- Email: [email protected]
time-of-flight mass spectrometry
Forensic Science International 235 (2014) 68–77
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Chemical analysis of pharmaceuticals and explosives in fingermarks using matrix-assisted laser desorption ionization/time-of-flight mass spectrometry Kimberly Kaplan-Sandquist a, Marc A. LeBeau b, Mark L. Miller c,* a Counterterrorism and Forensic Science Research Unit, Visiting Scientist Program, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States b Scientific Analysis Section, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States c Counterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 July 2013 Received in revised form 19 November 2013 Accepted 27 November 2013 Available online 16 December 2013
Chemical analysis of latent fingermarks, ‘‘touch chemistry,’’ has the potential of providing intelligence or forensically relevant information. Matrix-assisted laser desorption ionization/time-of-flight mass spectrometry (MALDI/TOF MS) was used as an analytical platform for obtaining mass spectra and chemical images of target drugs and explosives in fingermark residues following conventional fingerprint development methods and MALDI matrix processing. There were two main purposes of this research: (1) develop effective laboratory methods for detecting drugs and explosives in fingermark residues and (2) determine the feasibility of detecting drugs and explosives after casual contact with pills, powders, and residues. Further, synthetic latent print reference pads were evaluated as mimics of natural fingermark residue to determine if the pads could be used for method development and quality control. The results suggest that artificial amino acid and sebaceous oil residue pads are not suitable to adequately simulate natural fingermark chemistry for MALDI/TOF MS analysis. However, the pads were useful for designing experiments and setting instrumental parameters. Based on the natural fingermark residue experiments, handling whole or broken pills did not transfer sufficient quantities of drugs to allow for definitive detection. Transferring drugs or explosives in the form of powders and residues was successful for preparing analytes for detection after contact with fingers and deposition of fingermark residue. One downfall to handling powders was that the analyte particles were easily spread beyond the original fingermark during development. Analyte particles were confined in the original fingermark when using transfer residues. The MALDI/TOF MS was able to detect procaine, pseudoephedrine, TNT, and RDX from contact residue under laboratory conditions with the integration of conventional fingerprint development methods and MALDI matrix. MALDI/TOF MS is a nondestructive technique which provides chemical information in both the mass spectra and chemical images. Published by Elsevier Ireland Ltd.
Keywords: Touch chemistry Latent fingerprint Fingerprint powder Chemical imaging Drugs Cyanoacrylate fuming
1. Introduction ‘‘Touch chemistry’’ is a forensic technique that could provide investigative leads by analyzing chemical data from fingermark residue. Chemical intelligence gained from the analysis of latent fingermarks has the potential to provide information about the suspect even if the fingermarks are smudged or the patterns cannot
* Corresponding author at: Counterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, Quantico, 2501 Investigation Parkway, VA 22135, United States. Tel.: +1 703 632 7846; fax: +1 703 632 7801. E-mail address: [email protected] (M.L. Miller). 0379-0738/$ – see front matter . Published by Elsevier Ireland Ltd. http://dx.doi.org/10.1016/j.forsciint.2013.11.016
be matched through an automated fingerprint search. In order to collect latent fingermarks at a crime scene, the prints must first be made visible using development techniques such as fingerprint powder or cyanoacrylate fuming [1]. Once fingermarks are located, chemical information to be gained about the individual could include exposure to illicit substances/explosives, drug-use history, gender, and approximate age [2–8]. Past approaches for fingermark residue characterization include Fourier transform-infrared spectroscopy (FTIR) [9], gas chromatography mass spectrometry (GC/MS) [10], fluorescence microscopy [11], desorption electrospray ionization mass spectrometry (DESI/MS) [12], and secondary ion mass spectrometry (SIMS) [13,14]. An alternative approach that has recently
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
demonstrated high potential for characterizing a broad array of endogenous chemicals and exogenous components is matrixassisted laser desorption/time-of-flight mass spectrometry (MALDI/TOF MS) [2–4,14–17]. The combination of spectral and imaging information may help determine whether a detected chemical in a fingermark was from contact or excretion. The chemical imaging from MALDI/TOF MS allows for a more definitive answer than spectra alone because the analyte ion patterns should be uniformly visible in the print if deposition is through excretion. If the residue is simply left from the contact points after touching drugs or explosives, the chemical image will be unevenly dispersed on the print, reflecting the limited regions of contact [3]. Studies of limited sample sizes have demonstrated the ability to use ‘‘touch chemistry’’ to indicate if an individual smokes [4,11,17], has been exposed to or ingested illegal drugs of abuse [3,12], or has contacted explosives [2,18]. Many of these studies were simulations, and further efforts are needed to demonstrate applications in authentic samples and larger sample populations. In order to integrate touch chemistry with conventional fingerprint development methods, the requirements of the analytical technique must be considered. For example, MALDI requires an appropriate matrix for efficient ionization. Incorporating conventional fingerprint powders as the MALDI matrix has been shown to be effective in previous studies [4]. Alternative approaches from the literature include alpha-cyano-4-hydroxycinnamic acid (CHCA), a MALDI-specific matrix, as the fingerprint powder developer [15,16] or functionalized nanoparticle fingerprint powders specific for MALDI/TOF MS [2–4]. Investigating additional fingerprint developers that are compatible with MALDI is necessary for incorporating chemical analysis with current latent print identification processes. MALDI/TOF MS was used for this study to detect compounds in fingermark residue. A select group of drugs and explosives with different chemical properties was handled and analyzed. There were two main purposes of this research: (1) develop effective laboratory methods for detecting drugs and explosives in fingermark residues and (2) determine the feasibility of detecting drugs and explosives after casual contact with tablets, powders, and residues. As part of the laboratory study, we investigated synthetic latent print reference pads as mimics of natural fingermark residue to determine if the pads could be used for method development and quality control. 2. Materials and methods 2.1. Chemicals and materials Advil1 (200 mg ibuprofen tablets, Wyeth Consumer Healthcare, Madison, NJ), Tylenol1 Extra Strength (500 mg acetaminophen tablets, McNeil Healthcare LLC, Las Piedras, Puerto Rico), Bayer1 (81 mg aspirin tablets, Bayer Healthcare LLC, Morristown, NJ), and non-drowsy Sudafed1 24 h (240 mg pseudoephedrine hydrochloride tablets, McNeil Consumer Healthcare, Fort Washington, PA) were purchased from CVS1 Pharmacy. The weight percent based on label information (weight of drug/total weight of tablet 100) for each pill was 76% acetaminophen in Tylenol1, 64% pseudoephedrine hydrochloride in Sudafed1, 79% aspirin in Bayer1, and 42% ibuprofen in Advil1. Pseudoephedrine (1 mg/mL in methanol standard) was purchased from Cerilliant (Round Rock, TX). Angiotensin II human (>93% powder), a-cyano4-hydroxycinnamic acid (CHCA) (99% HPLC), procaine hydrochloride, and reserpine were purchased from Sigma–Aldrich (St. Louis, MO). TNT and RDX (1 mg/mL in 1:1 methanol:acetonitrile) were purchased from AccuStandard (New Haven, CT). TNT powder was obtained from our reference collection. Water (Optima grade), acetonitrile (Optima grade), acetic acid (certified ACS Plus), and
69
precleaned glass microscope slides were purchased from Fisher Scientific (Fair Lawn, NJ). Glass microscope slides (1 in. 3 in.) coated with approximately 100 nm aluminum were purchased from Deposition Research Lab, Inc (St. Charles, MO). Latent print reference pads for amino acid and sebaceous oil were purchased from Forensics Source (Jacksonville, FL). XYZ-Axis electrically conductive 3MTM double-sided tape (25 mm (W) 32.9 m (L)) was purchased from Electron Microscopy Sciences (Hatfield, PA). Ethyl cyanoacrylate adhesive INSTAbond1 S-100 was purchased from ACCRAbond, Inc (Olive Branch, MS). Conventional black latent fingerprint powder and white marabou feather brush were purchased from Arrowhead Scientific Inc. (Lenexa, KS). 2.2. Handling powders with artificial amino acid and sebaceous pad residue 2.2.1. Artificial fingermark residue All of the handling and residue experiments involving human subjects were approved and conducted in accordance to our Institutional Review Board. Artificial pads were used to create samples for initial development of MALDI/TOF MS methods. The pads were used to simulate fingermark residue in a controlled manner. Hands were washed with soap and water followed by an alcohol wipe prior to touching the artificial residue pads. For negative controls, artificial fingermark residues were directly deposited on aluminum-coated slides after rubbing the fingertips on the individual pads. With artificial residue still present on the fingertips, drug powders (aspirin, ibuprofen, and acetaminophen ground tablets) were handled before a second print was deposited next to the control print on the aluminum slide. Fingermarks were aged for approximately 30 min and developed using four separate processes: (1) dusted directly with black fingerprint development powder, (2) cyanoacrylate fumed followed by dusting with black fingerprint development powder, (3) dusted with black fingerprint development powder and lifted using tape, and (4) sprayed with MALDI matrix to apply a thin layer of CHCA. An improvised cyanoacrylate chamber was used for these initial experiments. 2.3. Handling tablets, powders, and transfer residues with natural fingermark residue 2.3.1. Analyte powder residue Hands were first washed with soap and water followed by an alcohol wipe to clean off external contaminants. ‘‘Groomed’’ control fingermarks were created by rubbing the fingers across the nose or neck to coat them with sebaceous gland secretions. Likewise, groomed fingertips were prepared for handling whole and broken aspirin, ibuprofen, and acetaminophen tablets. The whole tablets were placed between two fingers and held briefly to mimic the process of ingesting tablets. Tablets were then broken in half and the exposed drug/excipient side was dabbed once on the finger to imitate handling broken pills. Groomed fingertips were also used to touch ibuprofen, aspirin, acetaminophen, and pseudoephedrine hydrochloride powders from ground tablets. Likewise, ground analytical standards of procaine hydrochloride and TNT were touched with groomed fingertips to generate samples. Deposited fingermark residues were aged for approximately 30 min and developed with black fingerprint development powder, cyanoacrylate fuming followed by black fingerprint development powder, black fingerprint development powder and hinge lifting tape, and matrix sprayer with CHCA. 2.3.2. Analyte transfer residue An alternative to handling analyte powder was to establish a protocol for transferring residue to generate fingermark samples. First, calibration spots were generated with images to estimate the
70
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
quantity to use for the four compounds: pseudoephedrine, procaine, TNT, and RDX. Procaine and pseudoephedrine were used to mimic cocaine and methamphetamine, respectively. TNT and RDX were used as examples of high explosives. The amounts for initial trials were 1.3, 13, 130, 1300, and 2500 ng delivered from 1:1 water:methanol solutions. Samples were mixed with CHCA in a 1:1 ratio. Transfer experiments used 1300 ng pseudoephedrine and procaine hydrochloride and 2500 ng TNT and RDX. Three different volumes (1, 5, and 10 mL) of the four compounds were deposited on a glass slide and allowed to evaporate to dryness. Fingers were not washed prior to transfer but briefly rubbed together. The individual used fingertips to deposit a control print, then touched the dried residue and transferred the residue from the glass slide to a clean aluminum-coated slide. Fingermarks were aged for approximately 30 min, processed using black fingerprint development powder, cyanoacrylate fuming followed by black fingerprint development powder, black fingerprint development powder followed by lifting with hinge lifter, and matrix sprayer with CHCA.
technique [19] and allowed the samples to be archived. Positive and negative modes with a mass scan range between m/z 0 and 1200 were used for the detection of drugs and explosives. The raster area for all images was 70 mm 20 mm with 150 mm spacing and a total of 62,578 points. Two shots were accumulated per point. The laser power was optimized for the different chemical processes used to develop the fingermark residue. The rank order from high to low laser power was: lifts > cyanoacrylate fuming > black fingerprint development powder > CHCA. For all positive mode experiments, an ion gate was set to m/z 65 to prevent an overwhelming detection of sodium and potassium ions. The gate was not necessary for negative ionization mode. The pulsed extraction optimized m/z was set to 1047. The laser repetition rate was 50 Hz. The system was calibrated using the calibration spots of CHCA, reserpine, and angiotensin II prior to each analysis and checked after the run. The time taken to image one slide with four fingermark residues was about 4 h.
2.4. Fingermark development process
Images were processed using both Shimadzu Biotech MALDIMS software and BioMap (Novartis, Basel, Switzerland). In BioMap, regions of interest were created to extract the mass spectra and search for drugs and explosives present in the transfer residue. For prints to be considered positive, the extracted ion image had to show the correct location, and the mass spectra had to have a S/N ratio above 3 for the relevant ion.
Dusting with fingerprint powder. Fingermarks were developed with conventional fingerprint powder. Black fingerprint development powder and a white marabou feather brush were used to develop fingermark residues. Dusting created the challenge of confining the loose powder to the sample plate. Excess powder was removed by tapping the developed slides on a hard surface. However, more effective ways to remove excess powder are necessary as the MALDI vacuum chamber was eventually contaminated with black fingerprint development powder. Dusting followed by lifting. Conventional fingerprint powder was used to develop the prints then lifted using either conventional hinge lifters or double-sided conductive tape. The lifts were then mounted on an aluminum-coated slide. Cyanoacrylate fuming process. A Misonix1 CA-3000 cyanoacrylate fuming chamber was used following a standard operating procedure. Briefly, the chamber was set to reach 70% humidity prior to fuming. The hot plate was set to 300 8C and allowed approximately 5 min to equilibrate. Once the hot plate and humidity reached set values, 2 g cyanoacrylate in an aluminum boat were placed on the hot plate. The fumes were observed along with the development of the fingermark residue during the fuming process. When the fuming was completed, there was a chamber air purge cycle for approximately 10 min. The fumed prints were then dusted with black fingerprint development powder. Matrix sprayer parameters. The TM-SprayerTM (HTX Technologies, LLC, Carrboro, NC) was used to apply 5 mg/mL CHCA dissolved in 30:70 water:acetonitrile (ACN) with 0.1% acetic acid. The matrix sprayer has a 5 mL sample loop and double the loop volume of CHCA was used to load the loop to ensure no bubbles would get into the sprayer system. A syringe pump with 30:70 water:ACN with 0.1% acetic acid was set to a flow rate of 0.2 mL/min to push the matrix through the loop. One spray cycle was used with a velocity set to 800 mm/min at a temperature of 115 8C. The track spacing was also set to 3 mm with an analysis area set to 95 mm 40 mm. The sample was allowed to dry before analysis. 2.5. MALDI parameters The MALDI/TOF MS slide carrier held four slides at a time. The slides were mounted on the sample holder using double-sided carbon tape. The tape was placed on the top of the aluminumcoated slide with contact to the sample holder. A Shimadzu Axima Performance MALDI/TOF/TOF MS with a N2 laser (337 nm) was used in linear mode (TOF MS). MALDI/TOF MS is a nondestructive
2.6. Image processing
3. Results and discussion 3.1. Artificial fingermark residue and drug powder analysis The artificial amino acid and sebaceous oil pads were studied to determine if the pads mimic natural fingermark residue and could be used for method development. Natural fingermark residue consists primarily of material secreted by eccrine glands (located on the palmar side of hands or volar skin) and sebaceous glands (located on the scalp and face) [3]. Fingermark residues, described in the literature, consist primarily of sodium and potassium salts, amino acids, peptides, and lipids such as fatty acids, wax esters, squalene, and cholesterol [10,14–16]. While the exact composition of the latent reference pads is proprietary, the manufacturer disclosed that the base components of the sebaceous reference pad are linseed oil and a mixture of alcohols, and the base components of the amino acid reference pads are L-amino acids and water. Linseed oil consists of triglycerides and fatty acids which include palmitic, stearic, oleic, linoleic, and linolenic acids [20]. Palmitic, stearic, oleic, linoleic, and palmitoleic acids are reported to be major to moderate components found in sebaceous fingermark residue [7]. The mass spectra for the amino acid residue pads were distinguishable from the sebaceous oil residue pads (Fig. 1A and B). In the sebaceous oil residue, there are more intense peaks in the m/z range between m/z 450 and 570 (indicative of fatty acid masses). Additional differences between the sebaceous oil residue pad compared with the amino acid residue pad were found between the regions of m/z 150 and 200 (typical of masses for individual amino acids) and m/z 900 and 1200 (the high mass region was not shown in this window because the natural residues did not provide substantial signals above m/z 670 which are indicative of peptides and lipids). Distinct differences are shown among the mass spectra of the reference pads to natural secretions of three volunteers (Fig. 1A–E). Variations are also shown in the mass spectra for natural fingermark residue of the three volunteers (Fig. 1C–E), indicating the difficulty in creating a chemical mixture representative of natural fingermark residue. No attempt was made to identify the compounds in the natural fingermark residue.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
71
Fig. 1. Mass spectra of control fingermarks for artificial amino acid pad (a), sebaceous pad (b), female (c), male (d), and female (e) in positive ionization mode after being dusted with black fingerprint development powder.
Although the pads were not representative of natural fingermark residue, the pads were useful in setting initial instrumental and fingermark development parameters for experiments on handling targeted compounds. The amino acid and sebaceous synthetic residue were able to provide positive results after fingertips contacted analyte powders (ibuprofen, aspirin, and acetaminophen) with the most abundant ions being [M+Na]+ and [M+K]+. Protonated molecular ions were not detectable above the background in the images or spectra. In the ion image for ibuprofen powder, the signal intensity was higher in the amino acid residue than the sebaceous oil residue developed with black fingerprint powder (Fig. 2). The enhanced signal may be influenced by the difference in solubility of the drug in the amino acid over the sebaceous pad. The analyte powders were distributed throughout the fingermark extracted ion images made with the amino acid artificial pad fluid (Fig. 2). A cholesterol fragment ion from the loss of water [M+H-H2O]+ at m/z 369 [7] is a useful indicator to show the overall image (Fig. 2) for both the control and drug fingermarks. The same cholesterol fragment ion is shown in Fig. 1A in the amino acid pad example, and may or may not have originated from the reference pad. Although the individual’s fingermark was cleaned with an alcohol wipe, there are still endogenous compounds present in the fingermark residue and the cholesterol may have been deposited by the individual. Under the study conditions, the
pads were too dissimilar to the natural residues for use as a reference control. 3.2. Analysis of natural fingermark residue after handling powders of target compounds Initial experiments were designed to determine if casual contact with target drugs could be measured after handling common over-the-counter pharmaceuticals. Broken (to expose the drug interior) and intact tablets of aspirin, acetaminophen, and ibuprofen were handled before natural fingermark residues were deposited. Target drug ions were not detected, indicating the coating on the tablets may limit the transfer of sufficient amounts of the drugs onto fingermark residue. Broken tablets were used for additional handling experiments, but the small exposed surface area which had a high percentage of drug (range 42–79%) did not provide sufficient transfer for a definitive positive signal. Binding agents in the tablets may have minimized transfer of the drugs. As a consequence of the pills not providing adequate signal for detection, handling experiments using natural fingermark residue were conducted with powders from several compounds (aspirin, ibuprofen, acetaminophen, procaine, pseudoephedrine, and TNT). Four different development techniques were used: (1) dusting with fingerprint powder; (2) cyanoacrylate fuming followed by
72
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
Fig. 2. Comparison between control and ibuprofen powder prints deposited after coating fingers with either artificial amino acid or sebaceous fluid and developed with black fingerprint powder. Top picture is an extracted ion image for a fragment of cholesterol [M-H2O+H]+(m/z 369). The primary ions formed for ibuprofen detection were m/z 229 [M+Na]+ and m/z 245 [M+K]+ with the amino acid pad giving a higher response.
Fig. 3. Natural fingermark residue with control and aspirin powder as extracted ion image for [M H] for direct black fingerprint development powder analysis (A), cyanoacrylate fuming and black fingerprint development powder (B), black fingerprint development powder and commercial lifting tape (C), and matrix sprayer with CHCA (D).
Fig. 4. Extracted ion image for [M+H]+ of procaine powder. There are three fingermark residues present on the slide: (1) the control, (2) pseudoephedrine, and (3) procaine residue but only the extracted ion m/z 238 for procaine is demonstrated. The procaine powder is spread throughout the print and on the right side off of the print due to the development process.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
dusting with fingerprint powder; (3) dusting with fingerprint powder followed by lifting with conventional hinge lifters; and (4) spraying CHCA MALDI matrix directly on the residue. In contrast to the artificial amino acid residue pad images (Fig. 2B), the fingermark ridge detail was not observed in the drug signal from natural excretions (Fig. 3A). In order to view the ridge details of the fingermark, alternative ions were used (e.g., m/z 369 cholesterol fragment) in natural fingermarks. Images displayed distinct differences between the fingermark control residues and the target drug or explosives residues for dusting, cyanoacrylate fuming followed by dusting, and CHCA matrix. However, the results of the lifting experiments were compound-dependent because only aspirin and ibuprofen were detected. Lifts may have been unsuccessful due to the lack of conductivity of the hinge lifters which results in higher laser power requirements and broad
73
spectral peaks with low S/N. Peak broadening reduced the ability to find specific ions for extracting images. A concomitant factor is the unevenness of the surface of the hinge lifter relative to an aluminum-coated slide. The principle ions formed with fingerprint powder and cyanoacrylate fuming followed by fingerprint powder were as follows: aspirin and ibuprofen [M H] ; TNT [M] ; procaine and pseudoephedrine [M+H]+; and acetaminophen [M+Na]+. An example of natural fingermark residue extracted ion images for controls and aspirin powders using the four processes are found in Fig. 3A–D. The cyanoacrylate development helped confine the sample material to the fingermark residue area which prevented the powder from spreading during the dusting step. However, the chemical background in the spectra from the cyanoacrylate polymer was noticeably higher than the other processing methods.
Fig. 5. Mass spectra and image of [M+H]+ for pseudoephedrine mixed 1:1 with CHCA at 5 different amounts from top to bottom: 2500, 1300, 130, 13, 1.2 ng. Spot sizes are 3 mm.
74
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
The predominate analyte ions observed using CHCA matrix were protonated (positive mode) and deprotonated (negative mode) molecular compounds which are more amenable to the use of MS/MS when compared to other adduct ions (e.g., sodium and potassium). The relative success rate of fingermark development techniques that were able to provide positive results for target compounds based on signal intensity and image were as follows: MALDI matrix > black fingerprint development powder > superglue and black fingerprint development powder > black fingerprint development powder and lifts.
3.3. Analysis of natural fingermark residue after handling residues of target compounds An issue encountered with handling loose analyte powders was the spread of the analytes beyond the drug fingermark on the aluminum-coated slide during dusting with black fingerprint development powder. Vertical barriers were used to isolate neighboring fingermark residues from each other. The appearance of procaine signal beyond the fingermark is seen in Fig. 4A to the right side of the drug print where no barrier was used.
Fig. 6. Mass spectra and image of [M+H]+ for procaine mixed 1:1 with CHCA at 5 different amounts from top to bottom: 2500, 1300, 130, 13, 1.3 ng. Spot sizes 3 mm.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
In order to avoid the issues with loose powders, transferring dried residue was investigated. To establish the appropriate concentration to use for transferring residues, target analytes were spotted 1:1 with CHCA matrix and imaged. An example of mass spectra and an extracted ion image for pseudoephedrine is shown in Fig. 5 for five different amounts ranging from 1.3 ng to 2500 ng. For all five amounts there was a visible response in the image and discernible peaks in the mass spectra for m/z 166 [M+H]+. A duplicate experiment using procaine hydrochloride is demonstrated in Fig. 6. The ion at m/z 237 is the [M+H]+ for procaine. The signal intensity was more intense for procaine when compared with pseudoephedrine. Similar experiments were performed using fingerprint powder as the MALDI matrix (data not shown) and comparable responses were observed. The mass of 1300 ng was chosen for the direct deposit on a slide for fingermark
75
transfers. The drug residue was anticipated to have adequate signal after transferring with finger contact but expected to be at much lower levels. An example of a transfer experiment in fingermark residue for protonated procaine is demonstrated in Fig. 7. Three different volumes (1 mL, 5 mL, and 10 mL) are shown on a single print, and it was determined that a 5 mL spot size was sufficient for detection. When comparing the mass spectra of the analyte powders to the transfer residues, the [M+H]+ S/N is greater in the transfer residue. RDX and TNT were also evaluated using five different amounts ranging from 1.3 ng to 2500 ng. The primary ions measured in the standards were [M H] for TNT and a fragment of RDX [M– CH2N(NO2)2] . For all five amounts there was visible response in the image and the analyte peaks in the mass spectra were above the noise level. The RDX fragment was detected with smaller S/N
Fig. 7. Extracted ion images for [M+H]+ of transferred procaine residue from a glass slide to an aluminum-coated slide. There are three fingermark residues present on the slide: (1) the control, (2) pseudoephedrine, and (3) procaine residue but only the extracted ion m/z 238 for procaine is demonstrated. The transfer residue has three different volumes investigated (1 mL, 5 mL, and 10 mL) at the same amount (1300 ng) with the corresponding mass spectra shown.
76
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
Fig. 8. Extracted ion images (XIC) for transferred pseudoephedrine, procaine, TNT, and RDX residues. The latent fingermark experimental layout (A) is shown above the XICs. The control fingermark consists of the solvent background (1); the drug fingermark consists of pseudoephedrine (2) and procaine (3); and the explosive fingermark consists of TNT (4) and RDX (5). The fingermarks were developed with black powder and the ions observed were [M+H]+ for pseudoephedrine (B), procaine (C), [M H] for TNT (D), and [M–CH2N(NO2)2] fragment for RDX (E).
than the TNT ion for the different amounts. The mass 2500 ng was selected for direct deposit on the slide for fingermark transfers because the RDX fragment was harder to detect after transferring the explosives residue. The analytes from the transfer residues (procaine, pseudoephedrine, TNT, and RDX) were evaluated after developing the fingermarks using the four different processes. All target drugs and explosives were detected in the appropriate fingermark residues after developing using the four different processes. An example of extracted ion images for all four analytes is shown in Fig. 8 for fingermarks developed with black fingerprint powder. The primary ions detected using all development processes were [M+H]+ for procaine and pseudoephedrine, [M] and [M H] for TNT, and [M-CH2N(NO2)2] fragment for RDX. Transfer residues provided a reliable means of preparing fingermarks with known amounts of analytes as compared to using analyte powders. 4. Conclusion The MALDI/TOF MS was able to detect exogenous substances from contact residue under laboratory conditions with the integration of conventional fingerprint development methods. MALDI/TOF MS is a nondestructive method which provides
chemical information in both the mass spectra and chemical images. Touch chemistry using MALDI/TOF MS shows promising results for investigative and intelligence gathering applications for detecting exogenous compounds. In real-life scenarios of casually handling whole tablets, MALDI/TOF MS was unable to detect drug residue using the four different development processes. Casual interactions with whole or broken tablets are unlikely to result in detectable amounts under the experimental conditions described here. Reference pads were useful in creating MALDI/TOF MS methods for imaging fingermark residue; however, they did not model natural secretions chemically or physically enough to be used as a quality control measure for this study. Positive results were demonstrated for handling analyte powders. However, analyte powders were easily spread throughout the fingermark when developing the fingermarks using black fingerprint development powder and an application brush. In the future, if crime scene technicians collect evidence for chemical analysis of latent fingermark residue, it should be noted that loose powder on the fingermark residue will spread when the fingermark is developed using the powder and brush technique. In addition, it was difficult to reproducibly control the amount of analyte powder in the fingermark residue; field samples could demonstrate a large variability in the amounts of powders collected. Contact residues, which can be helpful for method development but may not resemble real circumstances, were useful in controlling the deposition location and amount of target compounds. Although touch chemistry is time consuming, the information gained may tie an individual to specific evidence. As faster lasers and newer technology evolve, the collection duration can be reduced. Without the chemical image it would be a challenge to draw conclusions due to the large chemical noise from the development process, the natural components in residue, and the background noise from the originating surface. Chemical imaging analysis can be used without impeding the latent print process. Smeared prints or prints that are not of value for identification can still be used for mass spectral analysis. Acknowledgements This is publication 13-14 of the Federal Bureau of Investigation Laboratory Division. This research was supported in part by an appointment to the Visiting Scientist Program at the Federal Bureau of Investigation Laboratory Division, administered by the Oak Ridge Institute for Science and Education, through an interagency agreement between the U.S. Department of Energy and the FBI. Names of commercial manufacturers are provided for identification purposes only and inclusion does not imply endorsement of the manufacturer or its products or services by the FBI. The views expressed are those of the authors and do not necessarily reflect the official policy or position of the FBI or the U.S. Government. References [1] T.A. Trozzi, R.L. Schwartz, M.L. Hollars, Processing guide for developing latent prints, U.S. Department of Justice, FBI, Laboratory Division, Forensic Sci. Commun. 3 (2001) 1 http://www.fbi.gov/about-us/lab/forensic-science-communications/ fsc/jan2001. [2] F. Rowell, J. Seviour, A.Y. Lim, C.G. Elumbaring-Salzar, J. Loke, J. Ma, Detection of nitro-organic and peroxide explosives in latent fingermarks by DART- and SALDITOF-mass spectrometry, Forensic Sci. Int. 221 (2012) 84–91. [3] F. Rowell, K. Hudson, J. Seviour, Detection of drugs and their metabolites in dusted latent fingermarks by mass spectrometry, Analyst 4 (2009) 701–707. [4] M. Benton, F. Rowell, L. Sundar, M. Jan, Direct detection of nicotine and cotinine in dusted latent fingermarks of smokers by using hydrophobic silica particles and MS, Surf. Interface Anal. 42 (2009) 378–385. [5] R.S. Croxton, M.G. Baron, D. Butler, T. Kent, V.G. Sears, Variation in amino acid and lipid composition of latent fingerprints, Forensic Sci. Int. 199 (2010) 93–102.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77 [6] K.M. Antonie, S. Mortazavi, A.D. Miller, L.M. Miller, Chemical differences are observed in children’s versus adults’ latent fingerprints as a function of time, J. Forensic Sci. 55 (2010) 513–551. [7] G.M. Mong, C.E. Petersen, T.R.W. Clauss, Advanced fingerprint analysis project fingerprint constituents, Pacific Northwest National Laboratory, 1999, PNNL13019, pp. 1–14. Retrieved from http://www.osti.gov/bridge/servlets/purl/ 14172-SQLzxz/webviewable/14172.pdf on June 12, 2013. [8] L.S. Ferguson, F. Wulfert, R. Wolstenholme, J.M. Fonville, M.R. Clench, V.A. Carolan, S. Francese, Direct detection of peptides and small proteins in fingermarks and determination of sex by MALDI mass spectrometry profiling, Analyst 137 (2012) 4686–4692. [9] T. Chen, Z.D. Schultz, I.W. Levin, Infrared spectroscopic imagining of latent fingerprints and associated forensic evidence, Analyst 134 (2009) 1902–1904. [10] B. Hartzell-Baguley, R.E. Hipp, N.R. Morgan, S.L. Morgan, Chemical composition of latent fingerprints by gas chromatography–mass spectrometry, J. Chem. Ed. 84 (2007) 689–691. [11] P. Hazarika, S.M. Jickells, D.A. Russell, Rapid detection of drug metabolites in latent fingermarks, Analyst 134 (2009) 93–96. [12] D.R. Ifa, N.E. Manicke, A.L. Dill, R.G. Cooks, Latent fingerprint chemical imaging by mass spectrometry, Science 321 (2008) 80. [13] M.I. Szynkowska, K. Czerski, J. Rogowski, T. Paryjczak, A. Parcewski, Detection of exogenous contaminants of fingerprints using ToF-SIMS, Surf. Interface Anal. 5 (2010) 393–397.
77
[14] M.J. Bailey, N.J. Bright, R.S. Croxton, S. Francese, L.S. Ferguson, S. Hinder, S. Jickells, B.J. Jones, B.N. Jones, S.G. Kazarian, J.J. Ojeda, R.P. Webb, R. Wolstenholme, S. Bleay, Chemical characterization of latent fingerprints by matrix-assisted laser desorption ionization, time-of-flight secondary ion mass spectrometry, mega electron volt secondary mass spectrometry, gas chromatography/mass spectrometry, Xray photoelcectron spectroscopy, and attenuated total reflection Fourier transform infrared spectroscopic imaging: an intercomparison, Anal. Chem. 84 (2012) 8514–8523. [15] R. Wolstenholme, R. Bradshaw, M.R. Clench, S. Francese, Study of latent fingermarks by matrix-assisted laser desorption/ionization mass spectrometry imaging of endogenous lipids, Rapid Commun. Mass Spectrom. 23 (2009) 3031–3039. [16] L. Ferguson, R. Bradshaw, R. Wolstenholme, M. Clench, S. Francese, Two-step matrix application for the enhancement and imaging of latent fingermarks, Anal. Chem. 83 (2011) 5585–5591. [17] M. Benton, M.J. Chua, F. Gu, F. Rowell, J. Ma, Environmental nicotine contamination in latent fingermarks from smoker contacts and passive smoking, Forensic Sci. Int. 200 (2010) 28–34. [18] P.H. Ng, S. Walker, M. Tahtouh, B. Reedy, Detection of illicit substances in fingerprints by infrared spectral imaging, Anal. Bioanal. Chem. 394 (2009) 2039–2048. [19] K. Dreisewerd, The desorption process in MALDI, Chem. Rev. 103 (2003) 395–425. [20] A.G. Vereshchagin, G.V. Novitskaya, The triglyceride composition of linseed oil, J. Am. Oil Chem. Soc. 42 (1965) 970–974.
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Chemical analysis of pharmaceuticals and explosives in fingermarks using matrix-assisted laser desorption ionization/time-of-flight mass spectrometry Kimberly Kaplan-Sandquist a, Marc A. LeBeau b, Mark L. Miller c,* a Counterterrorism and Forensic Science Research Unit, Visiting Scientist Program, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States b Scientific Analysis Section, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States c Counterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, Quantico, VA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 July 2013 Received in revised form 19 November 2013 Accepted 27 November 2013 Available online 16 December 2013
Chemical analysis of latent fingermarks, ‘‘touch chemistry,’’ has the potential of providing intelligence or forensically relevant information. Matrix-assisted laser desorption ionization/time-of-flight mass spectrometry (MALDI/TOF MS) was used as an analytical platform for obtaining mass spectra and chemical images of target drugs and explosives in fingermark residues following conventional fingerprint development methods and MALDI matrix processing. There were two main purposes of this research: (1) develop effective laboratory methods for detecting drugs and explosives in fingermark residues and (2) determine the feasibility of detecting drugs and explosives after casual contact with pills, powders, and residues. Further, synthetic latent print reference pads were evaluated as mimics of natural fingermark residue to determine if the pads could be used for method development and quality control. The results suggest that artificial amino acid and sebaceous oil residue pads are not suitable to adequately simulate natural fingermark chemistry for MALDI/TOF MS analysis. However, the pads were useful for designing experiments and setting instrumental parameters. Based on the natural fingermark residue experiments, handling whole or broken pills did not transfer sufficient quantities of drugs to allow for definitive detection. Transferring drugs or explosives in the form of powders and residues was successful for preparing analytes for detection after contact with fingers and deposition of fingermark residue. One downfall to handling powders was that the analyte particles were easily spread beyond the original fingermark during development. Analyte particles were confined in the original fingermark when using transfer residues. The MALDI/TOF MS was able to detect procaine, pseudoephedrine, TNT, and RDX from contact residue under laboratory conditions with the integration of conventional fingerprint development methods and MALDI matrix. MALDI/TOF MS is a nondestructive technique which provides chemical information in both the mass spectra and chemical images. Published by Elsevier Ireland Ltd.
Keywords: Touch chemistry Latent fingerprint Fingerprint powder Chemical imaging Drugs Cyanoacrylate fuming
1. Introduction ‘‘Touch chemistry’’ is a forensic technique that could provide investigative leads by analyzing chemical data from fingermark residue. Chemical intelligence gained from the analysis of latent fingermarks has the potential to provide information about the suspect even if the fingermarks are smudged or the patterns cannot
* Corresponding author at: Counterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, Quantico, 2501 Investigation Parkway, VA 22135, United States. Tel.: +1 703 632 7846; fax: +1 703 632 7801. E-mail address: [email protected] (M.L. Miller). 0379-0738/$ – see front matter . Published by Elsevier Ireland Ltd. http://dx.doi.org/10.1016/j.forsciint.2013.11.016
be matched through an automated fingerprint search. In order to collect latent fingermarks at a crime scene, the prints must first be made visible using development techniques such as fingerprint powder or cyanoacrylate fuming [1]. Once fingermarks are located, chemical information to be gained about the individual could include exposure to illicit substances/explosives, drug-use history, gender, and approximate age [2–8]. Past approaches for fingermark residue characterization include Fourier transform-infrared spectroscopy (FTIR) [9], gas chromatography mass spectrometry (GC/MS) [10], fluorescence microscopy [11], desorption electrospray ionization mass spectrometry (DESI/MS) [12], and secondary ion mass spectrometry (SIMS) [13,14]. An alternative approach that has recently
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
demonstrated high potential for characterizing a broad array of endogenous chemicals and exogenous components is matrixassisted laser desorption/time-of-flight mass spectrometry (MALDI/TOF MS) [2–4,14–17]. The combination of spectral and imaging information may help determine whether a detected chemical in a fingermark was from contact or excretion. The chemical imaging from MALDI/TOF MS allows for a more definitive answer than spectra alone because the analyte ion patterns should be uniformly visible in the print if deposition is through excretion. If the residue is simply left from the contact points after touching drugs or explosives, the chemical image will be unevenly dispersed on the print, reflecting the limited regions of contact [3]. Studies of limited sample sizes have demonstrated the ability to use ‘‘touch chemistry’’ to indicate if an individual smokes [4,11,17], has been exposed to or ingested illegal drugs of abuse [3,12], or has contacted explosives [2,18]. Many of these studies were simulations, and further efforts are needed to demonstrate applications in authentic samples and larger sample populations. In order to integrate touch chemistry with conventional fingerprint development methods, the requirements of the analytical technique must be considered. For example, MALDI requires an appropriate matrix for efficient ionization. Incorporating conventional fingerprint powders as the MALDI matrix has been shown to be effective in previous studies [4]. Alternative approaches from the literature include alpha-cyano-4-hydroxycinnamic acid (CHCA), a MALDI-specific matrix, as the fingerprint powder developer [15,16] or functionalized nanoparticle fingerprint powders specific for MALDI/TOF MS [2–4]. Investigating additional fingerprint developers that are compatible with MALDI is necessary for incorporating chemical analysis with current latent print identification processes. MALDI/TOF MS was used for this study to detect compounds in fingermark residue. A select group of drugs and explosives with different chemical properties was handled and analyzed. There were two main purposes of this research: (1) develop effective laboratory methods for detecting drugs and explosives in fingermark residues and (2) determine the feasibility of detecting drugs and explosives after casual contact with tablets, powders, and residues. As part of the laboratory study, we investigated synthetic latent print reference pads as mimics of natural fingermark residue to determine if the pads could be used for method development and quality control. 2. Materials and methods 2.1. Chemicals and materials Advil1 (200 mg ibuprofen tablets, Wyeth Consumer Healthcare, Madison, NJ), Tylenol1 Extra Strength (500 mg acetaminophen tablets, McNeil Healthcare LLC, Las Piedras, Puerto Rico), Bayer1 (81 mg aspirin tablets, Bayer Healthcare LLC, Morristown, NJ), and non-drowsy Sudafed1 24 h (240 mg pseudoephedrine hydrochloride tablets, McNeil Consumer Healthcare, Fort Washington, PA) were purchased from CVS1 Pharmacy. The weight percent based on label information (weight of drug/total weight of tablet 100) for each pill was 76% acetaminophen in Tylenol1, 64% pseudoephedrine hydrochloride in Sudafed1, 79% aspirin in Bayer1, and 42% ibuprofen in Advil1. Pseudoephedrine (1 mg/mL in methanol standard) was purchased from Cerilliant (Round Rock, TX). Angiotensin II human (>93% powder), a-cyano4-hydroxycinnamic acid (CHCA) (99% HPLC), procaine hydrochloride, and reserpine were purchased from Sigma–Aldrich (St. Louis, MO). TNT and RDX (1 mg/mL in 1:1 methanol:acetonitrile) were purchased from AccuStandard (New Haven, CT). TNT powder was obtained from our reference collection. Water (Optima grade), acetonitrile (Optima grade), acetic acid (certified ACS Plus), and
69
precleaned glass microscope slides were purchased from Fisher Scientific (Fair Lawn, NJ). Glass microscope slides (1 in. 3 in.) coated with approximately 100 nm aluminum were purchased from Deposition Research Lab, Inc (St. Charles, MO). Latent print reference pads for amino acid and sebaceous oil were purchased from Forensics Source (Jacksonville, FL). XYZ-Axis electrically conductive 3MTM double-sided tape (25 mm (W) 32.9 m (L)) was purchased from Electron Microscopy Sciences (Hatfield, PA). Ethyl cyanoacrylate adhesive INSTAbond1 S-100 was purchased from ACCRAbond, Inc (Olive Branch, MS). Conventional black latent fingerprint powder and white marabou feather brush were purchased from Arrowhead Scientific Inc. (Lenexa, KS). 2.2. Handling powders with artificial amino acid and sebaceous pad residue 2.2.1. Artificial fingermark residue All of the handling and residue experiments involving human subjects were approved and conducted in accordance to our Institutional Review Board. Artificial pads were used to create samples for initial development of MALDI/TOF MS methods. The pads were used to simulate fingermark residue in a controlled manner. Hands were washed with soap and water followed by an alcohol wipe prior to touching the artificial residue pads. For negative controls, artificial fingermark residues were directly deposited on aluminum-coated slides after rubbing the fingertips on the individual pads. With artificial residue still present on the fingertips, drug powders (aspirin, ibuprofen, and acetaminophen ground tablets) were handled before a second print was deposited next to the control print on the aluminum slide. Fingermarks were aged for approximately 30 min and developed using four separate processes: (1) dusted directly with black fingerprint development powder, (2) cyanoacrylate fumed followed by dusting with black fingerprint development powder, (3) dusted with black fingerprint development powder and lifted using tape, and (4) sprayed with MALDI matrix to apply a thin layer of CHCA. An improvised cyanoacrylate chamber was used for these initial experiments. 2.3. Handling tablets, powders, and transfer residues with natural fingermark residue 2.3.1. Analyte powder residue Hands were first washed with soap and water followed by an alcohol wipe to clean off external contaminants. ‘‘Groomed’’ control fingermarks were created by rubbing the fingers across the nose or neck to coat them with sebaceous gland secretions. Likewise, groomed fingertips were prepared for handling whole and broken aspirin, ibuprofen, and acetaminophen tablets. The whole tablets were placed between two fingers and held briefly to mimic the process of ingesting tablets. Tablets were then broken in half and the exposed drug/excipient side was dabbed once on the finger to imitate handling broken pills. Groomed fingertips were also used to touch ibuprofen, aspirin, acetaminophen, and pseudoephedrine hydrochloride powders from ground tablets. Likewise, ground analytical standards of procaine hydrochloride and TNT were touched with groomed fingertips to generate samples. Deposited fingermark residues were aged for approximately 30 min and developed with black fingerprint development powder, cyanoacrylate fuming followed by black fingerprint development powder, black fingerprint development powder and hinge lifting tape, and matrix sprayer with CHCA. 2.3.2. Analyte transfer residue An alternative to handling analyte powder was to establish a protocol for transferring residue to generate fingermark samples. First, calibration spots were generated with images to estimate the
70
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
quantity to use for the four compounds: pseudoephedrine, procaine, TNT, and RDX. Procaine and pseudoephedrine were used to mimic cocaine and methamphetamine, respectively. TNT and RDX were used as examples of high explosives. The amounts for initial trials were 1.3, 13, 130, 1300, and 2500 ng delivered from 1:1 water:methanol solutions. Samples were mixed with CHCA in a 1:1 ratio. Transfer experiments used 1300 ng pseudoephedrine and procaine hydrochloride and 2500 ng TNT and RDX. Three different volumes (1, 5, and 10 mL) of the four compounds were deposited on a glass slide and allowed to evaporate to dryness. Fingers were not washed prior to transfer but briefly rubbed together. The individual used fingertips to deposit a control print, then touched the dried residue and transferred the residue from the glass slide to a clean aluminum-coated slide. Fingermarks were aged for approximately 30 min, processed using black fingerprint development powder, cyanoacrylate fuming followed by black fingerprint development powder, black fingerprint development powder followed by lifting with hinge lifter, and matrix sprayer with CHCA.
technique [19] and allowed the samples to be archived. Positive and negative modes with a mass scan range between m/z 0 and 1200 were used for the detection of drugs and explosives. The raster area for all images was 70 mm 20 mm with 150 mm spacing and a total of 62,578 points. Two shots were accumulated per point. The laser power was optimized for the different chemical processes used to develop the fingermark residue. The rank order from high to low laser power was: lifts > cyanoacrylate fuming > black fingerprint development powder > CHCA. For all positive mode experiments, an ion gate was set to m/z 65 to prevent an overwhelming detection of sodium and potassium ions. The gate was not necessary for negative ionization mode. The pulsed extraction optimized m/z was set to 1047. The laser repetition rate was 50 Hz. The system was calibrated using the calibration spots of CHCA, reserpine, and angiotensin II prior to each analysis and checked after the run. The time taken to image one slide with four fingermark residues was about 4 h.
2.4. Fingermark development process
Images were processed using both Shimadzu Biotech MALDIMS software and BioMap (Novartis, Basel, Switzerland). In BioMap, regions of interest were created to extract the mass spectra and search for drugs and explosives present in the transfer residue. For prints to be considered positive, the extracted ion image had to show the correct location, and the mass spectra had to have a S/N ratio above 3 for the relevant ion.
Dusting with fingerprint powder. Fingermarks were developed with conventional fingerprint powder. Black fingerprint development powder and a white marabou feather brush were used to develop fingermark residues. Dusting created the challenge of confining the loose powder to the sample plate. Excess powder was removed by tapping the developed slides on a hard surface. However, more effective ways to remove excess powder are necessary as the MALDI vacuum chamber was eventually contaminated with black fingerprint development powder. Dusting followed by lifting. Conventional fingerprint powder was used to develop the prints then lifted using either conventional hinge lifters or double-sided conductive tape. The lifts were then mounted on an aluminum-coated slide. Cyanoacrylate fuming process. A Misonix1 CA-3000 cyanoacrylate fuming chamber was used following a standard operating procedure. Briefly, the chamber was set to reach 70% humidity prior to fuming. The hot plate was set to 300 8C and allowed approximately 5 min to equilibrate. Once the hot plate and humidity reached set values, 2 g cyanoacrylate in an aluminum boat were placed on the hot plate. The fumes were observed along with the development of the fingermark residue during the fuming process. When the fuming was completed, there was a chamber air purge cycle for approximately 10 min. The fumed prints were then dusted with black fingerprint development powder. Matrix sprayer parameters. The TM-SprayerTM (HTX Technologies, LLC, Carrboro, NC) was used to apply 5 mg/mL CHCA dissolved in 30:70 water:acetonitrile (ACN) with 0.1% acetic acid. The matrix sprayer has a 5 mL sample loop and double the loop volume of CHCA was used to load the loop to ensure no bubbles would get into the sprayer system. A syringe pump with 30:70 water:ACN with 0.1% acetic acid was set to a flow rate of 0.2 mL/min to push the matrix through the loop. One spray cycle was used with a velocity set to 800 mm/min at a temperature of 115 8C. The track spacing was also set to 3 mm with an analysis area set to 95 mm 40 mm. The sample was allowed to dry before analysis. 2.5. MALDI parameters The MALDI/TOF MS slide carrier held four slides at a time. The slides were mounted on the sample holder using double-sided carbon tape. The tape was placed on the top of the aluminumcoated slide with contact to the sample holder. A Shimadzu Axima Performance MALDI/TOF/TOF MS with a N2 laser (337 nm) was used in linear mode (TOF MS). MALDI/TOF MS is a nondestructive
2.6. Image processing
3. Results and discussion 3.1. Artificial fingermark residue and drug powder analysis The artificial amino acid and sebaceous oil pads were studied to determine if the pads mimic natural fingermark residue and could be used for method development. Natural fingermark residue consists primarily of material secreted by eccrine glands (located on the palmar side of hands or volar skin) and sebaceous glands (located on the scalp and face) [3]. Fingermark residues, described in the literature, consist primarily of sodium and potassium salts, amino acids, peptides, and lipids such as fatty acids, wax esters, squalene, and cholesterol [10,14–16]. While the exact composition of the latent reference pads is proprietary, the manufacturer disclosed that the base components of the sebaceous reference pad are linseed oil and a mixture of alcohols, and the base components of the amino acid reference pads are L-amino acids and water. Linseed oil consists of triglycerides and fatty acids which include palmitic, stearic, oleic, linoleic, and linolenic acids [20]. Palmitic, stearic, oleic, linoleic, and palmitoleic acids are reported to be major to moderate components found in sebaceous fingermark residue [7]. The mass spectra for the amino acid residue pads were distinguishable from the sebaceous oil residue pads (Fig. 1A and B). In the sebaceous oil residue, there are more intense peaks in the m/z range between m/z 450 and 570 (indicative of fatty acid masses). Additional differences between the sebaceous oil residue pad compared with the amino acid residue pad were found between the regions of m/z 150 and 200 (typical of masses for individual amino acids) and m/z 900 and 1200 (the high mass region was not shown in this window because the natural residues did not provide substantial signals above m/z 670 which are indicative of peptides and lipids). Distinct differences are shown among the mass spectra of the reference pads to natural secretions of three volunteers (Fig. 1A–E). Variations are also shown in the mass spectra for natural fingermark residue of the three volunteers (Fig. 1C–E), indicating the difficulty in creating a chemical mixture representative of natural fingermark residue. No attempt was made to identify the compounds in the natural fingermark residue.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
71
Fig. 1. Mass spectra of control fingermarks for artificial amino acid pad (a), sebaceous pad (b), female (c), male (d), and female (e) in positive ionization mode after being dusted with black fingerprint development powder.
Although the pads were not representative of natural fingermark residue, the pads were useful in setting initial instrumental and fingermark development parameters for experiments on handling targeted compounds. The amino acid and sebaceous synthetic residue were able to provide positive results after fingertips contacted analyte powders (ibuprofen, aspirin, and acetaminophen) with the most abundant ions being [M+Na]+ and [M+K]+. Protonated molecular ions were not detectable above the background in the images or spectra. In the ion image for ibuprofen powder, the signal intensity was higher in the amino acid residue than the sebaceous oil residue developed with black fingerprint powder (Fig. 2). The enhanced signal may be influenced by the difference in solubility of the drug in the amino acid over the sebaceous pad. The analyte powders were distributed throughout the fingermark extracted ion images made with the amino acid artificial pad fluid (Fig. 2). A cholesterol fragment ion from the loss of water [M+H-H2O]+ at m/z 369 [7] is a useful indicator to show the overall image (Fig. 2) for both the control and drug fingermarks. The same cholesterol fragment ion is shown in Fig. 1A in the amino acid pad example, and may or may not have originated from the reference pad. Although the individual’s fingermark was cleaned with an alcohol wipe, there are still endogenous compounds present in the fingermark residue and the cholesterol may have been deposited by the individual. Under the study conditions, the
pads were too dissimilar to the natural residues for use as a reference control. 3.2. Analysis of natural fingermark residue after handling powders of target compounds Initial experiments were designed to determine if casual contact with target drugs could be measured after handling common over-the-counter pharmaceuticals. Broken (to expose the drug interior) and intact tablets of aspirin, acetaminophen, and ibuprofen were handled before natural fingermark residues were deposited. Target drug ions were not detected, indicating the coating on the tablets may limit the transfer of sufficient amounts of the drugs onto fingermark residue. Broken tablets were used for additional handling experiments, but the small exposed surface area which had a high percentage of drug (range 42–79%) did not provide sufficient transfer for a definitive positive signal. Binding agents in the tablets may have minimized transfer of the drugs. As a consequence of the pills not providing adequate signal for detection, handling experiments using natural fingermark residue were conducted with powders from several compounds (aspirin, ibuprofen, acetaminophen, procaine, pseudoephedrine, and TNT). Four different development techniques were used: (1) dusting with fingerprint powder; (2) cyanoacrylate fuming followed by
72
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
Fig. 2. Comparison between control and ibuprofen powder prints deposited after coating fingers with either artificial amino acid or sebaceous fluid and developed with black fingerprint powder. Top picture is an extracted ion image for a fragment of cholesterol [M-H2O+H]+(m/z 369). The primary ions formed for ibuprofen detection were m/z 229 [M+Na]+ and m/z 245 [M+K]+ with the amino acid pad giving a higher response.
Fig. 3. Natural fingermark residue with control and aspirin powder as extracted ion image for [M H] for direct black fingerprint development powder analysis (A), cyanoacrylate fuming and black fingerprint development powder (B), black fingerprint development powder and commercial lifting tape (C), and matrix sprayer with CHCA (D).
Fig. 4. Extracted ion image for [M+H]+ of procaine powder. There are three fingermark residues present on the slide: (1) the control, (2) pseudoephedrine, and (3) procaine residue but only the extracted ion m/z 238 for procaine is demonstrated. The procaine powder is spread throughout the print and on the right side off of the print due to the development process.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
dusting with fingerprint powder; (3) dusting with fingerprint powder followed by lifting with conventional hinge lifters; and (4) spraying CHCA MALDI matrix directly on the residue. In contrast to the artificial amino acid residue pad images (Fig. 2B), the fingermark ridge detail was not observed in the drug signal from natural excretions (Fig. 3A). In order to view the ridge details of the fingermark, alternative ions were used (e.g., m/z 369 cholesterol fragment) in natural fingermarks. Images displayed distinct differences between the fingermark control residues and the target drug or explosives residues for dusting, cyanoacrylate fuming followed by dusting, and CHCA matrix. However, the results of the lifting experiments were compound-dependent because only aspirin and ibuprofen were detected. Lifts may have been unsuccessful due to the lack of conductivity of the hinge lifters which results in higher laser power requirements and broad
73
spectral peaks with low S/N. Peak broadening reduced the ability to find specific ions for extracting images. A concomitant factor is the unevenness of the surface of the hinge lifter relative to an aluminum-coated slide. The principle ions formed with fingerprint powder and cyanoacrylate fuming followed by fingerprint powder were as follows: aspirin and ibuprofen [M H] ; TNT [M] ; procaine and pseudoephedrine [M+H]+; and acetaminophen [M+Na]+. An example of natural fingermark residue extracted ion images for controls and aspirin powders using the four processes are found in Fig. 3A–D. The cyanoacrylate development helped confine the sample material to the fingermark residue area which prevented the powder from spreading during the dusting step. However, the chemical background in the spectra from the cyanoacrylate polymer was noticeably higher than the other processing methods.
Fig. 5. Mass spectra and image of [M+H]+ for pseudoephedrine mixed 1:1 with CHCA at 5 different amounts from top to bottom: 2500, 1300, 130, 13, 1.2 ng. Spot sizes are 3 mm.
74
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
The predominate analyte ions observed using CHCA matrix were protonated (positive mode) and deprotonated (negative mode) molecular compounds which are more amenable to the use of MS/MS when compared to other adduct ions (e.g., sodium and potassium). The relative success rate of fingermark development techniques that were able to provide positive results for target compounds based on signal intensity and image were as follows: MALDI matrix > black fingerprint development powder > superglue and black fingerprint development powder > black fingerprint development powder and lifts.
3.3. Analysis of natural fingermark residue after handling residues of target compounds An issue encountered with handling loose analyte powders was the spread of the analytes beyond the drug fingermark on the aluminum-coated slide during dusting with black fingerprint development powder. Vertical barriers were used to isolate neighboring fingermark residues from each other. The appearance of procaine signal beyond the fingermark is seen in Fig. 4A to the right side of the drug print where no barrier was used.
Fig. 6. Mass spectra and image of [M+H]+ for procaine mixed 1:1 with CHCA at 5 different amounts from top to bottom: 2500, 1300, 130, 13, 1.3 ng. Spot sizes 3 mm.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
In order to avoid the issues with loose powders, transferring dried residue was investigated. To establish the appropriate concentration to use for transferring residues, target analytes were spotted 1:1 with CHCA matrix and imaged. An example of mass spectra and an extracted ion image for pseudoephedrine is shown in Fig. 5 for five different amounts ranging from 1.3 ng to 2500 ng. For all five amounts there was a visible response in the image and discernible peaks in the mass spectra for m/z 166 [M+H]+. A duplicate experiment using procaine hydrochloride is demonstrated in Fig. 6. The ion at m/z 237 is the [M+H]+ for procaine. The signal intensity was more intense for procaine when compared with pseudoephedrine. Similar experiments were performed using fingerprint powder as the MALDI matrix (data not shown) and comparable responses were observed. The mass of 1300 ng was chosen for the direct deposit on a slide for fingermark
75
transfers. The drug residue was anticipated to have adequate signal after transferring with finger contact but expected to be at much lower levels. An example of a transfer experiment in fingermark residue for protonated procaine is demonstrated in Fig. 7. Three different volumes (1 mL, 5 mL, and 10 mL) are shown on a single print, and it was determined that a 5 mL spot size was sufficient for detection. When comparing the mass spectra of the analyte powders to the transfer residues, the [M+H]+ S/N is greater in the transfer residue. RDX and TNT were also evaluated using five different amounts ranging from 1.3 ng to 2500 ng. The primary ions measured in the standards were [M H] for TNT and a fragment of RDX [M– CH2N(NO2)2] . For all five amounts there was visible response in the image and the analyte peaks in the mass spectra were above the noise level. The RDX fragment was detected with smaller S/N
Fig. 7. Extracted ion images for [M+H]+ of transferred procaine residue from a glass slide to an aluminum-coated slide. There are three fingermark residues present on the slide: (1) the control, (2) pseudoephedrine, and (3) procaine residue but only the extracted ion m/z 238 for procaine is demonstrated. The transfer residue has three different volumes investigated (1 mL, 5 mL, and 10 mL) at the same amount (1300 ng) with the corresponding mass spectra shown.
76
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77
Fig. 8. Extracted ion images (XIC) for transferred pseudoephedrine, procaine, TNT, and RDX residues. The latent fingermark experimental layout (A) is shown above the XICs. The control fingermark consists of the solvent background (1); the drug fingermark consists of pseudoephedrine (2) and procaine (3); and the explosive fingermark consists of TNT (4) and RDX (5). The fingermarks were developed with black powder and the ions observed were [M+H]+ for pseudoephedrine (B), procaine (C), [M H] for TNT (D), and [M–CH2N(NO2)2] fragment for RDX (E).
than the TNT ion for the different amounts. The mass 2500 ng was selected for direct deposit on the slide for fingermark transfers because the RDX fragment was harder to detect after transferring the explosives residue. The analytes from the transfer residues (procaine, pseudoephedrine, TNT, and RDX) were evaluated after developing the fingermarks using the four different processes. All target drugs and explosives were detected in the appropriate fingermark residues after developing using the four different processes. An example of extracted ion images for all four analytes is shown in Fig. 8 for fingermarks developed with black fingerprint powder. The primary ions detected using all development processes were [M+H]+ for procaine and pseudoephedrine, [M] and [M H] for TNT, and [M-CH2N(NO2)2] fragment for RDX. Transfer residues provided a reliable means of preparing fingermarks with known amounts of analytes as compared to using analyte powders. 4. Conclusion The MALDI/TOF MS was able to detect exogenous substances from contact residue under laboratory conditions with the integration of conventional fingerprint development methods. MALDI/TOF MS is a nondestructive method which provides
chemical information in both the mass spectra and chemical images. Touch chemistry using MALDI/TOF MS shows promising results for investigative and intelligence gathering applications for detecting exogenous compounds. In real-life scenarios of casually handling whole tablets, MALDI/TOF MS was unable to detect drug residue using the four different development processes. Casual interactions with whole or broken tablets are unlikely to result in detectable amounts under the experimental conditions described here. Reference pads were useful in creating MALDI/TOF MS methods for imaging fingermark residue; however, they did not model natural secretions chemically or physically enough to be used as a quality control measure for this study. Positive results were demonstrated for handling analyte powders. However, analyte powders were easily spread throughout the fingermark when developing the fingermarks using black fingerprint development powder and an application brush. In the future, if crime scene technicians collect evidence for chemical analysis of latent fingermark residue, it should be noted that loose powder on the fingermark residue will spread when the fingermark is developed using the powder and brush technique. In addition, it was difficult to reproducibly control the amount of analyte powder in the fingermark residue; field samples could demonstrate a large variability in the amounts of powders collected. Contact residues, which can be helpful for method development but may not resemble real circumstances, were useful in controlling the deposition location and amount of target compounds. Although touch chemistry is time consuming, the information gained may tie an individual to specific evidence. As faster lasers and newer technology evolve, the collection duration can be reduced. Without the chemical image it would be a challenge to draw conclusions due to the large chemical noise from the development process, the natural components in residue, and the background noise from the originating surface. Chemical imaging analysis can be used without impeding the latent print process. Smeared prints or prints that are not of value for identification can still be used for mass spectral analysis. Acknowledgements This is publication 13-14 of the Federal Bureau of Investigation Laboratory Division. This research was supported in part by an appointment to the Visiting Scientist Program at the Federal Bureau of Investigation Laboratory Division, administered by the Oak Ridge Institute for Science and Education, through an interagency agreement between the U.S. Department of Energy and the FBI. Names of commercial manufacturers are provided for identification purposes only and inclusion does not imply endorsement of the manufacturer or its products or services by the FBI. The views expressed are those of the authors and do not necessarily reflect the official policy or position of the FBI or the U.S. Government. References [1] T.A. Trozzi, R.L. Schwartz, M.L. Hollars, Processing guide for developing latent prints, U.S. Department of Justice, FBI, Laboratory Division, Forensic Sci. Commun. 3 (2001) 1 http://www.fbi.gov/about-us/lab/forensic-science-communications/ fsc/jan2001. [2] F. Rowell, J. Seviour, A.Y. Lim, C.G. Elumbaring-Salzar, J. Loke, J. Ma, Detection of nitro-organic and peroxide explosives in latent fingermarks by DART- and SALDITOF-mass spectrometry, Forensic Sci. Int. 221 (2012) 84–91. [3] F. Rowell, K. Hudson, J. Seviour, Detection of drugs and their metabolites in dusted latent fingermarks by mass spectrometry, Analyst 4 (2009) 701–707. [4] M. Benton, F. Rowell, L. Sundar, M. Jan, Direct detection of nicotine and cotinine in dusted latent fingermarks of smokers by using hydrophobic silica particles and MS, Surf. Interface Anal. 42 (2009) 378–385. [5] R.S. Croxton, M.G. Baron, D. Butler, T. Kent, V.G. Sears, Variation in amino acid and lipid composition of latent fingerprints, Forensic Sci. Int. 199 (2010) 93–102.
K. Kaplan-Sandquist et al. / Forensic Science International 235 (2014) 68–77 [6] K.M. Antonie, S. Mortazavi, A.D. Miller, L.M. Miller, Chemical differences are observed in children’s versus adults’ latent fingerprints as a function of time, J. Forensic Sci. 55 (2010) 513–551. [7] G.M. Mong, C.E. Petersen, T.R.W. Clauss, Advanced fingerprint analysis project fingerprint constituents, Pacific Northwest National Laboratory, 1999, PNNL13019, pp. 1–14. Retrieved from http://www.osti.gov/bridge/servlets/purl/ 14172-SQLzxz/webviewable/14172.pdf on June 12, 2013. [8] L.S. Ferguson, F. Wulfert, R. Wolstenholme, J.M. Fonville, M.R. Clench, V.A. Carolan, S. Francese, Direct detection of peptides and small proteins in fingermarks and determination of sex by MALDI mass spectrometry profiling, Analyst 137 (2012) 4686–4692. [9] T. Chen, Z.D. Schultz, I.W. Levin, Infrared spectroscopic imagining of latent fingerprints and associated forensic evidence, Analyst 134 (2009) 1902–1904. [10] B. Hartzell-Baguley, R.E. Hipp, N.R. Morgan, S.L. Morgan, Chemical composition of latent fingerprints by gas chromatography–mass spectrometry, J. Chem. Ed. 84 (2007) 689–691. [11] P. Hazarika, S.M. Jickells, D.A. Russell, Rapid detection of drug metabolites in latent fingermarks, Analyst 134 (2009) 93–96. [12] D.R. Ifa, N.E. Manicke, A.L. Dill, R.G. Cooks, Latent fingerprint chemical imaging by mass spectrometry, Science 321 (2008) 80. [13] M.I. Szynkowska, K. Czerski, J. Rogowski, T. Paryjczak, A. Parcewski, Detection of exogenous contaminants of fingerprints using ToF-SIMS, Surf. Interface Anal. 5 (2010) 393–397.
77
[14] M.J. Bailey, N.J. Bright, R.S. Croxton, S. Francese, L.S. Ferguson, S. Hinder, S. Jickells, B.J. Jones, B.N. Jones, S.G. Kazarian, J.J. Ojeda, R.P. Webb, R. Wolstenholme, S. Bleay, Chemical characterization of latent fingerprints by matrix-assisted laser desorption ionization, time-of-flight secondary ion mass spectrometry, mega electron volt secondary mass spectrometry, gas chromatography/mass spectrometry, Xray photoelcectron spectroscopy, and attenuated total reflection Fourier transform infrared spectroscopic imaging: an intercomparison, Anal. Chem. 84 (2012) 8514–8523. [15] R. Wolstenholme, R. Bradshaw, M.R. Clench, S. Francese, Study of latent fingermarks by matrix-assisted laser desorption/ionization mass spectrometry imaging of endogenous lipids, Rapid Commun. Mass Spectrom. 23 (2009) 3031–3039. [16] L. Ferguson, R. Bradshaw, R. Wolstenholme, M. Clench, S. Francese, Two-step matrix application for the enhancement and imaging of latent fingermarks, Anal. Chem. 83 (2011) 5585–5591. [17] M. Benton, M.J. Chua, F. Gu, F. Rowell, J. Ma, Environmental nicotine contamination in latent fingermarks from smoker contacts and passive smoking, Forensic Sci. Int. 200 (2010) 28–34. [18] P.H. Ng, S. Walker, M. Tahtouh, B. Reedy, Detection of illicit substances in fingerprints by infrared spectral imaging, Anal. Bioanal. Chem. 394 (2009) 2039–2048. [19] K. Dreisewerd, The desorption process in MALDI, Chem. Rev. 103 (2003) 395–425. [20] A.G. Vereshchagin, G.V. Novitskaya, The triglyceride composition of linseed oil, J. Am. Oil Chem. Soc. 42 (1965) 970–974.