AC08CH21-Jackson
ARI
V I E W
14:48
Review in Advance first posted online on June 11, 2015. (Changes may still occur before final publication online and in print.)
A
N
I N
C E
S
R
E
26 May 2015
D V A
Forensic Mass Spectrometry Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
William D. Hoffmann1 and Glen P. Jackson1,2 1
Department of Forensic and Investigative Science and 2 C. Eugene Bennett Department of Chemistry, West Virginia University, Morgantown, West Virginia 26506; email: [email protected]
Annu. Rev. Anal. Chem. 2015. 8:21.1–21.22
Keywords
The Annual Review of Analytical Chemistry is online at anchem.annualreviews.org
field-portable mass spectrometers, ambient ionization, isotope ratio mass spectrometry, explosives, controlled substances, hair analysis
This article’s doi: 10.1146/annurev-anchem-071114-040335 c 2015 by Annual Reviews. Copyright All rights reserved
Abstract Developments in forensic mass spectrometry tend to follow, rather than lead, the developments in other disciplines. Techniques of great forensic utility or potential born independently of forensic applications include ambient ionization, imaging mass spectrometry, isotope ratio mass spectrometry, portable mass spectrometers, and hyphenated chromatography–mass spectrometry instruments, to name a few. Forensic science has the potential to benefit enormously from developments that are funded by other means, if only the infrastructure and personnel existed to adopt, validate, and implement the new technologies into casework. Perhaps one unique area in which forensic science is at the cutting edge is in the area of chemometrics and the determination of likelihood ratios for the evaluation of the weight of evidence. Such statistical techniques have been developed most extensively for ignitable-liquid residue analyses and isotope ratio analysis. This review attempts to capture the trends, motivating forces, and likely impact of developing areas of forensic mass spectrometry, with the caveat that none of this research is likely to have any real impact unless: (a) The instruments developed are turned into robust black boxes with red and green lights for positives and negatives, respectively, or (b) there are PhD graduates in the workforce who can help adopt these sophisticated techniques.
21.1
Changes may still occur before final publication online and in print
AC08CH21-Jackson
ARI
26 May 2015
14:48
1. INTRODUCTION
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
2009 NAS report: a detailed critique of traditional physical comparison methods; also hailed mass spectrometry as a mature discipline with sound chemical foundations
21.2
Forensic science operates within a complicated and difficult position; it must somehow simultaneously satisfy the scientific demands of the natural sciences, the judicial demands of the courts, the sensationalistic demands of the public, and the practical demands of the evidence. Students and administrators have financed an explosion in the number of forensic science degree programs; yet, remarkably, there are no PhD-granting institutions in forensic science in the United States with which to populate the growing number of tenure-track academic positions. The ivory towers have generally looked down on utility forensic science techniques (1), yet have not helped provide a leadership role in terms of providing adequate financial or scientific assistance. Despite these obstacles, and despite a practicing workforce that lacks the resources to help, forensic science research somehow inches along. Other review articles provide a more comprehensive and objective discussion of research articles in different areas of forensic chemistry (2, 3). This review intends to provide an unbiased but critical discussion of the status of forensic mass spectrometry (MS) research, and not at all an encyclopedic coverage of all the papers in the field. This article also attempts to capture and speak to the intangibles—the trends, motivating forces, and potential impact of developing areas of research. This review echoes others in the community with respect to the need for research and a research culture specific to forensic science (4–8). This is easier said than done. On the one hand, practitioners conduct the day-to-day analyses and are compelled to work with established technologies and within a rigid methodological framework. On the other hand, researchers and academicians, less involved in casework, conduct research into instrumentation and techniques with little regard for practical issues such as cost and time of analysis. What makes forensic science unique in this regard may be the unusually high barrier for new techniques to be adopted into casework. The foundation for this barrier is the understandable reluctance of practitioners to stray from established practices and protocols, even if new techniques provide useful capabilities or increase throughput and eliminate backlogs. In addition to financial, legal, and time constraints, the adoption of new techniques in crime laboratories is also limited by the absence of a research culture among laboratory personnel. Although many senior personnel have advanced degrees and have been active in research in the natural sciences (e.g., chemistry and biology), few can claim to have been cultured specifically in forensic science research; this situation should not continue. The United States has to develop research-based PhD programs in forensic science that serve the same goals as those of any other natural science discipline: (a) populate the next generation of teaching positions and (b) populate a workforce in which critical thinking skills serve as the foundation upon which hands-on specialized training can be added to the workforce. This is said to emphasize that whereas this review can help identify technologies and capabilities that ought to be of benefit to forensic practitioners, academicians and researchers are literally powerless to influence their adoption unless the workforce is populated with like-minded individuals. Addressing this lack of research culture has been the focus of multiple reports including the 2009 National Academy of Sciences (NAS) report (1). Prior to this report, a 2007 paper by Alan Jones concluded that the lack of incentive to write scholarly, peer-reviewed work by the bench practitioner led to the current disconnect between academic and practitioner communities (9). The relatively low impact factors of the five major forensic journals are a testament to the small size of the forensic research community and the lack of emphasis on the production of peer-reviewed articles (9). Whereas accredited forensic laboratories have strict standard operating procedures that they follow for different measurements, the diversity of state laws and operating modes of prosecutors
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
AC08CH21-Jackson
ARI
26 May 2015
14:48
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
results in methods and procedures that vary enormously among laboratories. Although the National Institute of Justice (NIJ) helped establish support working groups (SWGs) almost two decades ago to provide guidance to different forensic science disciplines, SWG guidelines are largely unenforceable. This situation is intended to change as the National Institute of Standards and Technology (NIST) is currently in the process of forming Organization for Scientific Area Committees (OSACs), which are intended to provide enforceable standards for forensic science practice. It is widely recognized that the first draft for each NIST standard will almost certainly derive from existing SWG guidelines, especially because the SWGs already have significant buy-in from practitioners.
2. FIELD-PORTABLE MASS SPECTROMETERS AND AMBIENT IONIZATION 2.1. Field-Portable Mass Spectrometers There is a level of irony in developing instruments and techniques for on-site forensic applications. For example, the take-home message of the NAS report (1) was to put the science back in the proverbial hands of scientists, but through on-site scientific instruments scientists are instead putting instruments into the physical hands of practitioners. NIJ offers research support for the validation of methods and instruments for portable instruments or on-site measurements. However, these small-scale R&D efforts typically fall short of addressing the entire implementation process: Crime scene technicians or first responders need to receive actual devices and be trained in the calibration, use, presentation, and defense of the technologies in court. Although the thought of trained technicians making on-site chemical analyses with sophisticated handheld chemical equipment scares almost all stakeholders in forensic science, plenty of police officers regularly implement breath alcohol tests with black boxes at the side of the road and then make important decisions based on those measurements. Field-portable techniques are of potential value to the forensic community because they can answer questions at the crime scene in pseudo real time and assist with evidence collection. Fieldportable techniques may also help alleviate backlogs by eliminating the collection and submission of samples that are of little or no evidentiary value. A number of field-portable techniques including Raman, infrared (IR), and X-ray fluorescence spectroscopy have been developed and are relatively mature. Although field-deployable mass spectrometers (10–14) are available and in use by certain military personnel and first responders, it will be several years before portable mass spectrometers make the same leap to everyday use (also see sidebar, Mass Spectrometry of Forensic DNA Polymerase Chain Reaction Products).
Ambient ionization: ion source for mass spectrometry in which analytes are ionized at atmospheric pressure with little or no sample preparation Direct analysis in real time (DART): uses a high-velocity plume of charged and excited species to volatilize and ionize semivolatile analytes on a surface Desorption electrospray ionization (DESI): uses a high velocity plume of charged droplets to pick up and ionize nonvolatile analyte molecules on a surface
2.2. Ambient Ionization Techniques MS methods of analysis that are gaining interest in various forensic applications are ambient ionization techniques such as direct analysis in real time (DART) and desorption electrospray ionization (DESI) (15–17). Figure 1 shows a schematic comparison of DESI and DART ion sources—the two complementary ion sources that have prompted enormous interest in ambient ionization technologies over the past decade (18). In DESI, an electrospray ionization (ESI) plume is directed at a sample surface and the desorbed droplets are sampled by a conventional, or slightly modified, atmospheric-sampling mass spectrometer (15). Common surfaces used for DESI include those made of polytetrafluorethylene for negative-mode analysis and polymethylmethacrylate for positive-mode analysis, although
www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.3
AC08CH21-Jackson
ARI
26 May 2015
14:48
MASS SPECTROMETRY OF FORENSIC DNA POLYMERASE CHAIN REACTION PRODUCTS
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
Forensic DNA analyses such as multiplex short tandem repeats or multiplex single nucleotide polymorphisms are primarily conducted on capillary electrophoresis (CE) instruments with separation times on the order of 20 min. Commercial CE systems typically analyze 16 samples in parallel, for an effective analysis time of ∼1.5 min per sample. However, using very similar polymerase chain reaction kits, but with modified primers, commercially available matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) analyzers can accomplish almost identical results in a small fraction of the time (149–151). With more validation, MALDI-TOFMS could be a serious contender to next-generation DNA analyzers.
almost any nonconductive surface will work (19, 20). Because the final multiply charged products of DESI are essentially the same as those from an ESI source, any mass spectrometer that can operate with an atmospheric pressure ionization source will work with DESI (20). DESI has been coupled with a variety of mass spectrometers including ion traps (20), Orbitraps (21), Fouriertransform ion cyclotron resonance (22), and ion mobility spectrometry (IMS) time-of-flight (TOF) (23) mass spectrometers. Although DESI can analyze molecules up to 66,000 Da (20, 24), this range is usually unnecessary in forensic applications where small molecules prevail. DESI Electrical potential kV
Solvent Nebulizing gas
Charged solvent spray
Vacuum interface and mass spectrometer
Gas jet Analyte ion desorption
Substrate
Analytes
He/N2
Discharge chamber Heating
Metastable He/N2
Vacuum interface and mass spectrometer Ion/neutral separation
Figure 1 Desorption electrospray ionization (top) and direct analysis in real time (bottom) analyses for ambient high-throughput mass spectrometric analysis of unprepared samples (skin, bricks, urine spots, clothing, c 2006, American Association for the tissue, etc.). Reprinted with permission from Reference 18. Copyright Advancement of Science. 21.4
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
a
b
c
d
Figure 2 (a) Desorption electrospray ionization (DESI) image of distribution of cocaine on a latent fingerprint blotted on glass. (b) Computer-generated fingerprint from DESI image. (c) Ink fingerprint blotted on paper and optically scanned. (d ) Computer-generated fingerprint from optical image. Some of the automatically detected points of interest (minutiae) are represented by red and blue dots in panels b and d, respectively. c 2008, American Association for the Figure reprinted with permission from Reference 33. Copyright Advancement of Science.
has been used in a variety of forensic applications including the detection of illicit drugs (25–27), explosives (28–30), alkaloids in plant matter (31), and imaging and analysis of latent fingerprints (32, 33). Figure 2 shows one of the first examples of imaging-specific chemical signatures in friction ridge residues using DESI (33). This example demonstrates the capability and type of application that ambient imaging methods could have, even if this particular example is a low-frequency problem. The US Drug Enforcement Administration (DEA) purchased and tested an early commercial DESI source, but because the DEA does not routinely publish or advertise everything they learn, the research community is left in the dark as to how successful the testing was (also see sidebar, Real-Time Drug Monitoring and Ambient Sampling Are Not New Ideas). After trying to develop an atmospheric pressure thermal electron source, Cody et al. (16) provided the initial report on DART. DART typically forms singly charged, protonated, or deprotonated species in either the positive or negative mode, respectively, but in some cases, it can also form radical ions (34). Ionization in DART normally occurs through proton transfer reactions, Penning ionization, or electron capture (35). Unlike DESI, DART is typically limited to analytes with a molecular weight below 800 Da (20, 34, 36, 37). Since its inception, DART has also been used for a variety of forensic applications including chemical warfare agents (16), explosives (38), www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.5
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
Figure 3 Schematic representation of the setup used by Green for real-time drug monitoring and ambient sampling of drugs of abuse. Reproduced from reference (152).
REAL-TIME DRUG MONITORING AND AMBIENT SAMPLING ARE NOT NEW IDEAS In 1972, Green (152) demonstrated that an ambient-sampling mass spectrometer (Figure 3) could detect volatile organics, like alcohol permeating through a subject’s arm within 100 s of ingesting an alcoholic beverage. The alcohol signal peaked around 45 min and lasted for more than 80 minutes. The same instrument set up was used to sniff vapors of aqueous solutions and powder forms of various drugs. A schematic of the setup is shown in Figure 3.
drugs (39–41), and ignitable liquids (16). There are limited demonstrations of imaging with DART, and it remains to be seen if the imaging capabilities can match those of DESI. NIJ is currently providing some funding for DART-MS library development and validation of DART ion sources, and several crime labs have begun using DART to prescreen samples for decision-making purposes. When SWGDRUG or the new OSACs weigh in on ambient ionization methods, they are likely to demand tandem mass spectrometry (MS/MS) for structural/compound identification as opposed to high mass accuracy. Commercially available DART-TOF-MS instruments can perform pseudoMS2 , but in the absence of precursor isolation and in the presence of other precursor ions, the assignment of specific product ions to specific precursor ions lacks reliability. However, with ion traps, linear ion traps, or triple quadrupole mass analyzers, precursor ion isolation is possible, so the assignment of specific product ions to specific precursors is strongly reliable.
3. IGNITABLE-LIQUID RESIDUES Between the United States and the United Kingdom, there are more than 330,000 arson cases reported annually (42, 43). Arson is responsible for 13% of the fires handled annually in the United States (42). The majority of arson cases in the United States involve the use of an ignitable liquid, and the majority of those use gasoline—the most abundant and accessible fuel for criminals. In 1959, Joseph Nicol, a firearms technician at the Chicago police crime lab, highlighted the potential of MS for identifying small quantities of volatile liquid recovered from a fire. Even then, 21.6
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
Nicol (44) recognized that large universities or oil companies would probably have to run the tests, as those were the only organizations that could afford a mass spectrometer at the time. It took more than a decade before the technique gained broad acceptance (45–48). MS introduced an independent dimension of analysis to gas chromatography (GC), which enabled the determination of coeluting substances and identities of compounds with higher confidence than provided by any other detector. Although GC/MS instruments have undergone technological improvements and now have vastly better limits of detection, the fundamental principles have remained unchanged in more than 50 years. However, the field is not stagnant. Significant changes continue to occur in the areas of sample preparation and data analysis (49). Several academic research groups have also made significant progress in automated interpretation of GC/MS data and the classification models for comparing questioned samples with reference materials and classes defined by ASTM (American Society for Testing and Materials) (e.g. ASTM E1618-11) (50). ASTM methods are available for extracting ignitable-liquid residues from fire debris, and static headspace extraction using activated charcoal strips is by far the most widely used technique, despite the very limited number of suppliers of activated carbon strips in the United States (we know of only two, and one might repackage the product of the other), which might explain the very high cost of >$5 per 10 × 20 mm strip. In contrast to the manual pattern matching suggested in ASTM E1618-11 [the NAS report (1) cautioned against subjective pattern-matching techniques], various chemometric approaches are reliable for classification and discrimination of ignitable liquids (51–55). In a totally different and perhaps simpler approach, Sigman’s group (56–59) has shown that a summed mass spectrum approach–essentially averaging mass spectra across an entire GC chromatogram and ignoring the chromatographic information–can effectively classify ignitable liquids according to the widely accepted ASTM classification scheme. In the absence of a strong external motivating force (such as the NIST OSACs), it seems unlikely that practitioners will ever receive the training or resources necessary to understand, validate, and implement these chemometric and automated approaches. This is unfortunate and a classic example of the chasm between research and practice in forensic science. This separation is present even in traditional and uncontroversial techniques such as GC/MS, which is often regarded as a gold standard of forensic laboratory techniques. In other areas of ignitable-liquid residue analysis, Turner & Goodpaster (60–63) showed through a series of experiments that biodegradation can alter the relative distribution of components in petroleum distillates. Geologists have long understood the importance of biological activity on hydrocarbon matter (64–66), but these recent studies emphasize the relevance to trace levels and on vastly shorter timescales. The results indicate that soil bacteria generally favor unbranched alkanes and lesser-functionalized aromatics as energy sources.
4. EXPLOSIVES Over the past 25 years, the detection of explosives and energetic materials has been a major research emphasis (67). Improvised explosive devices have become an effective weapon in the arsenal of terrorist organizations for use against civilians and military personnel as well as strategic infrastructure targets (67). The detection of explosives, explosive residues, and energetic materials are also of concern in military testing grounds, where exploded and unexploded ordinances can release significant quantities of explosives into the environment. Cyclotrimethylenetrinitramine (RDX), cyclotetramethylene-tetranitramine, trinitrotoluene (TNT) (68), and smokeless powders (69) are of particular interest because of their wide availability, relatively simple synthetic routes, and prior history of use in terrorist attacks. www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.7
ARI
26 May 2015
14:48
In 2002, Evans et al. (70) described a liquid chromatography (LC) separation of RDX, TNT, nitroglycerin, and pentaerythritol tetranitrate (PETN). Using atmospheric pressure chemical ionization, they studied the negative-ion mass spectra with and without the addition of a dopant, ammonium chloride. RDX response was particularly sensitive to the addition of ammonium chloride. When the NH4 Cl dopant was added, the response to RDX improved and the stable isotopes of the chloride adduct [M + Cl]− at m/z 257 and 259 were significantly enhanced. The wide availability of smokeless powders makes them particularly ubiquitous as the primary charge in improvised explosive devices (69). In an attempt to improve the detection of smokelesspowder components, de Perre et al. (69) used capillary electrochromatography combined with UV-visible spectroscopy and TOF-MS to detect the different explosives. Contrary to traditional capillary electrophoresis, capillary electrochromatography relies on a chromatographic stationary phase to separate analytes based on their retention factor and mobility, instead of just their mobilities. The goal was to classify smokeless powders on the basis of the various components present in the additive packages combined with the nitrocellulose base. The high resolution of the monolithic columns used in the capillary electrochromatographic method facilitated separation of the isomeric forms of dinitrotoluene, whereas the high mass accuracy afforded by TOF-MS allowed the unambiguous determination of many smokeless-powder components. Figure 4 shows the extracted ion chromatograms for 14 components present in smokeless powders analyzed in positive and negative mode using capillary electrochromatography (69). Hubert et al. (71) recently used high-resolution, high mass accuracy measurements to characterize the 30-Da losses during the fragmentation of TNT. Surprisingly, [M]−• and the [M − H]− ions have distinct fragmentation channels that have often been mislabeled. Collision-induced dissociation of the [M]−• ion at m/z 227 results in two dominant product ion peaks at m/z 210,
4 1.4 × 106
9, 10
2 5
1.2 × 106 1.0 × 106
Intensity (cps)
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
0.8 × 106
3
0.6 × 106 0.4 ×
11
1
106
12
67
13 14
0.2 × 106 0.0 ×
106
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Time (min)
m/z 288.9 to 289.1 m/z 243.9 to 244.2 m/z 170.0 to 170.2 m/z 215.0 to 215.2 m/z 279.1 to 279.2 m/z 199.0 to 199.1 m/z 269.1 to 269.2 m/z 223.0 to 223.2 m/z 241.1 to 241.2 m/z 195.0 to 195.1
ESI–
ESI+
Figure 4 Extracted ion chromatograms at specific m/z values of a mixture of 14 smokeless-powder components: (1) dimethylphthalate (DMP), (2) methyl centralite (MC), (3) diethylphthalate (DEP), (4) nitroglycerin (NG), (5) ethyl centralite (EC), (6) 3,4-dinitrotoluene (3,4-DNT), (7) 2,4-dinitrotoluene (2,4-DNT), (9) 4-nitrosodiphenylamine (4-NsDPA), (10) N-nitrosodiphenylamine (N-NsDPA), (11) dibutylphthalate (DBP), (12) 4-nitrodiphenylamine (4-NDPA), (13) diphenylamine (DPA), and (14) 2-nitrodiphenylamine c 2012 Elsevier. Abbreviation: ESI, electrospray ionization. (2-NPDA). Reprinted with permission from Reference 69. Copyright 21.8
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
corresponding to [M − OH]− , and at m/z 197, corresponding to [M − NO]− . Collision-induced dissociation of the [M − H]− precursor at m/z 226 results in a dominant product ion peak at m/z 196 (loss of 30 Da), which has always been assumed to be the neutral loss of NO• when fragmenting the [M]−• precursor ion. However, high-resolution, high mass accuracy measurements conducted in an Orbitrap mass analyzer revealed that this 30-Da loss instead comes from the unexpected loss of CH2 O. Single-point energy calculations revealed that loss of formaldehyde, although endothermic from the [M]−• precursor, is kinetically favorable to the loss of NO• , which has a relatively high barrier to isomerization prior to a stepwise reaction. Clemons et al. (72) used direct analyte-probed nanoextraction (DAPNe) coupled with DARTMS to extract and detect explosive residues. Previously applied to trace particulates found on fibers (73), micrometer-sized analyte particulates are dissolved in a solvent system that is subsequently aspirated into a nanoelectrospray capillary and subjected to MS analysis. Trace residues of TNT, RDX, and 1-methylaminoanthraquinone (MAAQ) were successfully detected from various surfaces (72). Analytes were extracted from metal surfaces and glass slides with particulates varying in size from
ARI
V I E W
14:48
Review in Advance first posted online on June 11, 2015. (Changes may still occur before final publication online and in print.)
A
N
I N
C E
S
R
E
26 May 2015
D V A
Forensic Mass Spectrometry Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
William D. Hoffmann1 and Glen P. Jackson1,2 1
Department of Forensic and Investigative Science and 2 C. Eugene Bennett Department of Chemistry, West Virginia University, Morgantown, West Virginia 26506; email: [email protected]
Annu. Rev. Anal. Chem. 2015. 8:21.1–21.22
Keywords
The Annual Review of Analytical Chemistry is online at anchem.annualreviews.org
field-portable mass spectrometers, ambient ionization, isotope ratio mass spectrometry, explosives, controlled substances, hair analysis
This article’s doi: 10.1146/annurev-anchem-071114-040335 c 2015 by Annual Reviews. Copyright All rights reserved
Abstract Developments in forensic mass spectrometry tend to follow, rather than lead, the developments in other disciplines. Techniques of great forensic utility or potential born independently of forensic applications include ambient ionization, imaging mass spectrometry, isotope ratio mass spectrometry, portable mass spectrometers, and hyphenated chromatography–mass spectrometry instruments, to name a few. Forensic science has the potential to benefit enormously from developments that are funded by other means, if only the infrastructure and personnel existed to adopt, validate, and implement the new technologies into casework. Perhaps one unique area in which forensic science is at the cutting edge is in the area of chemometrics and the determination of likelihood ratios for the evaluation of the weight of evidence. Such statistical techniques have been developed most extensively for ignitable-liquid residue analyses and isotope ratio analysis. This review attempts to capture the trends, motivating forces, and likely impact of developing areas of forensic mass spectrometry, with the caveat that none of this research is likely to have any real impact unless: (a) The instruments developed are turned into robust black boxes with red and green lights for positives and negatives, respectively, or (b) there are PhD graduates in the workforce who can help adopt these sophisticated techniques.
21.1
Changes may still occur before final publication online and in print
AC08CH21-Jackson
ARI
26 May 2015
14:48
1. INTRODUCTION
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
2009 NAS report: a detailed critique of traditional physical comparison methods; also hailed mass spectrometry as a mature discipline with sound chemical foundations
21.2
Forensic science operates within a complicated and difficult position; it must somehow simultaneously satisfy the scientific demands of the natural sciences, the judicial demands of the courts, the sensationalistic demands of the public, and the practical demands of the evidence. Students and administrators have financed an explosion in the number of forensic science degree programs; yet, remarkably, there are no PhD-granting institutions in forensic science in the United States with which to populate the growing number of tenure-track academic positions. The ivory towers have generally looked down on utility forensic science techniques (1), yet have not helped provide a leadership role in terms of providing adequate financial or scientific assistance. Despite these obstacles, and despite a practicing workforce that lacks the resources to help, forensic science research somehow inches along. Other review articles provide a more comprehensive and objective discussion of research articles in different areas of forensic chemistry (2, 3). This review intends to provide an unbiased but critical discussion of the status of forensic mass spectrometry (MS) research, and not at all an encyclopedic coverage of all the papers in the field. This article also attempts to capture and speak to the intangibles—the trends, motivating forces, and potential impact of developing areas of research. This review echoes others in the community with respect to the need for research and a research culture specific to forensic science (4–8). This is easier said than done. On the one hand, practitioners conduct the day-to-day analyses and are compelled to work with established technologies and within a rigid methodological framework. On the other hand, researchers and academicians, less involved in casework, conduct research into instrumentation and techniques with little regard for practical issues such as cost and time of analysis. What makes forensic science unique in this regard may be the unusually high barrier for new techniques to be adopted into casework. The foundation for this barrier is the understandable reluctance of practitioners to stray from established practices and protocols, even if new techniques provide useful capabilities or increase throughput and eliminate backlogs. In addition to financial, legal, and time constraints, the adoption of new techniques in crime laboratories is also limited by the absence of a research culture among laboratory personnel. Although many senior personnel have advanced degrees and have been active in research in the natural sciences (e.g., chemistry and biology), few can claim to have been cultured specifically in forensic science research; this situation should not continue. The United States has to develop research-based PhD programs in forensic science that serve the same goals as those of any other natural science discipline: (a) populate the next generation of teaching positions and (b) populate a workforce in which critical thinking skills serve as the foundation upon which hands-on specialized training can be added to the workforce. This is said to emphasize that whereas this review can help identify technologies and capabilities that ought to be of benefit to forensic practitioners, academicians and researchers are literally powerless to influence their adoption unless the workforce is populated with like-minded individuals. Addressing this lack of research culture has been the focus of multiple reports including the 2009 National Academy of Sciences (NAS) report (1). Prior to this report, a 2007 paper by Alan Jones concluded that the lack of incentive to write scholarly, peer-reviewed work by the bench practitioner led to the current disconnect between academic and practitioner communities (9). The relatively low impact factors of the five major forensic journals are a testament to the small size of the forensic research community and the lack of emphasis on the production of peer-reviewed articles (9). Whereas accredited forensic laboratories have strict standard operating procedures that they follow for different measurements, the diversity of state laws and operating modes of prosecutors
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
AC08CH21-Jackson
ARI
26 May 2015
14:48
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
results in methods and procedures that vary enormously among laboratories. Although the National Institute of Justice (NIJ) helped establish support working groups (SWGs) almost two decades ago to provide guidance to different forensic science disciplines, SWG guidelines are largely unenforceable. This situation is intended to change as the National Institute of Standards and Technology (NIST) is currently in the process of forming Organization for Scientific Area Committees (OSACs), which are intended to provide enforceable standards for forensic science practice. It is widely recognized that the first draft for each NIST standard will almost certainly derive from existing SWG guidelines, especially because the SWGs already have significant buy-in from practitioners.
2. FIELD-PORTABLE MASS SPECTROMETERS AND AMBIENT IONIZATION 2.1. Field-Portable Mass Spectrometers There is a level of irony in developing instruments and techniques for on-site forensic applications. For example, the take-home message of the NAS report (1) was to put the science back in the proverbial hands of scientists, but through on-site scientific instruments scientists are instead putting instruments into the physical hands of practitioners. NIJ offers research support for the validation of methods and instruments for portable instruments or on-site measurements. However, these small-scale R&D efforts typically fall short of addressing the entire implementation process: Crime scene technicians or first responders need to receive actual devices and be trained in the calibration, use, presentation, and defense of the technologies in court. Although the thought of trained technicians making on-site chemical analyses with sophisticated handheld chemical equipment scares almost all stakeholders in forensic science, plenty of police officers regularly implement breath alcohol tests with black boxes at the side of the road and then make important decisions based on those measurements. Field-portable techniques are of potential value to the forensic community because they can answer questions at the crime scene in pseudo real time and assist with evidence collection. Fieldportable techniques may also help alleviate backlogs by eliminating the collection and submission of samples that are of little or no evidentiary value. A number of field-portable techniques including Raman, infrared (IR), and X-ray fluorescence spectroscopy have been developed and are relatively mature. Although field-deployable mass spectrometers (10–14) are available and in use by certain military personnel and first responders, it will be several years before portable mass spectrometers make the same leap to everyday use (also see sidebar, Mass Spectrometry of Forensic DNA Polymerase Chain Reaction Products).
Ambient ionization: ion source for mass spectrometry in which analytes are ionized at atmospheric pressure with little or no sample preparation Direct analysis in real time (DART): uses a high-velocity plume of charged and excited species to volatilize and ionize semivolatile analytes on a surface Desorption electrospray ionization (DESI): uses a high velocity plume of charged droplets to pick up and ionize nonvolatile analyte molecules on a surface
2.2. Ambient Ionization Techniques MS methods of analysis that are gaining interest in various forensic applications are ambient ionization techniques such as direct analysis in real time (DART) and desorption electrospray ionization (DESI) (15–17). Figure 1 shows a schematic comparison of DESI and DART ion sources—the two complementary ion sources that have prompted enormous interest in ambient ionization technologies over the past decade (18). In DESI, an electrospray ionization (ESI) plume is directed at a sample surface and the desorbed droplets are sampled by a conventional, or slightly modified, atmospheric-sampling mass spectrometer (15). Common surfaces used for DESI include those made of polytetrafluorethylene for negative-mode analysis and polymethylmethacrylate for positive-mode analysis, although
www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.3
AC08CH21-Jackson
ARI
26 May 2015
14:48
MASS SPECTROMETRY OF FORENSIC DNA POLYMERASE CHAIN REACTION PRODUCTS
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
Forensic DNA analyses such as multiplex short tandem repeats or multiplex single nucleotide polymorphisms are primarily conducted on capillary electrophoresis (CE) instruments with separation times on the order of 20 min. Commercial CE systems typically analyze 16 samples in parallel, for an effective analysis time of ∼1.5 min per sample. However, using very similar polymerase chain reaction kits, but with modified primers, commercially available matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) analyzers can accomplish almost identical results in a small fraction of the time (149–151). With more validation, MALDI-TOFMS could be a serious contender to next-generation DNA analyzers.
almost any nonconductive surface will work (19, 20). Because the final multiply charged products of DESI are essentially the same as those from an ESI source, any mass spectrometer that can operate with an atmospheric pressure ionization source will work with DESI (20). DESI has been coupled with a variety of mass spectrometers including ion traps (20), Orbitraps (21), Fouriertransform ion cyclotron resonance (22), and ion mobility spectrometry (IMS) time-of-flight (TOF) (23) mass spectrometers. Although DESI can analyze molecules up to 66,000 Da (20, 24), this range is usually unnecessary in forensic applications where small molecules prevail. DESI Electrical potential kV
Solvent Nebulizing gas
Charged solvent spray
Vacuum interface and mass spectrometer
Gas jet Analyte ion desorption
Substrate
Analytes
He/N2
Discharge chamber Heating
Metastable He/N2
Vacuum interface and mass spectrometer Ion/neutral separation
Figure 1 Desorption electrospray ionization (top) and direct analysis in real time (bottom) analyses for ambient high-throughput mass spectrometric analysis of unprepared samples (skin, bricks, urine spots, clothing, c 2006, American Association for the tissue, etc.). Reprinted with permission from Reference 18. Copyright Advancement of Science. 21.4
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
a
b
c
d
Figure 2 (a) Desorption electrospray ionization (DESI) image of distribution of cocaine on a latent fingerprint blotted on glass. (b) Computer-generated fingerprint from DESI image. (c) Ink fingerprint blotted on paper and optically scanned. (d ) Computer-generated fingerprint from optical image. Some of the automatically detected points of interest (minutiae) are represented by red and blue dots in panels b and d, respectively. c 2008, American Association for the Figure reprinted with permission from Reference 33. Copyright Advancement of Science.
has been used in a variety of forensic applications including the detection of illicit drugs (25–27), explosives (28–30), alkaloids in plant matter (31), and imaging and analysis of latent fingerprints (32, 33). Figure 2 shows one of the first examples of imaging-specific chemical signatures in friction ridge residues using DESI (33). This example demonstrates the capability and type of application that ambient imaging methods could have, even if this particular example is a low-frequency problem. The US Drug Enforcement Administration (DEA) purchased and tested an early commercial DESI source, but because the DEA does not routinely publish or advertise everything they learn, the research community is left in the dark as to how successful the testing was (also see sidebar, Real-Time Drug Monitoring and Ambient Sampling Are Not New Ideas). After trying to develop an atmospheric pressure thermal electron source, Cody et al. (16) provided the initial report on DART. DART typically forms singly charged, protonated, or deprotonated species in either the positive or negative mode, respectively, but in some cases, it can also form radical ions (34). Ionization in DART normally occurs through proton transfer reactions, Penning ionization, or electron capture (35). Unlike DESI, DART is typically limited to analytes with a molecular weight below 800 Da (20, 34, 36, 37). Since its inception, DART has also been used for a variety of forensic applications including chemical warfare agents (16), explosives (38), www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.5
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
Figure 3 Schematic representation of the setup used by Green for real-time drug monitoring and ambient sampling of drugs of abuse. Reproduced from reference (152).
REAL-TIME DRUG MONITORING AND AMBIENT SAMPLING ARE NOT NEW IDEAS In 1972, Green (152) demonstrated that an ambient-sampling mass spectrometer (Figure 3) could detect volatile organics, like alcohol permeating through a subject’s arm within 100 s of ingesting an alcoholic beverage. The alcohol signal peaked around 45 min and lasted for more than 80 minutes. The same instrument set up was used to sniff vapors of aqueous solutions and powder forms of various drugs. A schematic of the setup is shown in Figure 3.
drugs (39–41), and ignitable liquids (16). There are limited demonstrations of imaging with DART, and it remains to be seen if the imaging capabilities can match those of DESI. NIJ is currently providing some funding for DART-MS library development and validation of DART ion sources, and several crime labs have begun using DART to prescreen samples for decision-making purposes. When SWGDRUG or the new OSACs weigh in on ambient ionization methods, they are likely to demand tandem mass spectrometry (MS/MS) for structural/compound identification as opposed to high mass accuracy. Commercially available DART-TOF-MS instruments can perform pseudoMS2 , but in the absence of precursor isolation and in the presence of other precursor ions, the assignment of specific product ions to specific precursor ions lacks reliability. However, with ion traps, linear ion traps, or triple quadrupole mass analyzers, precursor ion isolation is possible, so the assignment of specific product ions to specific precursors is strongly reliable.
3. IGNITABLE-LIQUID RESIDUES Between the United States and the United Kingdom, there are more than 330,000 arson cases reported annually (42, 43). Arson is responsible for 13% of the fires handled annually in the United States (42). The majority of arson cases in the United States involve the use of an ignitable liquid, and the majority of those use gasoline—the most abundant and accessible fuel for criminals. In 1959, Joseph Nicol, a firearms technician at the Chicago police crime lab, highlighted the potential of MS for identifying small quantities of volatile liquid recovered from a fire. Even then, 21.6
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
Nicol (44) recognized that large universities or oil companies would probably have to run the tests, as those were the only organizations that could afford a mass spectrometer at the time. It took more than a decade before the technique gained broad acceptance (45–48). MS introduced an independent dimension of analysis to gas chromatography (GC), which enabled the determination of coeluting substances and identities of compounds with higher confidence than provided by any other detector. Although GC/MS instruments have undergone technological improvements and now have vastly better limits of detection, the fundamental principles have remained unchanged in more than 50 years. However, the field is not stagnant. Significant changes continue to occur in the areas of sample preparation and data analysis (49). Several academic research groups have also made significant progress in automated interpretation of GC/MS data and the classification models for comparing questioned samples with reference materials and classes defined by ASTM (American Society for Testing and Materials) (e.g. ASTM E1618-11) (50). ASTM methods are available for extracting ignitable-liquid residues from fire debris, and static headspace extraction using activated charcoal strips is by far the most widely used technique, despite the very limited number of suppliers of activated carbon strips in the United States (we know of only two, and one might repackage the product of the other), which might explain the very high cost of >$5 per 10 × 20 mm strip. In contrast to the manual pattern matching suggested in ASTM E1618-11 [the NAS report (1) cautioned against subjective pattern-matching techniques], various chemometric approaches are reliable for classification and discrimination of ignitable liquids (51–55). In a totally different and perhaps simpler approach, Sigman’s group (56–59) has shown that a summed mass spectrum approach–essentially averaging mass spectra across an entire GC chromatogram and ignoring the chromatographic information–can effectively classify ignitable liquids according to the widely accepted ASTM classification scheme. In the absence of a strong external motivating force (such as the NIST OSACs), it seems unlikely that practitioners will ever receive the training or resources necessary to understand, validate, and implement these chemometric and automated approaches. This is unfortunate and a classic example of the chasm between research and practice in forensic science. This separation is present even in traditional and uncontroversial techniques such as GC/MS, which is often regarded as a gold standard of forensic laboratory techniques. In other areas of ignitable-liquid residue analysis, Turner & Goodpaster (60–63) showed through a series of experiments that biodegradation can alter the relative distribution of components in petroleum distillates. Geologists have long understood the importance of biological activity on hydrocarbon matter (64–66), but these recent studies emphasize the relevance to trace levels and on vastly shorter timescales. The results indicate that soil bacteria generally favor unbranched alkanes and lesser-functionalized aromatics as energy sources.
4. EXPLOSIVES Over the past 25 years, the detection of explosives and energetic materials has been a major research emphasis (67). Improvised explosive devices have become an effective weapon in the arsenal of terrorist organizations for use against civilians and military personnel as well as strategic infrastructure targets (67). The detection of explosives, explosive residues, and energetic materials are also of concern in military testing grounds, where exploded and unexploded ordinances can release significant quantities of explosives into the environment. Cyclotrimethylenetrinitramine (RDX), cyclotetramethylene-tetranitramine, trinitrotoluene (TNT) (68), and smokeless powders (69) are of particular interest because of their wide availability, relatively simple synthetic routes, and prior history of use in terrorist attacks. www.annualreviews.org • Forensic Mass Spectrometry
Changes may still occur before final publication online and in print
21.7
ARI
26 May 2015
14:48
In 2002, Evans et al. (70) described a liquid chromatography (LC) separation of RDX, TNT, nitroglycerin, and pentaerythritol tetranitrate (PETN). Using atmospheric pressure chemical ionization, they studied the negative-ion mass spectra with and without the addition of a dopant, ammonium chloride. RDX response was particularly sensitive to the addition of ammonium chloride. When the NH4 Cl dopant was added, the response to RDX improved and the stable isotopes of the chloride adduct [M + Cl]− at m/z 257 and 259 were significantly enhanced. The wide availability of smokeless powders makes them particularly ubiquitous as the primary charge in improvised explosive devices (69). In an attempt to improve the detection of smokelesspowder components, de Perre et al. (69) used capillary electrochromatography combined with UV-visible spectroscopy and TOF-MS to detect the different explosives. Contrary to traditional capillary electrophoresis, capillary electrochromatography relies on a chromatographic stationary phase to separate analytes based on their retention factor and mobility, instead of just their mobilities. The goal was to classify smokeless powders on the basis of the various components present in the additive packages combined with the nitrocellulose base. The high resolution of the monolithic columns used in the capillary electrochromatographic method facilitated separation of the isomeric forms of dinitrotoluene, whereas the high mass accuracy afforded by TOF-MS allowed the unambiguous determination of many smokeless-powder components. Figure 4 shows the extracted ion chromatograms for 14 components present in smokeless powders analyzed in positive and negative mode using capillary electrochromatography (69). Hubert et al. (71) recently used high-resolution, high mass accuracy measurements to characterize the 30-Da losses during the fragmentation of TNT. Surprisingly, [M]−• and the [M − H]− ions have distinct fragmentation channels that have often been mislabeled. Collision-induced dissociation of the [M]−• ion at m/z 227 results in two dominant product ion peaks at m/z 210,
4 1.4 × 106
9, 10
2 5
1.2 × 106 1.0 × 106
Intensity (cps)
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
0.8 × 106
3
0.6 × 106 0.4 ×
11
1
106
12
67
13 14
0.2 × 106 0.0 ×
106
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Time (min)
m/z 288.9 to 289.1 m/z 243.9 to 244.2 m/z 170.0 to 170.2 m/z 215.0 to 215.2 m/z 279.1 to 279.2 m/z 199.0 to 199.1 m/z 269.1 to 269.2 m/z 223.0 to 223.2 m/z 241.1 to 241.2 m/z 195.0 to 195.1
ESI–
ESI+
Figure 4 Extracted ion chromatograms at specific m/z values of a mixture of 14 smokeless-powder components: (1) dimethylphthalate (DMP), (2) methyl centralite (MC), (3) diethylphthalate (DEP), (4) nitroglycerin (NG), (5) ethyl centralite (EC), (6) 3,4-dinitrotoluene (3,4-DNT), (7) 2,4-dinitrotoluene (2,4-DNT), (9) 4-nitrosodiphenylamine (4-NsDPA), (10) N-nitrosodiphenylamine (N-NsDPA), (11) dibutylphthalate (DBP), (12) 4-nitrodiphenylamine (4-NDPA), (13) diphenylamine (DPA), and (14) 2-nitrodiphenylamine c 2012 Elsevier. Abbreviation: ESI, electrospray ionization. (2-NPDA). Reprinted with permission from Reference 69. Copyright 21.8
Hoffmann
·
Jackson
Changes may still occur before final publication online and in print
Annual Review of Analytical Chemistry 2015.8. Downloaded from www.annualreviews.org Access provided by New York University - Bobst Library on 07/04/15. For personal use only.
AC08CH21-Jackson
ARI
26 May 2015
14:48
corresponding to [M − OH]− , and at m/z 197, corresponding to [M − NO]− . Collision-induced dissociation of the [M − H]− precursor at m/z 226 results in a dominant product ion peak at m/z 196 (loss of 30 Da), which has always been assumed to be the neutral loss of NO• when fragmenting the [M]−• precursor ion. However, high-resolution, high mass accuracy measurements conducted in an Orbitrap mass analyzer revealed that this 30-Da loss instead comes from the unexpected loss of CH2 O. Single-point energy calculations revealed that loss of formaldehyde, although endothermic from the [M]−• precursor, is kinetically favorable to the loss of NO• , which has a relatively high barrier to isomerization prior to a stepwise reaction. Clemons et al. (72) used direct analyte-probed nanoextraction (DAPNe) coupled with DARTMS to extract and detect explosive residues. Previously applied to trace particulates found on fibers (73), micrometer-sized analyte particulates are dissolved in a solvent system that is subsequently aspirated into a nanoelectrospray capillary and subjected to MS analysis. Trace residues of TNT, RDX, and 1-methylaminoanthraquinone (MAAQ) were successfully detected from various surfaces (72). Analytes were extracted from metal surfaces and glass slides with particulates varying in size from
