM. They also contain characteristic components of the primer, Pb, Ba, and Sb, that can be detected by EDX [8, 12]. It is this combination of Pb, Ba, and Sb in a single particle that is considered characteristic of GSR, while other combinations are considered to be consistent with, but not unique to, GSR [12, 24]. SEM/EDX, however, can be very time consuming (8-12 hours) as the analyst must search and locate GSR particles over a large surface area potentially covered in non-firearm related debris (such as dirt and skin) [8, 12]. This has in some part, however, been overcome by automated instrumentation [12, 13], but even the automated system can take 2-3 hours to search debris on one item [8, 27]. Another pitfall with SEM/EDX is due to the recent development of heavy metal- free primers that were introduced due to the growing environmental and health concerns over leaded ammunition [22]. More common metals such as zinc and strontium are now used in primers, but these metals alone are not specific to GSR. Therefore, other means of identifying GSR must be explored [12, 22]. 1.4.2 Improvised Explosive Device (IED) Post-blast Residue The recent increase in terrorist activity [10] has elevated the need for analytical methods that can accurately identify explosives and post-blast explosive residue, as this may provide a crucial link between an explosive and a suspect. Post-blast residue may be found on a suspect’s clothing, skin, or hair and can be compared with the powder and/or residue found at the scene of an explosion. Another forensically important determination in any bombing is the type of explosive used, which can be determined by analyses of the powder’s chemical composition; the use of taggants was recently deemed impractical and unnecessary, as it is possible to use chemical composition for identification [3, 16, 19]. Thus, post-blast residues can be, and typically are, collected and analyzed for their additive content as a means of characterizing and identifying powders [5, 6, 17]. Two main laboratory methods utilized for detection of post-blast residue are gas chromatography/mass spectrometry (GC/MS) and gas chromatography with a thermal energy analyzer (GC/T EA) [13], but these techniques have limitations as discussed in the following section. 1.5 Developing Methods for the Detection and Analysis of Smokeless Powder by Identifying Organic Components While definitive results for GSR can be obtained by SEM/EDX, analysis of organic residues can provide complementary or additional information, especially if inorganic components are not found or are inconclusive [12, 19]. Thus one new promising approach is the analysis of the organic additives in smokeless powder because this can not only be used for the detection of firearm use, but it can also be used to identify the powder used in an IED and it can be used to provide composition information that can potentially associate residues and unfired powder. The three most commonly used stabilizers in smokeless powder, MC, EC, and DPA, are regarded as being the most characteristic organic material in smokeless powder [15, 28, 29]. EC and MC are found in celluloid and solid rocket propellant [12, 16], while DPA is used in rubber products and in the food industry [14, 29]. Since there are environmental contamination possibilities for all these compounds the simultaneous identification of a number of additives may be required for unambiguous forensic identification and association [16]. The presence of nitrated derivatives of DPA, however, is considered virtually unique to GSR because industrial and environmental uses of DPA are not normally associated with nitrating agents [14, 29]. DPA is nitrated by the nitrogen oxides released by the degradation of nitrocellulose to form various nitro- and nitroso- derivatives, including 2-nitroDPA, 4-nitroDPA, and N-nitrosoDPA [11, 14, 29]. The thermal instability of most explosives and the need for high sensitivity due to the low concentration of stabilizer in smokeless powder, limits the number of analytical methods that can be used [27, 28, 30, 31]. Gas chromatography (GC) and liquid chromatography (LC) are the most commonly used techniques, coupled with a variety of detectors including ultraviolet (UV), thermal energy analyzer (TEA), electron capture detector (ECD), and mass spectrometry (MS) [30]. UV detectors are nonsclcctive and are only good at trace levels for detection of materials that absorb strongly in the UV range. 10 TEA is also nonselective. ECD has good limits of detection for nitro compounds, but its linearity and sensitivity are significantly dependent on the electron capture properties of the solvent and any impurities. A detection system that can overcome the disadvantages of UV, TEA, and ECD is mass spectrometry, which is both sensitive and selective [30]. GC/MS and LC/MS have been reported by many groups and labs for the analysis of smokeless powder [12, 16, 17, 30, 32, 33]. Wu et al. used HPLC with a triple quadrupole mass spectrometer in order to analyze standard solutions of MC and nitroglycerine (N G) as well as GSR extracted from gloves worn by a shooter [31]. By using tandem mass spectrometry and multiple reaction monitoring (MRM), few interfering peaks were seen. Chromatography techniques, however, are not without problems for explosive analysis. GC is disadvantageous due to the instability and thermal degradation of some nitrated components, such as DPA and its derivatives [5, 20, 33]. While the thermally unstable and non-volatile constituents of smokeless powders can by analyzed by LC [11], isocratic LC methods are limited due to the wide range of components’ polarities and difficulty in separating geometric isomers [5, 20]. These disadvantages have led to the development of some alternative methods. A gradient reversed-phase liquid chromatography — electrospray ionization mass spectrometry (LC-ESI-MS) method using a quadrupole ion trap mass spectrometry has been developed. [21]. There have also been studies that do not utilize any chromatography system. A triple quadrupole mass spectrometer with electrospray ionization and a flow-injection system (C18 column) was used to examine MC standards and GSR extracted directly from shooters’ skin [28]. In another study by this same group, a triple quadrupole mass analyzer with ESI interface was successfully used to examine standard solutions of DPA and its nitrated dcrivatcs ll along with GSR extractions from skin [29]. These previous ESI-MS studies, however, were limited in the number of additives detected at any given time. 1.6 Aims of this Thesis The aims of this thesis are: l. to develop and optimize a method for the identification of the organic stabilizers commonly found in smokeless powders (MC, EC, DPA, and DPA nitrated derivates) by nanoelectrospray ionization tandem mass spectrometry (nESI-MS/MS) with a quadrupole ion trap, 2. to develop a rapid and efficient technique for the extraction of organic stabilizers from small samples of smokeless powder, 3. to characterize and differentiate unburned smokeless powders based on their physical appearances, extraction yields, and mass spectra, and; 4. to use nESI-MS/MS to analyze gunshot residue obtained at different firing distances for the existence of the stabilizers MC, EC, and DPA and to compare this with the known source powder. 12 CHAPTER TWO Mass Spectrometry 2.1 Mass Spectrometry Mass spectrometry involves the ionization, separation, and detection of a sample in order to determine the characteristic ions present by measuring their mass-to-charge ratios (m/z). The speed, sensitivity, and specificity of mass spectrometry make it particularly attractive for use in forensic applications. A mass spectrometer consists of the components illustrated in Figure 2.1. ------------------------------------------------------------------------------------------ IonIFormation Ion Separation Ion Detection Sample Ionization Mass Introduction Ci) Source :.> Analyzer d Detector , Vacuum Pump Data Handling Data System 331:... ll 1 I1 Mass spectrum Figure 2.1 Components of a mass spectrometer (adapted from “What is Mass Spectrometry”. www.asms.org). l3 2.1.1 Ionization 2.1.1.1 Electrospray Ionization (ESI) The first component of a mass spectrometer is the ionization source, which is responsible for the conversion of molecules to ions. During the mid 1980’s, electrospray ionization (ESI), a ‘soft’ ionization technique, was developed by Penn et a1. [34]. E81, illustrated in Figure 2.2, involves dissolving the sample in a liquid solvent and then pumping this through a small diameter capillary tubing. The capillary tip, located at atmospheric pressure, is floated at high potential causing the formation of an electric field which induces charge accumulation at the surface of the liquid. When a high enough voltage is applied to break the surface tension of the solvent, a Taylor cone forms and an electrospray of charged droplets begins. Gaseous ions are then produced as these charged droplets undergo solvent evaporation, due to high temperature or the presence of a sheath gas, and coulomb fissions, due to either the charge residue model (CRM) [35] or the ion evaporation model (IEM) [36, 37]. The CRM argues that a sequence of Rayleigh instabilities (where columbic repulsion becomes greater than the surface tension) and periods of solvent evaporation produce final droplets that contain only one ion each. This ion is then liberated into a gas-phase ion as the last of the solvent evaporates. The IEM, on the other hand, proposes that before a droplet reaches the final stage in the CRM model, where it contains only one ion, the field on the droplet’s surface becomes strong enough to overcome the solvation forces, allowing an ion to escape from the droplet surface and enter the gas phase. It is thought that large molecules ionize according to the CRM, while the dominant mechanism for smaller molecules may be the IEM [3 8]. l4 Capillary tip + O t ® ® @‘I‘ + +"'1L @ © 69 ~1- Gaseous I ions Taylor cone Charged droplets For positive ions +| High-voltage I" 1 Power supply I Figure 2.2 Schematic of electrospray ionization (ESI). (Reproduced and modified from reference 42) ESI can be coupled with high performance liquid chromatography (HPLC) or used in a direct infusion mode. When coupled with HPLC, the analytes are dissolved in the appropriate HPLC solvent and are subjected to ESI directly as they elute from the column. While HPLC is not always applicable in forensic work, its advantages are that it is a separation technique that offers another possible means of identifying an unknown (retention time) and it is amendable to high throughput because it can be automated. In the direct infusion mode, analytes are dissolved in an appropriate solvent, such as methanol/water (1:1), and introduced into the mass spectrometer at the desired flow rate via a syringe pump. Nanoelectrospray ionization (nESI) uses less solution than 15 conventional ESI by running at a lower flow rate, ~1 uL/min or less, and takes advantage of the smaller droplet sizes [39]. 2.1.1.2 Matrix-assisted Laser Desorption Ionization (MALDI) Matrix-assisted laser desorption ionization (MALDI) was developed in the late 1980’s by Karas and Hillenkamp [40]. MALDI utilizes the impact of high energy photons from a laser on a sample imbedded in a solid organic crystalline matrix to ionize analytes. First, the sample is dissolved in a matrix containing a UV absorbing chromophore and crystallized onto a metal target. This target is then placed under vacuum in the ion source and bombarded with short duration laser pulses. This causes the sample to heat up and the sample and matrix to sublime. While the exact mechanism of ionization has been the subject of much discussion [41], it is generally accepted that it is the interaction of the laser pulse with the samples that results in the ionization of both matrix and the analyte molecules. In contrast to ESI, MALDI predominantly produces pseudomolecular ions that are singly charged. Although ESI was exclusively employed in the studies reported herein, MALDI is a sensitive ionization technique, and would likely serve as a complementary ionization technique for the analysis of smokeless powders. 2.1.2 Mass Analyzers The purpose of the mass analyzer component of the mass spectrometer is to separate mixtures of gas-phase ions according to their m/z ratios. A range of mass analyzers exists for this purpose, each with their own unique capabilities and operational 16 performance characteristics (e.g., resolution, sensitivity, mass accuracy). Three types of mass analyzers commonly used in forensic analyses are: (i) quadrupole, (ii) triple quadrupole, and (iii) quadrupole ion trap. The quadrupole ion trap analyzer, which was used in this study, is discussed in the following section. 2.1.2.1 The Quadrupole Ion Trap Mass Analyzer The quadrupole ion trap is an electrodynamic mass analyzer consisting of a ring electrode and two end caps, as illustrated in Figure 2.3. A variable amplitude radio frequency (RF) is applied to the ring electrode to create a three dimensional quadrupole electric field to trap ions within the region bound by the electrodes. By adding a small amount of bath gas, such as helium, the motion of ions injected into the trap can be dampened, thereby increasing their trapping efficiency. 17 ions in endcap electrode ring electrode ions out endcap electrode Figure 2.3 Diagram of a quadrupole ion trap. (Reproduced and modified from reference 43) The potential applied to the ring electrode can be given by the equation (Do = + (U-Vcos cut) (1) where U is the amplitude of the direct current (DC) potential, V is the amplitude of the applied alternating current (AC) potential (in the RF range), a) is the angular fi'equency ((o = 211:1), where 1) is the frequency of the applied RF), and t is time. The field is seen in three dimensions, thus the motion of the ions under the influence of the applied potentials occurs in three dimensions: x, y, and 2. Due to the cylindrical symmetry of the trap x 18 equals y and the ion motion can be expressed using the coordinates z and r. Thus, the equations of motion inside the trap can be shown to be 2 d 22- 4 e (U—Vcoswt)z=0 (2) dt m(r02 + 2202) d2r 2 e 2 + (U—Vcoswt)r =0 (3) dt m(r02 + 2202) where r0 and 20 are the distances from the center of the trap to the ring and exit electrodes, respectively, and z is the charge on the ion. Similarities can be seen when comparing these two equations with the Mathieu equation, which describes the propagation of waves in membranes, given below. 2 Q + (au — 2qu cos 28,")u = 0 4 .152 ( ) In equation (4), u can stand for either 2 or r in equations (2) and (3) and {f = EJ—t Th , the equations of ion motion in the ion trap given can be written in the form of the Mathieu equation: —162eU 2 m(r0 + 220 )a) 3U zazz—Zar : 2 (5) 19 426V q. = qz = —2q,- = ——2—7 (6) ”Nb CU 82eV qz= 2 2 2 (7) m(r0 +220 )0) where U is the DC potential, V is the amplitude of the RF potential, in is the mass of an ion, rO is the distance from the center of the trap to the ring electrode, 20 is the distance from the center of the trap to the exit electrodes, and a) is the angular velocity of the RF potential; re, 20, and (r) are constant fora given trap. There is no applied DC potential in three dimensional ion trap mass analyzers, thus the au term from the Mathieu equation equals zero. The solution for the dimensionless trapping parameter qu (related to the RF potential, V) is therefore the critical parameter. Equation (6) is the solution for an ideal ion trap, while equation (7) is for a ‘stretched’ quadrupole ion trap [43]. A ‘stretched’ trap is one where the distance between the end caps has been increased to introduce a negative higher order field (i. e., octapole) component to offset the positive higher order field introduced by truncation of the potential field from the ion trap electrodes. Ions are stable in the ion trap as long as their trajectories do not reach the distances r0 and 20 (Figure 2.3). By plotting au versus qu, the areas where ions of a given mass-to-charge ratio (m/z) are stable, i.e., do not reach values above or equal to to or 20, as a function of the applied RF frequency and amplitude, can be determined. Figure 2.4 shows such a stability diagram for a 3D quadrupole ion trap [42]. As there is no applied DC potential, all of the ions line up along the x-axis of the stability diagram. According 20 to equations (6) and (7), q is directly proportional to V, but inversely proportional to m/z. Thus, for a given m/z, as V is increased ions move along the x-axis towards higher q values. Ions will become unstable as they reach the boundary of the stability region, at a q value of 0.908, and will be ejected from the trap. Thus, a mass spectrum may be acquired by scanning the amplitude of the RF field (V) applied to the ring electrode to progressively destabilize ions of increasing m/z value. This, however, is not the most efficient manner to eject ions from a trap because several m/z ions can be ejected at essentially the same time due to the slope of the potential well in which ions are trapped (discussed later), thereby resulting in loss of resolution. RF Ejection Stable Ions V Clu Figure 2.4 Typical Mathieu stability diagram for quadrupole ion trap. The larger balls represent high mass ions whereas the smaller balls represent low mass ions. (Reproduced and modified from reference 42) 21 One solution to overcome this is to use resonance ejection. It was noted earlier that u is the frequency of the applied RF field. Ions in the trap, however, do not oscillate at this fundamental frequency because of their inertia. They will oscillate at a secular frequency f that is lower than u. The relationship between f and 1) along the z axis is given by B v z = 8 f z 2 I ) where ,8 Z is a fundamental stability parameter given by the approximation qz 1/2 : + —u 9 flu all 2 ( ) for qu values lower than 0.4. Since ,Bz equals one at the q value of 0.908 (where ions become unstable), the maximum secular frequency an ion can have is half the applied RF frequency. If a supplementary auxiliary AC (RF) signal, or ‘tickle’, is applied to the endcap electrodes, typically at a q value of 0.86, an ion’s secular frequency can be slowly raised (by increasing V) until it matches the applied frequency on the endcaps, i.e., it is in resonance with the AC signal, and is ejected. The multistage tandem mass spectrometry (MS/MS and MS") capabilities of the ion trap make it particularly attractive as a mass analyzer for trace analysis studies by improving both the signal-to-noise ratio (SW) and detection limits. Tandem mass 22 spectrometry involves the repeated isolation and fragmentation of ions, which occurs 11 times for an MSn analysis. For example, an MS/MS experiment would involve a series of events consisting of isolation of a precursor ion, an intermediate reaction event (typically involving energetic dissociation), followed by mass analysis of the product ions [44]. Isolation of the selected precursor ion is performed by the application of a high amplitude ‘notchcd’ broadband resonance ejection supplementary RF signal applied to the end caps in order to eject all ions except the precursor ion of interest. The isolated ion is then subjected to fragmentation by collision induced dissociation (CID), through the application of a low amplitude RF resonance excitation signal applied to the end cap electrodes. This signal is at a frequency corresponding to the secular frequency of motion of the ion of interest, enabling energetic collisions with the background He gas that is present. This causes a fraction of the kinetic energy to be converted into internal energy, bringing the ion to a vibrationally excited state and resulting in fragmentation if sufficient energy is deposited. The secular frequency of motion of the product ions are not in resonance with the supplemental RF signal and the ions are therefore ‘cooled’ by collisions with the background gas to the center of the trap. It is possible, however, for ion ejection to occur prior to fragmentation. The Dehmelt pseudopotcntial well associated with ion storage in the ion trap is given by the equation D—_q _V__ zeV2 z 28 m(r02+22§)m2 (10) 23 where D—z is the Dehmelt potential. This is illustrated in Figure 2.5, where the energy ERE represents the energy that is required at a specified q value in order for an ion to be ejected due to resonant excitation. Dissociation is typically performed at a q value of 0.25 to maintain a balance between the requirement for obtaining efficient ion fragmentation (rather than ejection), while keeping an appropriate low mass cutoff value for storage of the resultant product ions. | l 1 1 l ’ 0 0.2 0.4 0.6 0.8 1.0 Figure 2.5 Dehmelt pseudo well potential D (eV) as a function qz. Higher mass ions for a given V value have lower Dehmelt trapping energies. (Reproduced and modified from reference 42) 24 2.1.3 Detectors The remaining components of the mass spectrometer are the detector and a data system. The detector converts the ion flux into a proportional electric current. One of the most common detectors is a continuous-dynode electron multiplier with a conversion dynode (Figure 2.6). Electron multiplier Positive ions C9 69 ® Secondary ' particles / I I I I l 7 - HV Conversion dynode Cascade of electrons Resistive conductive / surface ‘ To ground via amplifier Figure 2.6 Diagram of a continuous-dynode electron multiplier with a conversion dynode. (Reproduced and modified from reference 45) When an ion strikes the conversion dynode, secondary particles (negative ions and electrons for positive ions) are emitted and accelerated into the electron multiplier where they strike the surface of the electron multiplier with enough force to cause secondary electrons to be ejected. Drawn towards the ground potential at the end of the multiplier, the emitted secondary electrons move further into the electron multiplier. As they do so, they too strike the surface of the multiplier causing the emission of more and more 25 electrons (cascade effect), thus amplifying the original signal. The data system then records the magnitude of this electrical signal from the detector as a function of mass-to- charge (i.c., a mass spectrum). 26 CHAPTER THREE Experimental 3.1 Materials Methanol (HPLC grade), 1,3-dimethyl-1,3-diphenylurea (methyl centralite, MC), 1,3-diethyl-1,3-diphenylurea (ethyl centralite, EC), and diphenylamine (DPA) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deionized water was purified by a Bamstead nanopure diamond purification system (Dubuque, Iowa, USA). All reagents were used as supplied without further purification. Men’s short sleeve white t-shirts by Fruit of the Loom (100% cotton, size XXL) were obtained from a local retail store. Smokeless powder samples were obtained by removing the powder from the following cartridges: (1) 9 mm, 124 grain, full metal jacket (FMJ) by Federal, (2) 9 mm +P, 135 grain, jacketed hollow point (JHP) by Federal, (3) 0.45 automatic colt pistol (ACP), 230 grain, JHP by Federal, (4) 0.45 ACP, 230 grain, PM] by Federal, (5) 0.40 Smith and Wesson (S&W), 155 grain, JHP by Federal, (6) 0.40 S&W, 155 grain, .IHP (‘Silvertip’) by Winchester, and (7) 0.357 Magnum, 125 grain, semi-jacketed hollow point by Remington-Peters. Cartridges 1, 3, 4, and 5 were provided by the Michigan State University Police Department, while cartridges 2, 6, and 7 were personally provided by a certified firearm instructor employed by the Michigan State University Police Department. All cartridges were stored in a cool, dry location prior to use. Gunshot residue samples were obtained by firing the aforementioned cartridges from the following handguns: (1) Glock Model 19, 9 mm, 4" barrel (for cartridges 1 and 2), (2) Sigarms Model P220, .45 ACP (for cartridges 3 and 4), (3) Sigarms Model P229, 27 4" barrel (for cartridges 5 and 6), and (4) Smith and Wesson Model 66, 2.5" barrel (for cartridge 7). The Sigarms Model P229 firearm was owned by the Michigan State University Police Department. All other firearms, and the MPro7 cleaning system (bore cleaning gel and gun cleaner) that was used, were personally owned by a certified firearm instructor employed by the Michigan State University Police Department. 3.2 Standard Solutions of the Smokeless Powder Stabilizers MC, EC, and DPA for Analysis by Mass Spectrometry MC was dissolved in methanol to make a stock solution of 100 pmol/pL. This solution was stored in a -20 °C freezer and dilutions with methanol were made daily to obtain concentrations of ~ 0.000001, 0.00001, 0.0001, 0.001, 0.01, 0.1, l, and 10 pmol/ttL. This process was repeated for EC and DPA. All solutions were centrifuged prior to injection into the mass spectrometer 3.3 Smokeless Powder Samples 3.3.1 Stereomicroscopy of Smokeless Powder Samples A Nikon SMZ800 stereomicroscope equipped with a Nikon digital DXM1200F camera was used to photograph the smokeless powders (Nikon Corporation, Tokyo, Japan). The software used was Nikon ACT-1, version 2.62. Direct lighting (used for the powders from cartridges 2, 4, and 7) was supplied by a Schott Fostec light (Schott North America Inc., Elmsford, NY), while oblique lighting (used for the powders from cartridges 1, 3, 5, and 6) was supplied by a standard light bulb. The photographs were 28 adjusted for contrast, color, and brightness using Adobe PhotoShop, version 8.0 (Adobe Systems Incorporated, San Jose, CA). 3.3.2 Extraction of Stabilizers from Smokeless Powder Samples for Analysis by Mass Spectrometry 1 mg of each smokeless powder was extracted with 1 mL of methanol by vortexing for about 30 seconds. The subsequent solutions were diluted to 0.1 mg/mL total concentration with methanol. In order to compare extraction efficiency, 10 mg of powder 2 (9 mm +P) was extracted with 10 mL of methanol by vortexing for roughly 30 seconds, yielding a solution with a total concentration of 0.1 mg/mL. All solutions were centrifiiged prior to injection into the mass spectrometer. 5 mg of each smokeless powder was extracted with 1 mL of methanol by either (1) vortexing for approximately 30 seconds or (2) crushing with the end of a spatula for about 1 minute and then vortexing for about 10 seconds. The powder was weighed before and after extraction in order to determine and compare extraction yields. Dilutions were then performed with methanol to obtain solutions with the desired concentration of extracted material (0.02 mg/mL, 0.5 mg/mL, and 0.1 mg/mL). All solutions were centrifuged prior to injection into the mass spectrometer. 5 mg of each of the smokeless powders was extracted with 1 mL of deionized water by crushing with the end of a spatula for about 1 minute and then vortexing for about 10 seconds. No dilutions were made. All solutions were centrifuged prior to injection into the mass spectrometer. 29 3.4 Gunshot Residue (GSR) 3.4.1 Collection of GSR Target paper was fastened to a foam target backer and 12" x 12" squares of cloth cut from white t-shirts were pinned to the paper one at a time to avoid the accidental deposition of gunshot residue on cloth not being fired at. Each cloth square was pinned to a different location on the target paper to avoid gunshot residue transferring from the target paper to the cloth. To avoid possible cross contamination between cartridges, the target paper was changed between each cartridge type and the firearms were extensively cleaned between cartridges. Each cartridge was fired twice at a firing distance of 3": once into a square of unwashed cloth and once into a square of cloth that had been extensively washed with deionized water to remove any potential contaminants and air dried. Cartridge 2 (9 mm +P) was also fired into a square of cloth that had been machine washed and dried. Once fired into, the square of cloth was placed between two pieces of wax paper, folded, and sealed in a plastic storage bag. This process was repeated at a firing distance of 12". These distances were chosen because they are within the range of distances commonly used by forensic labs for test firings. 3.4.2 Extraction of GSR from Cloth for Analysis by Mass Spectrometry A sample of cloth (~ %” by M1”) was cut from directly beside the bullet hole where visible gunshot residue was present. This cloth was placed in 1 mL of methanol and vortexed for approximately 30 seconds. No dilutions were made and all solutions were 30 centrifuged prior to injection into the mass spectrometer. Samples were run on the day of collection. 3.5 Mass Spectrometry All experiments were performed using a Thermo model LCQ quadrupole ion trap mass spectrometer (Themio Scientific, San Jose, CA). Solutions were introduced to the mass spectrometer at a flow rate of l pL/min by nanoelectrospray ionization (nESI). The nESI conditions were optimized to maximize the intensity of the protonated precursor ion and to minimize the appearance of in-source fragmentation peaks. Typical nESI conditions were: spray voltage 2.5 kV, heated capillary temperature 125 °C, capillary voltage 0 V, and tube lens voltage 0 V. Mass spectra were acquired using the normal resonance ejection scan mode of the quadrupole ion trap mass spectrometer. The precursor ion accumulation time for each scan was controlled by the automatic gain control (AGC) function of the instrument in order to maintain an ion target number of 2 x 107 (arbitrary value). CID MS/MS and MS3 spectra were acquired at an activation q value of 0.25 using isolation widths of 3-10 Da, normalized collision energies of 10 - 50%, and an activation time of 30 ms or 300 ms. The values were chosen such that the gentlest conditions were used in order to completely dissociate the selected precursor ion (1'.e., some precursor ions required a larger normalized collision energy and/or longer activation time). Full and selected reaction monitoring (SRM) scan types were used. The MS spectra and the MS/MS and MS3 product ion spectra shown throughout this thesis are an average of 100 and 60 individual mass analysis scans, respectively. 31 CHAPTER FOUR Nanoelectrospray Ionization Tandem Mass Spectrometry for the Analysis and Identification of Ethyl Centralite, Methyl Centralite, and Diphenylamine 4.1 Introduction Smokeless powder is one of the most common explosives used today in gunpowder and improvised explosive devices (IEDs). The ability to rapidly identify and characterize smokeless powder samples and their residues is therefore of great forensic value. The most common method for the forensic analysis of smokeless gunpowder or gunshot residue is to identify inorganic components contained in the primer [8, 12]. With the relatively recent introduction of heavy metal-free primers, however, there is the need for the development of a different approach, such as the analysis of organic components in the propellant [12, 22]. Stabilizers, added to slow the decomposition ofnitrocellulose, are considered to be the most unique organic components of smokeless powder [15, 28, 29]. The three most commonly used stabilizers are methyl centralite (MC, Structure 1), ethyl centralite (EC. Structure 2), and diphenylaminc (DPA, Structure 3). \ 0 Structure 1: Methyl Centralite (MC) 32 Structure 2: Ethyl centralite (EC) Structure 3: Diphenylamine (DPA) As discussed in Section 1.5 there have been many techniques used to analyze smokeless powder, but mass spectrometry offers the much desired sensitivity and specificity. The quadrupole ion trap mass spectrometer is an ideal instrument due to its ability to perform multiple stages of dissociation (e.g., MS"), thus allowing further characterization and possible identification of unknown components. Although mass spectrometry is commonly coupled with either gas chromatography (GC) or liquid chromatography (LC), GC can cause the thermal degradation of some compounds, such as DPA, while LC is limited due to the potentially wide range of polarities present in a powder. Nanoelectrospray ionization (nESI), however, helps overcome these difficulties and minimizes analysis time by allowing direct introduction of the sample to the mass spectrometer. In this study, standard solutions of the organic stabilizers commonly found 33 in smokeless powder (MC, EC, and DPA) were examined by nESI-MS using a quadrupole ion trap in order to develop and optimize a method for the routine detection and compositional analysis of smokeless powder. 4.2 Development and Optimization of Conditions for the Detection of MC, EC, and DPA It is important to maximize the signal for the ions of interest in order to achieve a low limit of detection (LOD). This is especially true for residues of smokeless powder, which are more likely to contain only trace amounts of the stabilizers compared to the amounts in unburned powder. The conditions needed to maximize the signal from one stabilizer, however, may cause the signals from the other stabilizers to decrease. It is therefore crucial that the instrumental conditions be optimized for the simultaneous identification of all three stabilizers, since it would not be known at first Which, if any, stabilizer was present in a sample. Thus, a compromise must be reached whereby the conditions used maximize the signals obtained from MC, EC, and DPA such that all have a signal high enough for trace analysis. Typical nESI-MS conditions were optimized to maximize the intensity of the protonated MC, EC, and DPA ions, while minimizing the appearance of in-source fragmentation peaks. Iii-source fragmentation, or source-collision induced dissociation, occurs when sufficient kinetic energy is imparted to an ion while in the source to cause the ion to fragment when it collides with solvent or air molecules. This type of fragmentation can reduce the intensity of the signal from the precursor ion, in this case MC, EC, or DPA, which can hinder detection and identification if the compounds are 34 only present in trace amounts. Most of the nESI-MS conditions (1'.e., the entrance lens, interoctapole lens, multipole 1 and 2 etc.) did not greatly affect the peaks observed, so standard settings were used without problems. The capillary voltage, tube lens voltage, and heated capillary temperature, however, were found to have a significant affect on the abundance of the protonated precursor ion and on the extent of in-source fragmentation. The capillary voltage was examined from 0 V to -60 V and the effects of the tube lens voltage were analyzed from 0 V to -250 V. At the higher voltages for both the capillary and the tube lens, in-source fragmentation became pronounced. The heated capillary temperature was examined from 100 °C up to 325 °C. If the temperature was set too low, the signal was lost, but if the temperature was too high, in-source fragmentation occurred. For these reasons, the capillary and tube lens voltages were set at 0 V and the heated capillary temperature was set at 125 °C for all three compounds of interest. 4.3 Analysis of the Mass Spectra Obtained for MC, EC, and DPA Standards Once the nESI-MS conditions were optimized, spectra for MC, EC, and DPA could be obtained under these conditions. In a forensic context, an analyst will not be able to spend a lot of time on the analysis of any given sample. It is therefore necessary for a forensic scientist to be able to specifically look for peaks in a spectrum that are known to be associated with one of these stabilizers instead of having to analyze an entire spectrum. Tandem mass spectrometry can then be used to conclusively identify peaks of interest, if the fragmentation pathway has been previously established with standards. Thus, by running standards of MC, EC, and DPA, key peaks and their dissociation behavior can be established such that unknown powders can later be rapidly analyzed. 35 Figure 4.1 shows the mass spectra acquired for MC (Figure 4.1A), EC (Figure 4.18), and DPA (Figure 4.1C) under optimized conditions. MC and EC were analyzed at concentrations of 10 pmol/uL, while DPA was run at 100 pmol/uL. DPA did not ionize as efficiently as MC and EC and thus required a higher concentration in order to achieve a signal-to—noise (S/N) ratio comparable to that of MC and EC. Even at the 100 pmol/uL concentration, however, the spectrum for DPA (Figure 4.1C) is noisier than the others, but DPA could not be run more concentrated without contaminating the system. DPA is therefore likely to have a higher limit of detection than that of MC and EC. At these high concentrations (10 pmol/uL for MC and EC, 100 pmol/uL for DPA), MC and EC form dimers and adducts with Na and K, while DPA forms a dimer, trimer, and quadramer and Na adducts, but the S/N ratio is such that background ions are minimized and the minimal in-source fragmentation obtained under the optimized conditions is easily observed. A good S/N ratio is initially required in order to clearly identify peaks in the spectra of the standards. Once peaks of interest are identified, however, high background ions would be inconsequential in a forensic setting as long as the ion of interest could be detected. 36 100- [2M+Na]* A 503 I [M+H-Nc,11,-cor 4 106 [M+H-NC7H9]‘ ”W”? 134 481 [2M+K]* ‘ [WNW 519 [M+11]+ 7'63 r 241 [M+K]* 279 . ll/- .1?" 1-1 . 63°C I I I I I I I I I I I l 50 150 250 350 450 550 650 1001 [2M+Na]+ a, B 559 Q g: —1 ('3 "g [M+H-NC,H,,-cor B -1 120 [2M+H]+ <3 [M+H-NC,H,,]+ 537 O.) _ I48 [2M+K]+ .2. [MtNal’ 575 E |M+Hj+ 291 _ 26 odd) 9 [M+I(]+ / o\° [1,307 ‘33 [H 628 1 1 1 1 ’ 1 1 1 1 1 i l I 1 ’ 1 50 150 250 350 450 550 650 100— [M+H]+ C 170 - [4M+11]+ 676 [2M+H]* d 339 [2M+Na]’ 361 [3M+H]‘ 508 ‘ 355 r331,408 1 LJIL I N "41m“? 1 IMLJ1 I’Auige'" I I Ll 50 150 250 350 450 550 650 m/z Figure 4.1 Mass spectra under optimized nESI-MS conditions for standard solutions of (A) MC at 10 pmol/uL, (B) EC at 10 pmol/pL, and (C) DPA at 100 pmol/pL. 37 4.3.1 MC Tandem mass spectrometry was used to identify and characterize the peaks in the spectra obtained for MC (Figure 4.1A) and this data is shown in Table 4.1 below. Dissociation of the protonated precursor ion for MC (m/z 241 in Figure 4.1A) yielded a product ion at m/z 134, corresponding to [NC7H3CO]+, and a product ion at m/z 106, which could be the ion [NC7H3]+, produced by the loss of CO from the product ion at m/z 134. This was confirmed by performing MS3 on the m/z 134 product ion, which yielded the ion at m/z 106. The mechanistic pathway is shown in Scheme 4.1. The ion at m/z 481 has a mass consistent with the protonated MC dimer. When MS/MS was performed 011 this precursor ion, the product ion formed was at m/z 241, the same m/z as protonated MC. MS3 on this m/z 241 product ion resulted in the formation of ions at m/z 106 and 134, confirming that the product ion at m/z 241 is [MC+H]+ and the precursor ion at m/z 481 is [2MC+H]+. While the ions at m/z 263 and 279 did not readily dissociate, the mass difference between these ions and the [MC+H]+ ion at m/z 241 is 22 and 38, respectively. These characteristic mass differences are indicative that the ion at m/z 263 is the sodium adduct [MC+Na]+, while the ion at m/z 279 is the potassium adduct [MC+K]+. This same logic holds for the ions at m/z 503 and 519, which have masses that are consistent with Na and K adducts for the MC dimer, respectively. Further evidence of this is obtained through MS/MS experiments, whereby the ion at m/z 503 (probable [2MC+Na]+ ion) fragments to give m/z 263 (the m/z of the likely [MC+Na]+ adduct) and the ion at m/z 519 (probable [2MC+K]+ ion) fragments to yield m/z 279 (the m/z ofthe [MC+K]+ adduct). The ion at m/z 600 dissociates to yield product ions at m/z 263, 320, 360, and 429, none of which 38 would fragment further to give MS3 data. The product ion at m/z 263, however, could be the adduct [MC+Na]+, thus the ion at m/z 600 could be a complex containing MC and a Na adduct. 1,1 GIG MS/MS - 107 M53 -28 Z— / Scheme 4.1 Possible mechanism for the fragmentation of protonated MC (m/z 241) to form product ions at m/z 134 and 106. 39 The other ions seen in the mass spectra for MC either did not fragment due to low abundance, such that characterization was not possible, or their fragmentation behavior did not readily lead to identification as there were no characteristic losses. Some of these ions could be tentatively identified based on comparisons between the MC and EC spectra (discussed in section 4.3.2) All of these ions, however, have low relative abundances compared to the main MC peaks, which would be the peaks of interest in a forensic analysis aimed at identifying an unknown as smokeless powder, since MC, EC, and DPA are the organic components considered most characteristic of smokeless powder. Table 4.1 MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 10 pmol/ 11L standard solution of MC in methanol. Component Precursor ions MS/MS M83 (m/z) product ions (m/z) product ions (m/z) (% relative abundance) (% relative abundance) MC 241 106 (4%) 134 (100%) 106 (100%) MC+Na 263 MC+K 279 Unidentified 380 260 ( 100%) 216 ( 100%) 2MC 481 241 (100%) 106 (1%) 134 (100%) *MC 493 *MC 500 2MC+Na 503 263 (100%) 2MC+K 519 279 (100%) *MC 600 263 (94%) 320 (86%) 360 (100%) 429 (30%) * Unidentified MC complex/adduct 40 4.3.2 EC The fragmentation behavior of EC is similar to that seen for MC due to their similar structures (Structures 2 and 1, respectively). The MS and MSn data for EC is summarized in Table 4.2 below. The protonated precursor ion at m/z 269 (Figure 4.1B) dissociates to give product ions at m/z 148 and 120. The product ion at m/z 148 is likely the ion [NC3H10CO]+, while the ion m/z 120 would be [NC8H10]+, formed by the loss of CO from the m/z 148 product ion. As expected, MS3 of the m/z 148 product ion gave the ion at m/z 120. MS4 of the ion at m/z 120 yielded an ion at m/z 92, or [NC6H6]+ formed by the loss of CHZCHZ. This is illustrated mechanistically in Scheme 4.2. The ion at m/z 537 has a mass that is consistent with the protonated EC dimer. MS/MS of m/z 537 gave the product ion m/z 269, which could be [EC+H]+. MS3 of this product ion at m/z 269 yielded ions at m/z 148 and 120 confirming that the product ion at m/z 269 is [EC+H]+ and that the precursor ion at m/z 537 is [2EC+H]+. The mass difference between the [EC+H]+ ion at m/z 269 and the ions at m/z 291 and 307 was 22 and 38, respectively indicating that the ion at m/z 291 is the sodium adduct [EC+Na]+, while the ion at m/z 307 is the potassium adduct [EC+K]+. Similarly, the ions at m/z 559 and 575 have masses and fragmentation behavior consistent with [2EC+Na]+ and [2EC+K]+, respectively. 41 1 NH if MS/MS -121 _/ uri- MS3 28 H2 /2\c H2|C AV /(I. MS4 28 Scheme 4.2 Possible mechanism for the fragmentation of protonated EC (m/z 269) to form product ions at m/z 148, 120, and 92. 42 The ion at m/z 628 fragmented to give five product ions that included m/z 269 and m/z 291, which could be [EC+H]+ and [EC+Na]+, respectively. MS3 of the product ion at m/z 269 yielded fragmentation that is consistent with [EC+H]+. The product ion at m/z 291, however, did not fragment, but this would be consistent with observations for the [EC+Na]+ ion. As was the case of the ion at m/z 291, none of the other three product ions would fragment and thus further characterization of the ion at m/z 628 was limited. The data from the ions at m/z 269 and 291, however, provide strong evidence that the ion at m/z 628 is a complex containing EC and a Na adduct. This also provides further evidence that the ion observed at m/z 600 for MC (Figure 4.1A) is indeed an MC and Na containing complex as proposed because the mass difference between the EC ion at m/z 628 and the MC ion at m/z 600 is 28, or the mass difference between MC and EC (mass of 28 = C2H4). This mass difference of 28 is seen for EC and MC monomers identical in composition except that one contains EC and one contains MC (such as [EC+H]+ at m/z 269 and [MC+H]+ at m/z 241 or [EC+Na]+ at m/z 291 and [MC+Na]+ at m/z 263). The ions at m/z 549 and 556 in the EC spectrum have a mass difference between them and the ions at m/z 493 and 500 seen in the MC spectra of 56, or the mass difference between 2MC and 2EC (mass of 56 = C4H3). This difference is seen for other ions such as [2EC+Na]+ at m/z 559 and [2MC+Na]Jr at m/z 503. This suggests these ions may have identical compositions except that the ions at m/z 549 and 556 contain the EC dimer and the ions at m/z 493 and 500 contain the MC dimer. This is further confirmed by comparing the mass of these ions with the respective protonated monomer. The ions at m/z 549 and 556 in the EC spectrum have a mass difference between them and the [EC+H]+ ion (m/z 269) of 280 and 287, respectively. The mass difference for the ions at 43 m/z 493 and 500 seen in the MC spectra are 252 and 259 when compared to the [MC+H]+ ion (m/z 241). The difference between 280 and 252 and the difference between 287 and 259 is 28, the mass difference between MC and EC. As was the case for EC, though, these unidentified ions are low in abundance and therefore of little forensic value compared to the other identified peaks. 44 m/z 493 and 500 seen in the MC spectra are 252 and 259 when compared to the [MC+H]+ ion (m/z 241). The difference between 280 and 252 and the difference between 287 and 259 is 28, the mass difference between MC and EC. As was the case for EC, though, these unidentified ions are low in abundance and therefore of little forensic value compared to the other identified peaks. 44 Table 4.2 MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 10 pmol/uL standard solution of EC in methanol. Component Precursor MS/MS MS3 MS4 ions product ions product ions product ions (m/Z) (m/Z) (m/Z) (m/Z) (% relative (% relative (% relative abundance) abundance) abundance) EC 269 120 (14%) 148 (100%) 120 (100%) 92 (100%) EC+Na 291 EC+K 307 Unidentified 423 288 (100%) 181 (70%) 223 (98%) 274 ( 100%) 2EC 537 269 (100%) 120 (9%) 148 (100%) *EC 549 414 (100%) 280 (12%) 289 (65%) 423 (40%) 298 ( 100%) *EC 556 2EC+Na 559 291 ( 100%) 2EC+K 575 307 (100%) *EC 628 269 (15%) 120 (7%) 148 (100%) 120 (100%) 291 (100%) 360 (82%) 434 (55%) 457 (32%) * Unidentified EC complex/adduct 45 MS/MS -78 HN Scheme 4.3 Possible mechanism for the fragmentation of protonated DPA (m/z 170) to form a product ion at m/z 92. 47 Table 4.3 MS and MSn data obtained by nESI quadrupole ion trap mass spectrometry for a 100 pmol/ 11L standard solution of DPA in methanol. Component Precursor ions MS/MS MS3 (m/z) product ions (m/z) product ions (m/z) (% relative abundance) (% relative abundance) DPA 170 92 (100%) 2DPA 339 170 (3%) 92 (100%) 303 (52%) 321 (100%) 303 (100%) Unidentified 355 339 (100%) 2DPA+Na 361 Unidentified 391 296 (100%) Unidentified 408 221 (100%) 80 (100%) 249 (88%) 266 (32%) 294 (32%) 362 (23%) 3DPA 508 170 (4%) 92 (100%) 339 ( 100%) 4DPA 676 339 (100%) 170 (100%) 4.4 Limits of Detection and Calibration Curves In order to determine the limit of detection (LOD) and create calibration curves for MC, EC, and DPA, each standard was run at concentrations ranging from 0.000001 pmol/pL to 10 pmol/uL. Selected reaction monitoring (SRM) was used to help determine the LOD as this is an ideal scan type for trace analysis. As the name suggests, SRM monitors a specific reaction, such as a specific fragmentation pathway for a given ion. First, the parent ion is trapped in the mass analyzer and all other ions are ejected. The selected precursor ion is then fragmented into its product ions. The product ion of 48 interest is ejected from the mass analyzer to produce an SRM product ion mass spectrum. Data is not collected for any of the other precursor or product ions. The advantages of SRM compared to the full scan mode are that it allows for the rapid analysis of trace components in a complex mixture by improving S/N and detection limits. In this work, at low concentrations the protonated precursor ions were sufficiently low in abundance as to be in the noise. As the fi‘agmentation behavior of these ions, however, was previously determined (Section 4.3), SRM could be used to monitor the most abundant product ion produced by the dissociation of the [M+H]+ ion. For MC this product ion was 1n/z 134. Similarly, the product ion at m/z 148 was monitored for EC and the product ion at m/z 92 was used for DPA. Using this method, the LOD for MC and EC was found to be 0.001 pmol/uL, while the LOD for DPA was found to be 0.01 pmol/uL. The slightly higher LOD for DPA is consistent with its ionization efficiency being less than that of MC and EC, as mentioned in Section 4.2. While these LOD are very good, they could be improved further by using a linear two dimensional (2D) ion trap, which has improved sensitivity. This is because linear 2D quadrupole ion traps exhibit higher acceptance (i. e., more efficient ion injection) compared to 3D traps, due to the lack of a quadrupolar field along the z-axis. 2D traps also have orders of magnitude greater ion storage capacity due to the larger volume of the device and the fact that ions are focused along the entire length of the quadrupole due to the radial nature of the quadrupolar confinement field [47, 48]. For each of the stabilizers, calibration curves were generated that ranged from the LOD for each standard up to 1 pmol/pL. As an example. Figure 4.2 shows the calibration curve for EC in methanol. 49 4.5136 4,0136 ~ y=4000000x+47534 3.5136 , R2=0.9988 3.0136 . 2.5136 . 2.0E6 . 1.5E6 4 1.0136 « SRM intensity of m/z 148 5.0E5 : 3.0E3 z ' 1 ' ' 1 - ' ‘ 0 0.2 0.4 0.6 0.8 1.0 [EC] (pmol/uL) Figure 4.2 Calibration curve for EC in methanol (0.001 pmol/uL to 1 pmol/uL) using selected reaction monitoring of m/z 269 fragmenting to m/z 148. The results from the standard solutions at 10 pmol/pL were not included in the calibration curves because at such high concentrations adducts and other complexes formed (Figure 4.1) to the point of causing the relative abundances of the [MC+H]+, [EC+H]+, and [DPA+H]+ ions to decrease. Thus, adduct and complex formation makes quantification above the 1 pmol/uL concentration difficult. Also, the adduct ions are themselves not reliable candidates for quantification because of the widespread existence of Na and K that would greatly influence the concentration of these adducts. Even a simple change in the tubing of the nESI source could cause a drastic change in the abundance of these adducts as the new tubing may have very different concentrations of Na and K compared to the old tubing. 50 The calibration curves were also slightly non-linear at the lower concentrations. Thus quantification is not reliable without plotting a significantly large number of points at low concentrations to account for the non-linearity. Overall, due to the formation of complexes and adducts at high concentrations and non-linearity at low concentrations. quantification using nESI-MS is not reasonable. Although it may be possible to quantify using an internal standard or isotopic labeling, from a forensic standpoint, quantification of the components of smokeless powder is not as important as identification of the components. Hence, qualitative analysis is sufficient to identify a powder or residue as being smokeless powder and to differentiate between smokeless powders and their residues. 4.5 Summary Standard solutions of the organic stabilizers MC, EC, and DPA, which are commonly found in smokeless powder, were examined by tandem mass spectrometry (MS/MS and M83) in a quadrupole ion trap using nESI in order to develop a method for the routine and rapid identification and comparison of smokeless powders. Following optimization of the ionization and mass analyzer conditions, mass spectra for standard solutions of MC, EC, and DPA were obtained. MC gave the characteristic fragmentation pathway m/z 241 —> 134 —> 106. EC’s protonated precursor ion at m/z 269 dissociated to give 148 —> 120 —* 92. DPA fragmented from m/z 170 to m/z 92. This method was able to detect all three stabilizers using the same nESI-MS conditions. Acceptable LOD (0.001 pmol/ttL, or 1 fmol/pL, for MC and EC and 0.01 pmol/uL, or 10 fmol/jiL, for DPA) were obtained by utilizing the selected reaction monitoring (SRM) scan mode. 51 Calibration, however, was not possible due to non-linearity of the calibration curves at low concentrations and due to the formation of adducts (Na, K) and complexes at higher concentrations. 52 CHAPTER FIVE Nanoelectrospray Ionization Tandem Mass Spectrometry for the Analysis of Smokeless Powder and its Residue 5.1 Introduction While methyl centralite (MC), ethyl centralite (EC), and diphenylaminc (DPA) are considered to be the most unique organic components in smokeless powders, they are used for other purposes. Methyl centralite (Structure 1) is found in celluloid and solid rocket propellant, as is ethyl centralite (Structure 2) [12, 16]. Diphenylamine (Structure 3) is used in rubber products and in the food industry. The nitrated derivatives of DPA that form as nitrocellulose degrades, however, are thought to be unique to smokeless powder as the other uses for DPA do not involve nitrating agents or processes [14, 29]. Thus, the identification of more than one of these stabilizers in a sample or identification of a nitrated DPA compound could provide more certainty that the substance being analyzed was smokeless powder, burned or unburned [16]. Previous experiments using mass spectrometry to analyze the organic stabilizers used in smokeless powder, however, have been limited in the number of stabilizers detected simultaneously or in the number of samples analyzed [28, 29]. In this study, samples of smokeless powder were obtained from seven different cartridges and assigned a number that will be used to refer to each powder: (1) 9 mm, 124 grain, full metal jacket (FMJ) by Federal, (2) 9 mm +P, 135 grain, jacketed hollow point (JHP) by Federal, (3) 0.45 automatic colt pistol (ACP), 230 grain, JHP by Federal, (4) 0.45 ACP, 230 grain, PM] by Federal, (5) 0.40 Smith and Wesson (S&W), 155 grain, 53 JHP by Federal, (6) 0.40 S&W, 155 grain, JHP (‘Silvertip’) by Winchester, and (7) 0.357 Magnum, 125 grain, semi-jacketed hollow point by Remington-Peters. These powders were then compared based on their physical characteristics, i.c., color, shape, texture. and size, and on their extraction characteristics. Finally, the smokeless powder samples and their residues were examined by using the previously optimized method for nESI-MS (Chapter 4) in a quadrupole ion trap. 5.2 Physical Characteristics of Smokeless Powder Samples The initial characterization of unburned smokeless powder is typically a survey of the powder morphology using a stereomicroscope. If the powder morphologies are vastly different, the powders can be excluded as having originated from a common source. When the powders are not distinguished based on morphology, however, further chemical analyses are necessary. Figure 5.1 shows photographs taken of the seven powders used in this study using a stereomicroscope. 54 Figure 5.1 Stereomicroscopy photos of smokeless powder samples. It can be seen than powders 1, 3, 5, and 6 have similar morphology. They are all relatively round in shape with slightly rough surfaces and a gray-brown color, almost like rocks or pebbles. Slight differences, however, can be seen among these four powders. Powder 3 is significantly rougher in texture than the other three powders, with the order of roughness (from most to least) being 3, 5, 6, and 1. Also, the color of powders 3 and 5 55 is somewhat different than it is for powders 1 and 6, which have a more brown hue to them. The sizes of particles in powders 1 and 6 are more widespread than for powders 3 and 5, with powders 1 and 6 have particles ranging from small (~ 300 um) smooth spheres to large (~ 600 um), circular-like shapes. Powders 3 and 5, however, have particles that are all roughly 600 um. Powders 2, 4, and 7 consist of circular discs that are approximately 600 pm and grainy in texture. Powders 2 and 4 are dark gray in color, while powder 7 consists of some particles that are black and some that are light gray. The thickness of the particles for powders 2 and 7 is more variable than it is for powder 4. Overall, powders 1 (9 mm), 3 (0.45 ACP, JHP), 5 (0.40 S&W, Federal) and 6 (0.40 S&W, Winchester) are distinctly different from powders 2 (9mm +P), 4 (0.45 ACP, FMJ), and 7 (0.357 Magnum). Powder 1, 3, 5, and 6 show enough similarities that it would be difficult to differentiate between them based on physical properties alone. Only the roughness of powder 3’s surface might allow it to be matched to an unknown powder and for powders 1, 5, and 6 to be excluded. Powders 2 and 4 are also too similar in nature to be differentiated from one another based on physical appearance, but powder 7 can easily be identified because it is the only powder to contain particles of two different colors. Since only one of the seven powders could be definitively differentiated from the other powders, further analysis is needed. 5.3 Extraction Efficiency of Smokeless Powder Samples As a rapid, initial survey of the extractability of smokeless powder in methanol, 1 mg of powder 2 (9 mm +P) was extracted with 1 mL of methanol by vortexing for about 56 30 seconds and diluted to 0.1 mg/mL concentration. 10 mg of this same powder was also extracted with 10 mL of methanol by vortexing for roughly 30 seconds, yielding a solution with a concentration of 0.1 mg/mL in methanol. The mass spectra for these solutions did not show any major differences in the ions seen or their abundances, indicating that similar amounts of the same material were being extracted with both methods. Thus, there is no need to use a large quantity of powder when extracting, which is significant in forensic cases since the amount of sample may be very limited. In order to determine the mass of material extracted from the smokeless powders, 5 mg of each powder was extracted with 1 mL of methanol by vortexing for about 30 seconds. The supernatant solution was carefully pipetted off and the powder dried and reweighed in order to determine the amount of material extracted and allow calculation of an extraction yield. This same procedure was repeated for all the powders, but instead of simply vortexing the solutions, a spatula was used to thoroughly crush the powder and then the solution was vortexed for about 10 seconds. These methods were performed at least three times for each powder and the average extraction yields, along with relative standard deviations (RSDs), are given in Table 5.1. Also, 5 mg of each of the smokeless powders was extracted with 1 mL of deionized water using the crushing method. This was only done once for each powder, since it was clear that less material was being extracted into water than was extracted into methanol. Extraction yields for this solvent are also listed in Table 5.1. 57 Table 5.1 Extraction data for smokeless powder samples in methanol (MeOH) and in water (H20). % Extracted in °/o % Extracted in % % Extracted in Powder MeOH, RSD MeOH, RSD H20, Vortexed* Crushed* Crushed 1 6.03 34.60 12.05 6.67 8.68 2 43.06 14.53 53.52 4.29 5.21 3 20.63 14.48 20.05 2.00 4.67 4 40.24 13.50 39.50 6.68 6.26 5 3.19 33.30 25.27 1.73 7.05 6 7.52 10.52 12.48 11.85 10.75 7 37.45 13.09 46.34 7.15 10.38 * Average of 3 or more extractions. The mass of extracted material for a given powder varied more when the solution was vortexed only versus when the powder was crushed first. This is evident in the RSDs for the two techniques as the RSDs for all the powders are higher when the powders were vortexed only, with the exception of powder 6 (0.40 S&W, Winchester) where the RSDs are about the same. Not only was there less variability in the amount of extracted material with the crushing technique, but more material was extracted for most of the powders — powders 3 (0.45 ACP, JHP) and 4 (0.45 ACP, FMJ) had essentially equivalent amounts of extracted material for both techniques. While optimizing this extraction procedure, it became evident that this data could also be used to differentiate between samples of unburned smokeless powder as some trends were observed in the extraction data for the powders crushed in methanol. 58 Powders l (9 mm) and 6 (0.40 S&W, Winchester) had the least amount of material extracted (less than 15%), while powders 2 (9mm +P), 4 (0.45 ACP, FMJ), and 7 (0.357 Magnum) had over 40% extraction yields. Powders 3 (0.45 ACP, JHP) and 5 (0.40 S&W, Federal) yielded extractions of around 20-25%. This suggests that powders 2, 4, and 7 have the highest concentration of extractable organic components, while powder 1 and 6 have the least. The grouping of powders seen here is the same as the grouping based on similarities in physical characteristics (section 5.2). Thus, extractability can add another level of certainty when trying to determine if two powders are the same or different. Here, this would be especially useful for differentiating powders 3 and 5 from powder 1 and 6, which were difficult to distinguish based on physical traits. 5.4 Analysis of Smokeless Powder Samples by nESI-MS The detection of the compounds MC, EC, and/or DPA in a powder or residue can provide strong evidence that the explosive used was smokeless powder. This can be used, for example, to determine the explosive used in a bombing or to look for gunshot residue surrounding a hole that is thought to be a bullet hole. Also, comparison of the presence or absence of these compounds in a powder can help distinguish it from other powders, especially when combined with differences in the powders’ physical attributes and extraction yields. For the seven powders used in this study, dilutions were performed in methanol on the solutions obtained by crushing in order to obtain concentrations of extracted material of 0.002 mg/mL, 0.02 mg/mL, 0.5 mg/mL, and 0.1 mg/mL. Mass spectra were 59 then collected using the previously optimized conditions (Section 4.2). The spectra obtained for powder 2 (9 mm +P) are shown in Figure 5.2 as an example. 60 m w .m a m m . m a NE E: m L own owe 0mm Omv omm 0mm of 0m own owe omm omv 0mm 0mm of cm w W __1_ __.c_ 1.12 _ _ p ___.r .E1p_ p. _1_ mm . _L 1 . . l .4 . 1 :_.1_ a- % 1 \ \ “0 m .J S 1 am 1 P... 1m mu; Suzi): .7214: now. a m a 1 m: 1 M m d .QEQ .3122 n w. .1 a 1 1 n mm, L, u SN amm\ w. + m an .513): 1 #21123 1 w m We gaze/E 5. D H Sm 1L SN 0 a mv 5 3.23 .2: 3.2. .513: 12: 2 0. a 0 1 d ) 6 w m 0 ., own owe own 0mv omm cmm 02 om own omo 0mm 0? 0mm 0mm of 0m WL T Wiflli—q _ _ <_ _ .—_l