Test Sample for the Spatially Resolved Quantification of Illicit Drugs on

Apr 27, 2015 - Netherlands Forensic Institute (NFI), Ministry of Security and Justice, The Hague 1098 XH, The Netherlands. § van't Hoff ... Citation ...
0 downloads 9 Views 1MB Size
Subscriber access provided by NATIONAL CENTRAL UNIV

Article

A Novel Test Sample for the Spatially Resolved Quantification of Illicit Drugs on Fingerprints using Imaging Mass Spectrometry Shin Muramoto, Thomas P. Forbes, Arian C. van Asten, and J. Greg Gillen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b01060 • Publication Date (Web): 27 Apr 2015 Downloaded from http://pubs.acs.org on May 3, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A Novel Test Sample for the Spatially Resolved Quantification of Illicit Drugs on Fingerprints using Imaging Mass Spectrometry Shin Muramoto1, Thomas P. Forbes1, Arian C. van Asten2,3,4, and Greg Gillen1 1

National Institute of Standards and Technology (NIST), US Department of Commerce, Gaithersburg, Maryland, USA 2 Netherlands Forensic Institute (NFI), Ministry of Security and Justice, The Hague, The Netherlands 3 van ‘t Hoff Institute for Molecular Sciences, Faculty of Science, University of Amsterdam, Amsterdam, The Netherlands 4 Amsterdam Center for Forensic Science and Medicine (CLHC), University of Amsterdam, Amsterdam, The Netherlands

*Corresponding Author: Shin Muramoto National Institute of Standards and Technology, Gaithersburg, MD, USA 1-301-975-5997 (phone) 1-301-417-1321 (fax) [email protected]

Keywords: tof-sims; desi; chemical imaging; fingerprints; forensics; illicit drugs

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 16

Abstract A novel test sample for the spatially resolved quantification of illicit drugs on the surface of a fingerprint using time-of-flight secondary ion mass spectrometry (ToF-SIMS) and desorption electrospray ionization mass spectrometry (DESI-MS) was demonstrated. Calibration curves relating the signal intensity to the amount of drug deposited on the surface was generated from inkjet-printed arrays of cocaine, methamphetamine, and heroin with a deposited-mass ranging nominally from 10 pg to 50 ng per spot. These curves were used to construct concentration maps that visualized the spatial distribution of the drugs on top of a fingerprint, as well as being able to quantify the amount of drugs in a given area within the map. For the drugs on the fingerprint on silicon, ToF-SIMS showed great success as it was able to generate concentration maps of all three drugs. On the fingerprint on paper, only the concentration map of cocaine could be constructed using ToF-SIMS and DESI-MS as the signals of methamphetamine and heroin were completely suppressed by matrix and substrate effects. Spatially resolved quantification of illicit drugs using imaging mass spectrometry is possible, but the choice of substrates could significantly affect the results.

ACS Paragon Plus Environment

Page 3 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Introduction Chemical imaging of latent fingerprints using mass spectrometry (MS) techniques is an area of increasing interest due to the enormous amount of forensic information that can potentially be extracted. In addition to imaging of fingerprint details for the identification of the donor (dactyloscopic investigation), its molecular specificity allows the discrimination of specific endogenous compounds related to an individual’s age,1, 2 and quite possibly gender,3 that can aid in the identification of a suspect. It also allows identification of exogenous compounds such as explosives and drugs that can potentially connect a suspect to a crime scene.4, 5 The technique’s strength lies in its ability to detect multiple chemical species, which can significantly reduce analysis time and expense, at the same time completely eliminating the need for markers such as tagged-antibodies.4, 6-8 One MS imaging technique, time-of-flight secondary ion mass spectrometry (ToF-SIMS), offers precise identification of molecules with a mass resolving power (m/∆m) of 5,000 to 10,000 with picogram to even femtogram sensitivity for organic molecules on surfaces.9, 10 Its sub-micrometer spatial resolution can offer detailed images of fingerprints,5, 11 able to show not only secondary-level details such as bifurcations and ridge ending features, but also tertiary-level details such as shape and size of the sweat pores, which are starting to be recognized as discriminative features for identification.12, 13 When operated in static-conditions (primary ion dose of 1012 ions/cm2, where 1% of the surface is theoretically sampled), mass spectra with sufficient data can be generated while keeping material consumption to a minimum and the spatial distribution of the molecules intact. The ability to mitigate sample charging allows visualization of fingerprints directly on a wide variety of surfaces such as glass and paper.14 Another potential MS imaging approach relevant to the chemical imaging of fingerprints is the droplet-based ambient ionization technique desorption electrospray ionization (DESI). It works in ambient pressure by using a high velocity spray of charged solvent droplets directed at the surface for desorption and ionization of molecules, followed by mass analysis at high vacuum. It has been used to produce chemical images of latent fingerprints doped with explosives and/or illicit drugs,15, 16 with the latter showing a sensitivity of 10 ng/mm2 as tested on the fingerprint. The spatial resolution is rather low (> 100 µm) due to physical limitations in minimal jet diameter and flow rate for stable jetting,16, 17 which currently precludes the imaging of tertiary level details. The biggest advantage is its simplicity and lower cost, and its ability to analyze samples in atmosphere for a significantly reduced analysis time. Analysis at ambient pressures also mitigates any potential changes in the chemistry of the fingerprint through exposure to vacuum.18

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 16

In comparison to bulk quantification methods such as gas or liquid chromatography (GC- or LCMS), spatially resolved quantification at the surface can provide essential forensic information regarding the fingerprint donor. For example, analyte concentration coupled with spatial information may reveal the age of a fingerprint through the extent of surface diffusion of endogenous compounds with time.19 Concentration maps can also show local regions of high concentrations or “hot spots” of drugs on the surface, and help reveal the order of deposition of drugs relative to the fingerprint by identifying where in the fingerprint the drugs are localized (i.e., valley or ridge of a fingerprint). This manuscript introduces a novel method for the spatially resolved quantification of drugs on the surface of a fingerprint. The proposed approach is an inkjet-printed deposition of a known amount of drugs on a matrix, from which a calibration curve relating the signal intensity to the amount of drugs is generated. This information is then used to construct a concentration map of the drug. A piezo-driven drop-on-demand printer capable of depositing material with a precision of 0.2%20 was used to print an array of droplets containing a nominal mass of 8 pg to 50 ng of cocaine, methamphetamine, and heroin. A simulated fingerprint was deposited using a 3D-printed plastic finger and artificial sebum21 to control the friction ridge pattern and chemistry, since the chemical composition of a natural sebum changes with time,18, 22 and from fingerprint to fingerprint due to contributions from both endogenous and exogenous constituents. The drugs were deposited onto four surfaces; a bare silicon wafer, bare paper, simulated fingerprint on a silicon wafer, and simulated fingerprint on paper, to investigate the effect of matrix and substrate effects on analyte signal response.

Experimental Sample Preparation. Powdered cocaine hydrochloride, heroin hydrochloride, and methamphetamine hydrochloride were purchased from Sigma-Aldrich Co.1 (St. Louis, MO), and dissolved in ultrapure water at a concentration of 25 mg/mL. 1 inch diameter Si(100) wafers were purchased from Virginia Semiconductors (Fredericksburg, VA), and ultrasonicated sequentially in acetone, methylene chloride, and methanol (Sigma-Aldrich Co., St. Louis, MO) for 10 min. A thin layer of fluorocarbon was then self-assembled onto the surface by molecular vapor deposition, where the silicon wafers were exposed overnight to 5 mg of 1H,1H,2H,2H-perfluorodecyltrichlorosilane (FDTS) purchased from Alfa Aesar (Ward Hill, MA) inside a 1 L Pyrex vacuum desiccator. The water contact 1

Certain commercial equipment, instruments, or materials are identified in this paper to adequately specify the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

ACS Paragon Plus Environment

Page 5 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

angle of the film was (107.9 ± 3.1)°, and its coverage appeared uniform, which can be seen in Figure S-1, Supporting Information. A Hammermill® copy paper (International Paper Co., Memphis, TN) was also used as a substrate, which was used as received. Artificial fingerprint sebum. The artificial fingerprint sebum is a mixture of equal parts sebaceous and eccrine components (see Table S-1 in Supplemental Information for a list of the ingredients and directions for preparation), created by mixing together chemical constituents that were identified in literature,22 and also those that were detected in natural fingerprints residue using SIMS, FTIR, and GCMS.21 The artificial finger mold was printed using an Objet350 Connex 3D printer (Stratasys, Eden Prairie, MN), using their proprietary photocurable ink. The ridge details were altered from the original donor’s fingerprint for privacy. Simulated fingerprints were made by spreading 10 µL of the artificial sebum onto the finger mold, and stamping onto substrates. The drug-sebum mixture was prepared by combining 2 µL of analyte (2.5 mg/mL in methanol) with 20 µL of artificial sebum. 10 µL of this emulsion was applied to the finger mold. Drop-on-Demand Inkjet Printer. The drug molecules dissolved in water were printed using a Jet Lab 4 (MicroFab, Plano, TX) piezoelectric drop-on-demand materials deposition printer system equipped with a 55 µm ID orifice print head. Operating conditions for the piezoelectric dispenser varied slightly day-to-day, but typically were 19 µs for Dwell time, 3 µs for the Rise, Fall, and Rise2 times, 6 µs for Echo time, 23 V for the Dwell voltage, and -9 V for the Echo voltage. The droplet frequency was kept at 250 Hz. Masses of the drug molecules deposited were determined by weighing the drops with an analytical microbalance; the mass of a burst of 20,000 drops in 100 drop intervals were obtained and averaged to determine the average mass of a single drop after correcting for solvent evaporation. Desorption Electrospray Ionization. The DESI source consisted of a Prosolia Omni Spray® ion source described in detail elsewhere.23 Briefly, acetonitrile (Sigma Aldrich) was delivered at a flow rate of 2.5 µL/min with the spray directed toward the surface at an incidence angle of 45° using a charging potential of +4000 V. Droplet nebulization was pneumatically-assisted by a N2 coaxial carrier gas supplied at (690 ± 35) kPa (100 psi). Mass analysis was performed using the Applied Biosystems/MDS Sciex 4000 QTrap mass spectrometer (Framingham, MA), with the following parameters: curtain gas pressure of 138 kPa (20.0 psi), ion source gas pressure of 83 kPa (12.0 psi), interface heater temperature of 150 °C; declustering potential of +130 V; and an entrance potential of +100 V. The motorized stage moved at a scan rate of 350 µm/s with a step/pixel size of 100 µm. A

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 16

MATLAB-based freeware package, MSiReader (v0.04, W. M. Keck FT-ICR Mass Spectrometry Laboratory, North Carolina State University),24 was used for the construction of the images. Time-of-Flight Secondary Ion Mass Spectrometry. Samples were imaged using an IONTOF IV (Münster, Germany) instrument equipped with a 25 kV Bi3+ analysis source, operated at a current of 0.12 pA pulsed at 10 kHz. The beam was rastered within a 500 µm × 500 µm area with a pixel density of 128 × 128 pixels, with 5 pulses at each pixel. These scans were stitched together to create larger images. The ion dose density used was 3.27 × 109 ions/cm2 and 1.64 × 109 ions/cm2 for the analysis of the inkjet-printed arrays and the pre-mixed fingerprint sebum, respectively, both of which were below the static limit of 1 × 1012 ions/cm2. A low energy electron flood gun was used during the analysis on paper for charge mitigation. The peaks for cocaine, methamphetamine, and heroin appeared as molecular ions ([M+H]+) at m/z 304, m/z 150, and m/z 370, respectively. All data points reflect the average of at least 5 replicate samples, and the error bars represent their standard deviations. Particle size measurements were made and concentration maps were constructed from ion images processed using ImageJ (National Institutes of Health, Bethesda, MD).25 A circular region of interest (ROI) with a diameter of 500 µm was used to extract the intensity of analytes at each spot, large enough to encompass the coffee rings seen around the drops.

Results and Discussion A 10 × 10 array of drug molecules was printed onto simulated fingerprints to investigate optimal deposition parameters. First, inkjet printing was optimized by adjusting parameters that yielded a solid droplet without any satellites.26 Figure 1 shows the ToF-SIMS ion images of the arrays containing cocaine, illustrating the printer’s capability to deposit a consistent volume of solution at spatially defined positions. Each spot in the array contained just 1 drop of the drug solution, with a volume of roughly (33.7 ± 0.1) pL that led to the deposition of (0.84 ± 0.01) ng of analyte per spot. The substrate was an FDTS-modified silicon to prevent droplets from wetting the surface and spreading during printing. The deposition of the simulated fingerprint was optimized by varying the number of times the sebum-laden finger mold was “pre-stamped” onto a clean aluminum foil before being stamped onto the silicon wafer, where Figures 1a, 1b, and 1c correspond to 0 to 3 to 6 “pre-stamps”, respectively. Fewer stamps onto the aluminum foil corresponded with a thicker sebum film on the silicon wafer. This was important because the behavior of the droplet once it touches the surface is dictated by the thickness of the sebum layer (as observed visually); a thinner layer causes the droplets to bead up on the fingerprint

ACS Paragon Plus Environment

Page 7 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

from contributions of the low energy surface, while a thicker layer causes the droplets to wet the fingerprint, presumably by shielding the hydrophobic effects of the surface. As seen in Figures 1a and 1b, spots appear to be very well confined on the fingerprint with three “pre-stamps” while those on the fingerprint with zero “pre-stamps” show a greater degree of wetting. Excessive “pre-stamping” led to an incomplete coverage of sebum on the silicon wafer. On a natural fingerprint, the droplets were seen to bead up on the surface, so a thinner sebum layer was definitely preferred. Thus, three “pre-stamps” were performed for the preparation of all simulated fingerprints in this study. (a) 0 Pre-Stamps 304

(b) 3 Pre-Stamps K

+

C3H7O

1 mm

+

304

+

+

304/C3H7O /K

(c) 6 Pre-Stamps K

+

C3H7O

+

304

+

+

304/C3H7O /K

1 mm

K

+

C3H7O

+

+

+

304/C3H7O /K

1 mm

Figure 1. ToF-SIMS ion images showing the array of cocaine printed on top of a simulated fingerprint, deposited after (a) zero, (b) three, and (c) six “pre-stamps” on a clean aluminum foil. The images show the distribution of the molecular ion of cocaine, a constituent of the eccrine component K+, a constituent of the sebum component C3H7O+, and an overlay of those images. Image size is 3 mm × 3 mm. ToF-SIMS analysis of calibration arrays on silicon. To build a calibration curve for quantitative analysis, calibration arrays containing different amounts of analyte per spot were printed. To cover a range of 8 pg to 50 ng, two arrays were generated using two solution concentrations; a 0.1 mg/mL solution was used to deposit a range of nominal mass of (8, 40, 80, 200, 800) pg/spot, and an 8.3 mg/mL solution was used to deposit spots containing (0.8, 5, 10, 25, and 50) ng/spot. This was printed on both the bare FDTS-modified silicon wafer and on the simulated fingerprint to investigate the effects of sebum matrix on the ionization suppression of the analyte. An example of this calibration array is shown in Figure 2, showing the ToF-SIMS ion images of the heroin array printed on the two surfaces (ion images of the cocaine and methamphetamine arrays appeared very similar and are provided in Figure S-1, Supplemental Information). The diameters of the spots were larger on the fingerprint as expected, a 60% increase relative to those on the silicon surface (see Table S-2 for a list of spot diameters on silicon and on the fingerprint, Supplemental Information). On the silicon surface, the spots

ACS Paragon Plus Environment

Analytical Chemistry

were surrounded by a faint ring of analyte, showing the footprint of the initial drop which shrank over time as the droplet evaporated. A portion of the ring may have formed as a result of the droplet spreading and rebounding during impingement,27, 28 but this could not be confirmed without a highspeed camera. Figure 3a shows the secondary ion intensity of the drug molecules plotted as a function of the mass deposited. Here, absolute ion intensities were used to compare ion intensities between ToF-SIMS and DESI, but the use of intensity ratios (matrix adjusted by normalizing to C3H7O+) also gave very similar plots. The calibration curves for the most part were linear on the log-log plot, showing that discrete intensities could be seen for all three drugs down to 8 pg on both the bare surface and on the fingerprint. One exception to this observation was methamphetamine on fingerprint, where the minimum detectable mass was around 1 ng.

(b) ToF-SIMS Image of Array on Fingerprint

1 mm

44.0 ng/spot

+

4.4 ng/spot

0.9 ng/spot

44.0 ng/spot

22.0 ng/spot

8.8 ng/spot

4.4 ng/spot

370/C3H7O /K

22.0 ng/spot

+

370

8.8 ng/spot

(a) ToF-SIMS Image of Array on Si

0.9 ng/spot

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 16

1 mm

Figure 2. ToF-SIMS images of the heroin calibration array on (a) the bare fluorinated silicon wafer and (b) the simulated fingerprint deposited on the fluorinated silicon wafer. For (b), the overlay shows the distribution of heroin in green, C3H7O+ in red, and K+ in blue. Five replicate rows are shown, with deposited mass per spot increasing from left to right within each row. Image size is 9 mm × 6 mm. The figure also showed that the presence of sebum can significantly suppress the signal of the drug molecules. This matrix-induced signal suppression is thought to be caused by two factors. One is due to the free fatty acids in sebum, whose pKa varies from 7.3 to 8.8 depending on chain length;29, 30 they would neutralize charge by abstracting the protons from the matrix, leaving the analyte molecular ions deprived of protons for ionization. The other possible cause of signal suppression is the topography of the fingerprint surface, where microscale height differences and curved surfaces would induce signal loss through the distortion of the extraction field.31 Interestingly, some spots were more affected by

ACS Paragon Plus Environment

Page 9 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

signal suppression than others (e.g., spots in the 22 ng/spot and 44 ng/spot columns in Figure 2b displayed varying intensities), showing that the matrix effect was not consistent. More interestingly, the molecules experienced increasing suppression for larger deposited mass. For example, cocaine molecules on the fingerprint exhibited a factor of 2 lower intensity in the 0.01 ng to 0.1 ng range, but showed a loss of a factor of ten around the 10 ng to 100 ng range. The increased suppression with deposited mass suggests a “saturation” of the ions due to an excessive number of analytes competing for charge,32, 33 but this did not seem likely since analytes on the bare silicon wafer did not experience a similar phenomenon. The cause of mass-dependent suppression is not clear at this time. Finally, an overlap of mass at 0.8 ng/spot was made to examine the effect of spot diameter on signal intensity. A larger deposit was made using 100 drops of the 0.1 mg/mL solution, and a smaller deposit was made using 1 drop of the 8.3 mg/mL solution, but both contained roughly the same number of analytes per spot. The diameter of the 100-drop spot was larger by roughly a factor of two with respect to the smaller deposit, equating to a roughly five times increase in surface area. Despite this difference, the average intensity was higher only by up to 20%. The increase in intensity is due to the spreading of the molecules over a larger area, exposing more molecules to the sampling volume of the technique,34 but the disproportionate increase in intensity suggests that the thicknesses of the deposits may not be drastically different. Furthermore, the intensities at the overlapping mass were well within error of each other, and did not significantly affect the trend of the curves. This was especially true for drugs on sebum, where the difference in intensity was generally less than 20%, with occasionally the smaller deposit showing higher intensity. Therefore, for the spatially resolved quantification of these molecules using ToF-SIMS, it was concluded that the deposited volume does not play a significant role. ToF-SIMS analysis of calibration arrays on paper. The same experiment was repeated using paper as the substrate, a porous material that makes analysis difficult due to analyte absorption.35 Figure 3b shows the secondary ion intensity of the drug molecules plotted as a function of the mass deposited. The relationship between intensity and deposited mass was generally linear on the log-log plot similar to what was seen on silicon. The exception was heroin which displayed a constant intensity of (375 ± 55) counts below a deposited-mass of 1 ng, statistically similar to the baseline of (341 ± 67) counts obtained from a blank fingerprint, which indicated that a detection limit was reached. Interpolation led to an instrument detection limit of 4.1 ng, corresponding to an intensity of 540 counts (three standard deviations above the observed baseline). The intensities of the three drug molecules were much lower

ACS Paragon Plus Environment

Analytical Chemistry

(a) On Silicon Wafer

y = 6960x0.434 R² = 0.9736

1.0E+03 1.0E+02

1401.1x0.2231

y= R² = 0.884

1.0E+01 0.001

Intensity (counts)

1.0E+05 1.0E+04

0.01

0.1

1

10

Intensity (counts)

Cocaine

y = 6804.1x0.3975 R² = 0.982

1.0E+03 y = 59.428x + 294.72 R² = 0.9841

1.0E+01 0.001

0.01

0.1

1

Heroin

1.0E+03

y = 1285.2x0.4805 R² = 0.9717

10

1.0E+02

y = 171.37x0.2212 R² = 0.9221 0.1

1

10

100

y = 131.95x0.2073 R² = 0.8063

y = 72 0.01

0.1

1

10

100

Methamphetamine y = 142x0.2086 R² = 0.913

1.0E+02

y = 76 1.0E+01 0.001

100

1.0E+02

0.01

1.0E+03

y = 444.44x0.2601 R² = 0.9435

1.0E+04

1.0E+04

1.0E+01 0.001

Cocaine

1.0E+03

Methamphetamine

1.0E+02

1.0E+04

1.0E+01 0.001

100

Intensity (counts)

1.0E+04

(b) On Paper

Intensity (counts)

Intensity (counts)

1.0E+05

Intensity (counts)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 16

0.01

0.1

1

10

100

Heroin

1.0E+03

y = 61.552x + 449.32 R² = 0.9822

1.0E+02

1.0E+01 0.001

Deposited Mass (ng)

y = 257

0.01

0.1

1

10

100

Deposited Mass (ng)

Figure 3. ToF-SIMS secondary ion intensity as a function of deposited-mass from calibration arrays of cocaine, methamphetamine, and heroin printed on (a) fluorinated silicon wafers and (b) paper. The arrays were inkjet-printed onto bare substrates (solid markers) and on the simulated fingerprint (open markers).

compared to their intensities on silicon; by about a factor of two for heroin, and roughly a factor of ten for cocaine and methamphetamine. In addition to the loss of analyte into the paper fiber, a significant

ACS Paragon Plus Environment

Page 11 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

source of signal loss is thought to be due to ionization suppression caused by the presence of hydroxyl groups in cellulose. These groups are basic with a pKa of roughly 13,36, 37 and is thought to suppress the ionization of the drug molecules more strongly than the fatty acids with a pKa between 7.3 to 8.8. The presence of sebum further suppressed their intensities, suggesting that matrix and substrate effects may be additive. Relative to the intensities of cocaine molecules on paper, their intensities on the fingerprint were lower by a factor of three across the range of deposited-mass tested. A constant intensity of (72 ± 5) counts was seen below a deposited-mass of 100 pg, corresponding to a detection limit of 130 pg. The baseline intensity of (63 ± 7) counts obtained from a blank fingerprint suggested that there were no intensity contributions from another molecule. For methamphetamine and heroin, constant intensities of (76 ± 19) counts and (257 ± 32) counts, respectively, were seen for all depositedmass below 50 ng. It appeared that the additional suppression from the sebum was enough to suppress their intensities to the baseline intensities obtained the blank fingerprint, which were (62 ± 23) counts and (233 ± 41) counts, respectively. The detection limits could not be determined since a deposited-mass higher than 50 ng was not prepared. In the ion images, the methamphetamine and heroin arrays could not be visualized, but the location of the individual spots was identified by the suppressed K+ signal (Figure S-2, Supplemental Information). ToF-SIMS analysis of drug-sebum mixtures. The calibration curves generated in Figure 3 were used to construct a concentration map of the drug molecules on top of a fingerprint on both the silicon wafer and paper. 10 µL (2.5 µg) of the drug-sebum mixture was applied to the finger mold, “prestamped”, then stamped onto the substrates. On silicon, the measured intensities of the drug molecules within the 5 mm × 5 mm images were 5.93 × 103 counts for cocaine, 1.70 × 103 counts for methamphetamine, and 5.13 × 102 counts for heroin. Using their respective calibration curves obtained in Figure 3a, the corresponding amount of drugs present inside the image was 57.0 ng of cocaine, 23.6 ng of methamphetamine, and 12.4 ng of heroin (for cocaine and heroin, the calculation was performed on half of the image at a time so that their intensities fell within the range expressed in the calibration curves). Figure 4a shows the ToF-SIMS images of a 5 mm × 5 mm area of the fingerprint converted into concentration maps (see Figure S-3 in Supplemental Information for details). Physically, there appeared to be large variations in the friction ridge patterns of the fingerprints, despite the fact that similar forces were applied during deposition and roughly the same location on the fingerprint was analyzed (the “ribbon-shaped” artifact at the bottom center portion of the images can be seen). Chemically, the

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 16

concentration maps revealed that the distributions of the molecules were also entirely different; cocaine molecules seemed to be present everywhere within the ridges in uniform concentrations while the methamphetamine molecules were found in high density areas or “hot spots”. In comparison, heroin molecules seemed to be present uniformly throughout, with higher concentrations localized around the perimeter of the ridges.

(a) ToF-SIMS Images of Drug/Sebum Mixture on Silicon Cocaine

3.8 pg/ Methamphetamine pix

1.9 pg/ pix

0 pg/pix

0 pg/pix

(b) ToF-SIMS Images of Mixture on Paper Cocaine

1.4 pg/ pix

Heroin

0.5 pg/ pix

0 pg/pix

(c) DESI-MS Images of Mixture on Paper Cocaine

0 pg/pix

416 pg/ pix

0 pg/pix

Figure 4. ToF-SIMS concentration maps showing the distribution of the drug molecules in a drug-sebum mixture deposited on (a) a modified silicon wafer and (b) paper. On paper, only the cocaine molecules could be visualized. Image size is 5 mm × 5 mm (320 pixels × 320 pixels). (c) DESI-MS concentration map of cocaine deposited on paper. Image size is 14 mm × 9 mm (140 pixels × 90 pixels). The concentration map is useful for calculating the amount of drugs present within a region of interest (ROI). For example, the “ribbon-shaped” artifact was found to contain roughly 2.6 ng of cocaine and 1.2 ng of methamphetamine on the surface of the fingerprint. The calculation is performed by selecting an ROI within the image, extracting a histogram from within the ROI, and using the concentration scale to calculate the amount of drug present. At first glance, the artifact containing methamphetamine appears to have more material due to its hotter temperature color, but it turns out the

ACS Paragon Plus Environment

Page 13 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

artifact containing cocaine is slightly bigger in size at 2,769 pixels (versus 2,043 pixels for methamphetamine) and uses a higher concentration for the scale bar. On paper, only the distribution of cocaine molecules could be visualized because molecular signals of both methamphetamine and heroin were completely suppressed. The intensity of cocaine within the image was 3.72 × 102 counts, corresponding to roughly 10.5 ng of molecules on the surface of the fingerprint within the 5 mm × 5 mm image. For methamphetamine and heroin, their molecular ion intensities were 1.21 × 102 counts and 1.45 × 102 counts, respectively, which are roughly the same baseline intensities seen in Figure 3b. The concentration map of cocaine on sebum is displayed in Figure 4b, and shows that the fingerprint ridge patterns can still be visualized on paper. The cocaine molecules are seen only on the paper fibers and not in the voids, suggesting that a large fraction of the sebum mixture was absorbed deep into the substrate beyond the sampling depth of the instrument. A large fraction of the image appeared to have suffered from signal suppression, as can be seen by the dark areas in the image. Whether this is an artifact of the fingerprint deposition process (i.e., uneven pressure across the finger during deposition) or caused by charging is not known at this time. Analysis of calibration arrays using DESI-MS. While ToF-SIMS is fully capable of spatially quantifying the drug molecules on a fingerprint on both silicon and paper, due to the high cost of implementing and maintaining a ToF-SIMS instrument, less expensive systems that are easier to operate need to be evaluated as potential chemical imaging instruments for implementation in a crime lab. Here, DESI-MS is evaluated as the next generation chemical imaging tool for the spatially resolved quantification of analyte. Paper was used as the sole substrate since the high velocity liquid droplets were seen to displace the arrays on the modified silicon surface, limiting the ability to acquire spatially resolved images. Similar to the plots in Figure 3, the relationship between the DESI intensity and deposited-mass was linear on the log-log plot for all three drugs deposited on bare paper (shown in Figure S-4, Supplemental Information, due to page limitations). For the drug molecules printed on top of the simulated fingerprints, considerable signal suppression was seen for all drug molecules; cocaine molecules were seen to decrease by a factor of 10 across the range of mass tested, while the intensities of methamphetamine and heroin were completely suppressed and could not be detected. Analysis of lower deposited mass of cocaine (8 pg/spot to 800 pg/spot) showed that no signal could be detected, indicating a limit of detection of above 1 ng of cocaine for this instrument on the fingerprint.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 16

DESI-MS analysis of drug-sebum mixtures. Since the signals for methamphetamine and heroin were completely suppressed on the fingerprint, only the calibration curve for cocaine was evaluated. 10 µL (2.5 µg) of the cocaine-sebum was applied to the finger mold and stamped onto the bare paper. Figure 4c shows the DESI-MS images of a 14 mm × 9 mm area of the fingerprint deposited on paper, showing the distribution of cocaine molecules. Similar to what was observed for ToF-SIMS, the fingerprint ridge patterns could be clearly visualized, with areas of suppressed intensities. Whether this is due to pressure points during fingerprint deposition or local charging is still not clear. Also, the image shows a gradient in intensity from top to bottom, potentially indicating that the solvent spray may have moved the molecules around on the surface. Even though the image was acquired from top to bottom with the solvent spray pointing in the upward direction (top of the paper), there are micro-eddies that form behind the incident spray38 which could lead to undesired movement of molecules. The molecular intensity versus deposited-mass plot in Figure S-4 (supplemental information) showed that the relationship between intensity and deposited-mass of cocaine can be expressed using the equation 2.397 × 10 + 5.098 × 10 . Using the measured intensity of 4.06 × 107 counts, the amount of cocaine was determined to be 374 ng within the 14 mm × 9 mm image. The amount of cocaine extracted was roughly three orders of magnitude higher than what was seen using ToF-SIMS. This highlights one of its potential advantages, which is its much deeper information depth that can be exploited to extract more analyte from the surface. For samples with trace levels of material, being able to extract more signal is always desirable.

Conclusion A novel test sample for the spatially resolved quantification of illicit drugs on top of a fingerprint using ToF-SIMS and DESI-MS was demonstrated. The results indicated that drugs printed in an array with increasing deposited-mass per spot could be used to generate a calibration curve, from which a concentration map can be constructed for the quantitative visualization of drugs on the surface. For the drugs printed on the fingerprint on silicon, ToF-SIMS showed great success as it was able to obtain discrete intensities of drugs for deposited-mass between 8 pg and 50 ng, with the exception of methamphetamine whose minimum detectable mass was around 1 ng. The calibration curves were able to generate concentration maps of the drugs on the fingerprint. Unfortunately DESI-MS was not able to obtain intensity information due to its high velocity spray displacing the printed spots.

ACS Paragon Plus Environment

Page 15 of 16

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

On the fingerprint on paper, both ToF-SIMS and DESI-MS were met with some success for the quantitative analysis of cocaine. The analysis of methamphetamine and heroin, in comparison, was difficult as their signals were completely suppressed by matrix effects originating from the fatty acids in sebum and substrate effects from the hydroxyl groups of the sugar molecules. What was learned was that the lowest detectable amount of drugs on the fingerprint on paper was very similar for both ToFSIMS and DESI-MS; around 1 ng for cocaine, and above 50 ng for methamphetamine and heroin. The application of the calibration curve for quantification of drugs on the surface of a fingerprint is still a work in progress, but initial indications are promising. Future work will focus on increasing the number of illicit drugs for quantitative analysis, as well as seeking ways to improve analyte sensitivity on porous or fibrous substrates. To enhance the quality and reliability of the data, sample preparation methodologies will be optimized for the consistent deposition of the simulated fingerprint. Also, controlling the amount of drugs within a fingerprint will be investigated by examining how the sequential deposition of the fingerprint changes the concentration and distribution of drugs on the surface.

Acknowledgement The authors would like to thank Matthew Staymates for the printing of the plastic finger, and Edward Sisco Ph.D. for the formulation of artificial sebum. Research was performed in part at the NIST Center for Nanoscale Science and Technology. This work was the result of a collaboration with NIST and the NFI through the research visit of A.C. van Asten at NIST on May 2014.

References 1. K. M. Antoine, S. Mortazavi, A. D. Miller and L. M. Miller, J. Forensic Sci., 2010, 55, 513-518. 2. D. K. Williams, C. J. Brown and J. Bruker, Forensic Sci. Int., 2011, 206, 161-165. 3. K. G. Asano, C. K. Bayne, K. M. Horsman and M. V. Buchanan, J. Forensic Sci., 2002, 47, 805807. 4. R. Leggett, E. E. Lee-Smith, S. M. Jickells and D. A. Russell, Angew. Chem., 2007, 119, 41784181. 5. M. I. Szynkowska, K. Czerski, J. Rogowski, T. Paryjczak and A. Parczewski, Surf. Interface Anal., 2010, 42, 393-397. 6. L. A. McDonnel and R. M. A. Heeren, Mass Spectrom. Rev., 2006, 26, 606-643. 7. P. Hazarika, S. M. Jickells, K. Wolff and D. A. Russell, Angew. Chem. Int. Ed., 2008, 47, 1016710170. 8. P. Hazarika, S. M. Jickells, K. Wolff and D. A. Russell, Anal. Chem., 2010, 82, 9150-9154. 9. R. W. Odom, Appl. Spectrosc. Rev., 1994, 29, 67-116. 10. A. Benninghoven, Angew. Chem. Int. Ed. (English), 1994, 33, 1023-1043.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Page 16 of 16

M. J. Bailey, M. Ismail, S. Bleay, N. Bright, M. L. Elad, Y. Cohen, B. Geller, D. Everson, C. Costa, R. P. Webb, J. F. Watts and M. de Puit, Analyst, 2013, 138, 6246-6250. M. Arora and V. Sharma, International Journal, 2013. A. K. Jain, C. Yi and M. Demirkus, IEEE Trans. Pattern Anal. Mach. Intell., 2007, 29, 15-27. S. J. Hinder and J. F. Watts, Surf. Interface Anal., 2010, 42, 826-829. D. R. Ifa, N. E. Manicke, A. L. Dill and R. G. Cooks, Science, 2008, 321, 805. T. P. Forbes and E. Sisco, Analyst, 2014, 139, 2982-2985. T. P. Forbes and E. Sisco, Anal. Chem., 2014, 86, 7788-7797. N. J. Bright, T. R. Willson, D. J. Driscoll, S. M. Reddy, R. P. Webb, S. Bleay, N. I. Ward, K. J. Kirkby and M. J. Bailey, Forensic Sci. Int., 2013, 230, 81-86. A. C. van Asten, Science & Justice, 2014, 54, 170-179. R. M. Verkouteren and J. R. Verkouteren, Anal. Chem., 2009, 81, 8577-8584. E. Sisco, http://drum.lib.umd.edu/bitstream/1903/15270/1/Sisco_umd_0117E_15070.pdf, University of Maryland, College Park, MD, 2014. A. Girod, R. Ramotowski and C. Weyermann, Forensic Sci. Int., 2012, 223, 10-24. C. Szakal and T. M. Brewer, Anal. Chem., 2009, 81, 5257-5266. G. Robichaud, K. Garrard, J. Barry and D. Muddiman, J. Am. Soc. Mass. Spectrom., 2013, 24, 718-721. W. S. Rasband, http://imagej.nih.gov/ij/, National Institutes of Health, Bethesda, MD, 19972014. G. D. Martin, S. D. Hoath and I. M. Hutchings, JPCS, 2008, 105, 012001. J. E. Field, Wear, 1999, 233–235, 1-12. T. Deng, K. K. Varanasi, M. Hsu, N. Bhate, C. Keimel, J. Stein and M. Blohm, Appl. Phys. Lett., 2009, 94, 133109. R. Lieckfeldt, J. Villalaín, J.-C. Gómez-Fernández and G. Lee, Pharm. Res., 1995, 12, 16141617. J. R. Kanicky, A. F. Poniatowski, N. R. Mehta and D. O. Shah, Langmuir, 1999, 16, 172-177. J. S. Lee, I. Gilmore, M. Seah and I. Fletcher, J. Am. Soc. Mass. Spectrom., 2011, 22, 1718-1728. C. G. Enke, Anal. Chem., 1997, 69, 4885-4893. K. Tang, J. Page and R. Smith, J. Am. Soc. Mass. Spectrom., 2004, 15, 1416-1423. S. Muramoto, J. Brison and D. G. Castner, Anal. Chem., 2011, 84, 365-372. C. Roux, N. Jones, C. Lennard and M. Stoilovic, J. Forensic Sci., 2000, 45, 761-769. J. B. Calkin, Tappi, 1951, 34, 108A-112A. S. P. Saric and R. K. Schofield, Proc. R. Soc. London, Ser. A, 1946, 185, 431-447. S. Muramoto, T. P. Forbes, M. E. Staymates and G. Gillen, Analyst, 2014, 139, 2668-2673.

Supporting Information Available. This information is available free of charge via the Internet at http://pubs.acs.org/.

ACS Paragon Plus Environment