University of Huddersfield Repository Determining the chronology of deposition of natural fingermarks and inks on paper using secondary ion mass spectrometry

This study thoroughly explores the use of time-of- ﬂ ight secondary ion mass spectrometry (ToF-SIMS) for determining the deposition sequence of ﬁ ngermarks and ink on a porous paper surface. Our experimental work has demonstrated that mapping selected endogenous components present in natural ﬁ ngermarks enables the observation of friction ridges on a laser-printed surface, only when a ﬁ ngerprint is deposited over this layer of ink. Further investigations have shown limited success on ink-jet printing and ballpoint pen inks. 51 blind tests carried out on natural, latent ﬁ ngermarks on laser-printed surfaces; up to 14 th depletion with samples aged for up to 421 days have resulted in a 100% success rate. Development with ninhydrin was found to a ﬀ ect the ﬁ ngermark residue through mobilisation of ions, therefore sequencing determination was compromised; whilst iodine fuming and 1,2-indanedione developers did not. This implied that selected development methods a ﬀ ected success in ﬁ ngermark-ink deposition order determination. These results were further corroborated through inter-laboratory validation studies. The adopted protocol and extensive series of tests have therefore demonstrated the e ﬀ ectiveness and limitations of ToF-SIMS in providing chronological sequencing information of ﬁ ngermarks on questioned documents; successfully resolving this order of deposition query.


Introduction
Fingerprint evidence is routinely used in forensic investigations and has been a widely accepted form of identication evidence for over 100 years. 1 Dactyloscopy is, however, a particularly challenging eld of forensic science as ngermarks lack consistency, with a composition that depends on the individual, as well as his/her diet, stress levels and grooming regime. Further discrepancies in the appearance of friction ridges are also affected by exerted pressure on application, duration and angle of contact between the nger and a receiving surface. [2][3][4][5] Porous-surfaces in particular affect surface residue deposits because of their inhomogeneous composition and ability to rapidly absorb components into the substrate. 6 The development treatment necessary to visualise latent prints and analyse ridge detail, which is selected on the basis of factors including substrate and ngermark age, also affects the components within the ngermark. This entire spectrum of factors is further complicated by the fact that dactyloscopy heavily relies on the skill of the ngerprint examiner to correctly discern the ridge features in the ngermark evidence. 7,8 When investigating cases of fraud or counterfeiting, besides recovering the ngermark ridge pattern on a handled document, it is necessary to establish whether the ngerprint has been deposited before or aer the surface was written or printed over with compromising ink material. This would allow the forensic document investigator to establish the chronology of a ngerprint on a surface and therefore identify whether an individual is actually associated with the ink-related evidence or simply handled a blank sheet of paper. If it was possible to tell whether a document was handled aer inked evidence was deposited onto the surface, a forensic investigator would be able to avoid claims of a suspect handling a pre-printed/signed document: if touched aer ink deposition, then the suspect/ donor would have handled a pre-printed document. This problem is one of the major challenges in ngermarks associated with document examination, as existing development techniques do not provide any information on chronology or depth of penetration of ngerprints into porous surfaces, making it impossible to determine the order of deposition of ngerprints and inks.
An increasing number of established and emerging analytical characterisation technologies are being implemented to study and retrieve information from ngerprint evidence. These include infrared and chromatographic methods, as well as desorption/ionisation-mass spectrometry techniques, which all have particular features that at present allow or are still being explored to investigate ngermarks. Composition, alteration, presence of contaminants, imaging of ridge detail and increasing the level of understanding of interactions experienced between ngermarks and surfaces, environments, development agents and other processes are all aspects being examined for ngermark analysis, which have been described elsewhere. 4,[9][10][11][12][13][14][15] In cases of fraud or counterfeiting, the examination of questioned documents and ngermark components on these porous surfaces necessitates an investigative method with good spatial resolution, highly surface-sensitive capabilities and minimal sample preparation owing to the complexity of the chronological information required. A recent study into sequencing of ngermarks and inks by Fieldhouse et al. 3 reports a good degree of efficiency by electrostatic detection (EDSA) on laser printed inks and ngermarks on paper from surface indentations. In our research we explore the application of timeof-ight secondary ion mass spectrometry (ToF-SIMS) as a suitable technique to establish the chronological sequence of printed inks and ngermarks on paper. This instrumental method targets the chemical composition of a sample, providing a combination of elemental and molecular fragment signatures which can be directly linked to individual constituents within a sample complex. It is therefore much more sample-specic and does not purely rely on physical surface indentations. Although other methods such as matrix-assisted laser desorption ionization (MALDI) and desorption electrospray ionization (DESI) are imaging mass spectrometry techniques with proven success in ngermark analysis, 10,13 neither has achieved the spatial resolution or surface sensitivity of ToF-SIMS. 16 SIMS uses a nely focussed primary ion beam for the 'so'-ionization of a surface through desorption of secondary ions. When coupled with a time-of-ight mass spectrometer, this instrumental method allows for rapid mass-spectral analysis over a wide mass range with good mass resolution. It is also capable of producing high resolution chemical images by rastering the ion beam across the sample surface. 17 ToF-SIMS has excellent lateral resolution capabilities, is particularly surface sensitive, and has recently demonstrated applications in the forensic analysis of ngerprints, 18-21 papers and paper coatings, 22 and in determining the deposition order of ngerprints and ball point inks, however with limited success. 18,23 The present work is a development and expansion of our proof of concept study, 24 and investigates the use of this technique for chronological determination of inks and natural ngerprints deposited on paper. Rigorous testing has been carried out on a variety of printing papers, overlaid with latent ngerprints acquired from various donors on solid, text and dra laser-printed black and coloured inks. Investigations have explored both latent and developed ngermarks, which were examined in various deposition sequences and at staggered times, with extended ageing periods of over 1 year. Our experimental work has demonstrated that mapping various endogenous components within the ngermark enables the observation of friction ridges from ngermarks on the ink surface when a ngerprint is deposited above a layer of ink. Developers were found to affect the chemical components and resulting ion signal from a ngermark residue to various extents, with certain development methods compromising the ability to determine the order of deposition of these ngerprints on inks. This work therefore also advises the order of which latent ngermarks should be developed on this type of evidence to effectively target the deposition order query.

Materials
Four types of white office 80 g m À2 printing paper were used in these experiments. The inks and printing techniques are detailed in Table 1, with a total of 156 known samples and 51 blind-test samples analysed. Both natural and sebaceousloaded samples were initially investigated. Natural samples were, however, considered more representative of the types of marks encountered in eld applications and were therefore the principal focus of this study. 25 Donors were asked not to wash their hands or apply cosmetics for at least 30 minutes prior to ngermark deposition. Fingertips were also rubbed together to improve distribution of sample deposit from both le and right hands. 25 Fingerprint sampling usually involved acquiring rst, second and h depletions of all ngers per hand; though in ball point pen tests only rst depletions were investigated. Some ngermarks were applied over a layer of ink (ngermark over ink: FOI), while other marks were deposited on plain paper and subsequently printed over with ink (ngermark under ink: FUI). Printing over the ngermarks in FUI samples was executed within 30 minutes of ngermark deposition, unless the ngermark was intentionally le to age before ink deposition. The analysis day in FUI prints therefore generally determined the age of both ngerprint and ink. Ball point pen ink experiments necessitated manual application of ink, with both solid and discrete lines of ball point pen ink marked onto the standard 80 g m À2 paper. Ink was applied uniformly in all samples, although in solid ink examples this implied that certain areas were inked over more than once. The ink in these ball point pen ink experiments was allowed to dry for four days and 1 month in two separate consecutive studies before ngermarks were deposited onto the inked surface in FOI studies. Fresh and aged ink and ngermark sequences were analysed for all samples. Samples were aged together in a ling cabinet located in a normal office environment, and no particular care taken to avoid inter-sample contact during ageing, packaging or transportation.

Development
A number of samples were developed with ninhydrin and/or iodine for the development trials as detailed in Table 1. Ninhydrin solution was prepared by mixing 5 g of ninhydrin (WA UK, SIRCHIE®) with 45 mL of ethanol, 5 mL acetic acid and 2 mL ethyl acetate (>99% purity, Sigma-Aldrich [UK]); to which 1 litre of 3M™ Novec HFE 7100 was added. This solution was either sprayed or swabbed onto the latent ngermarks with a cotton bud onto the surface, and development was accelerated by placing the samples in an oven with a beaker of water to avoid total desiccation, at 60 C for 20 minutes. Documents developed in iodine (>99.8% Sigma-Aldrich) were suspended in an airtight chamber with 1 g of iodine crystals dispersed in a Petri dish at the base of the chamber. The latent ngermarks were fumed naturally without any additional application of heat or air ow for 45 minutes.

Instrumental analysis
Initial examination of materials was executed using a Zeiss Supra 35VP eld emission scanning electron microscope (FEG-SEM), coupled with an Oxford Instruments Inca energy dispersive X-ray analyser (EDX). Samples were coated with approximately 11 nm of gold to increase surface conductivity and enhance contrast. Samples were mounted on aluminium stubs using a carbon-loaded pressure sensitive adhesive. Imaging and elemental composition assessment was conducted under high vacuum at 20 kV.
ToF-SIMS was used to individually analyse paper, printed ink and ngermarks deposited on silicon wafer as well as the ngermark-ink FOI/FUI sequencing samples prepared (Table 1). Secondary ion spectra and images were obtained using a Kore Technology Ltd. SurfaceSeer ToF-SIMS spectrometer. Two types of primary ion sources were used: an 114 In + and a 69 Ga + FEI liquid metal ion source. The spectrometer was operated in a pulsed, positive mode with 25 kV applied voltage and 1 mA current. Secondary ions were analysed in a reectron time of ight mass spectrometer and detected with a dual microchannel plate assembly. A ood gun with low energy electrons was operated simultaneously to compensate for surface charging in insulating samples. Spatial resolution was $1 mm diameter, and ight times were recorded with a 0.5 ns time-to-digital converter. Mass spectra acquisition time was set at 60 seconds at a magnication of Â100, and acquired spectra m/z range was between 0-1300 Daltons (Da). Well-dened cation peaks on these spectra between 10-60 Da were generally used to select mapping regions. 512 Â 512 pixel chemical maps were acquired over 40-50 minutes at 20 cycles per pixel for 4/5 frames, with a minimum of three areas scanned per sample. Acquired maps measured $750 Â 750 mm, and were processed where necessary to improve contrast using imaging soware. The images themselves were visually assessed to establish whether ngermark traces could be observed over an inked area. All ToF-SIMS spectra were calibrated before analysis.

Blind tests
Blind test samples were acquired from one female and two male donors on both solid, dra and text printing in a similar procedure as outlined in 2.1 above. A range of 1 st to 14 th depletions were prepared, with ink deposited in various ink and ngermark sequences on different days. These were analysed at varied times between 1 to 421 days aer ngermark deposition. Samples were cut up and arbitrarily numbered in polyethylene sample bags, and were submitted for ToF-SIMS analysis without providing any additional sample information. Quasi-operational blind tests were also investigated, which replicated an in-eld scenario much as possible. Here a text, laser-printed document was rst developed with iodine (IOD) with ngermark boundaries marked out post development to establish ngerprint location. Samples were then documented, packaged and analysed using ToF-SIMS to determine ngermark chronology (FOI/FUI). Each identied ngerprint was further developed by spraying with ninhydrin (NIN) solution and le in an oven for 45 minutes to visually attest whether a 'standard' developmental procedure was possible aer the ngermark sequencing determination protocol.

Inter-laboratory validation experiments
Six known and six unknown, latent and developed samples acquired from 3 donors were analysed in an inter-laboratory validation study executed in collaboration with the University of Surrey and the Netherlands Forensic Institute (NFI). Samples were prepared at the NFI, with ngermarks from thumb, middle and ring ngers deposited 1 h aer donors had washed their hands with soap and water. 80 g m À2 paper (Fastprint extra) was used in these experiments, and lines of ink were printed over/ under the ngerprint-paper setup with a standard 1200 dpi laser printer. All blind test samples were developed, with two examples provided for NIN development, and four for 1,2indanedione (IND).
Development was carried out using standard NFI stock solutions. NIN was prepared by mixing 25 g of ninhydrin (BVDA, Haarlem) and 225 mL of ethanol; with 10 mL ethyl acetate and 25 mL acetic acid (>99% purity, Sigma-Aldrich [Zwijndrecht, the Netherlands]) subsequently added to the stock solution. 1 litre of HFE 7100 (3M St. Paul, USA) was then mixed with 52 mL of NIN until a homogeneous solution was formed. A similar stock solution of IND was prepared by combining 10 g of indanedione (BVDA, Haarlem) and 60 mL of ethyl acetate, 10 mL of acetic acid and 900 mL HFE 7100. A working solution was prepared by mixing 8 mL ZnCl 2 (>99% purity, Sigma-Aldrich, prepared with ethanol, ethyl acetate and HFE 7100) with 100 mL indanedione. Both NIN and IND development was executed via immersion, with samples dipped in solutions until wet. These were then air dried for a few minutes and heated for 10 minutes at 80 C and 20 minutes at 100 C for NIN-and IND-treated samples respectively.
ToF-SIMS investigations for the inter-laboratory experiments were carried out at the University of Surrey 68-72 days aer the samples were developed, and no particular care was taken in the packaging or transportation of the samples. Analysis was executed on an IONTOF GmbH (Münster, Germany) ToF-SIMS 5 spectrometer, employing a 25 keV Bi 3+ primary ion beam delivering 0.35 pA of current. Images were acquired at 128 Â 128 resolution in the MacroRaster mode over a 5 Â 5 mm area. Image data were acquired using 256 cycles per pixel point with 1 scan per pixel and a cycle time of 100 ms.

Laser printed ink and photocopying
Laser printers and photocopiers engage electrostatic liing of toner particles onto a charged, rotating drum. When the drum comes into contact with a sheet of paper, an image is printed onto the paper surface through the application of direct contact (pressure) and heat, which fuses the ink to the paper. SEM images of laser, dra-printed and photocopied paper all demonstrated this fused toner, which formed a discrete layer of ink with several surface irregularities and voids that completely covered the underlying interwoven mass of paper bres. In dra-printed ink at 300 dpi, ink covered less than half the paper surface in a checkerboard pattern at a relatively small scale, yet still masked the paper bres ( Fig. 1a and b). EDX analysis of these toner-based inks showed their fused-silica, organic-based nature. Additional elemental colorants, opaciers and/or additives were also identied, with iron consistently detected in black inks, copper and titanium in cyan, and titanium with traces of sulphur in both magenta and yellow printed toners. Sodium and calcium peaks were present in all EDX spectra, identied as well-known paper additives sodium chloride and calcium carbonate. 24,22 ToF-SIMS data of samples of ngermarks over and under laser-printed ink (FOI/FUI) acquired in this study corroborated the preliminary results published in our proof of concept study. 24 Ions in the lower mass range were consistently more prominent in all mass spectra owing to analyte composition and ionising properties of the primary ion beam. 22 Mapping of these ions presented clear secondary ion maps with improved ion contrast and material visibility. These lower m/z cations were therefore identied as the most useful species to differentiate between ngermark, paper and ink constituents. Mapped ions included the paper derivatives calcium (Ca 2+ ) and to a lesser extent quaternary nitrogen (C 3 H 8 N + ) at 40.078 and 58.102 Daltons respectively. 22 Silicon (Si + ) and SiOH + at m/z 28.086 and 45.093 Da were identied as the embedding silicon components in laser-printed inks. The organic nature of the samples implied that various hydrocarbon fragments at, m/z 15.034, 27.045, 29.061, 41.071 and 55.098 Da, for example, were identied in all materials being examined (paper, ink, ngerprint). The most prominent hydrocarbon peak generally appeared at C 3 H 5 + (41.071 Da). Relative heights of each component to the surface also affected constituent ion counts. This was as a result of the surface sensitivity of the instrument and subsequent limitations in height resolution over a specic area, with larger distances increasing the time taken for secondary ions to reach the time of ight mass analyser, resulting in contrast differences and identication of laser-printed ink and paper boundaries (Fig. 2).
Sodium (Na + ) and potassium (K + ) secondary ions of 22.990 Da and 39.098 Da, consistently corresponded to the ngermark deposit. Na + ions typically dominated as a base peak in ngermark spectra, though these salts sometimes appeared as accretions in maps around pore areas from certain donors. Replicated samples analysed from donors indicated that certain individuals consistently deposited poorer ngermarks, though certain other deposits appeared to be affected by cold weather. This was demonstrated by analysing average ion counts and image contrast in SIMS maps. When a mass spectrum was acquired at a non-pore/Na + accretion area in poorer donors, this was reected through comparatively lower peak heights in the mass spectra at these particular areas. Sodium ions also predominated in all examined paper substrates, especially in the Office Depot paper batches. The similarity between Na + and C x H y + mass fragments in paper and ngerprint component fractions implied that locating a ngermark deposited on paper from these ions was difficult; whether or not the paper was subsequently printed over with ink (FUI). Occasionally, the sodium ion counts in the ngermark were notably higher than those in the paper, implying that ridges could be identied, but these concentrations varied widely between paper and ngerprint donors, as well as in depletion, day and time of deposition. Sodium ions were additionally observed as base peaks in mass spectra acquired from cyan and yellow toners, as well as being fairly prominent in magenta mass spectra (Fig. 3). K + was also abundant in all coloured laser inks, with ion counts in secondary ion maps approximating those in FOI ngermarks, therefore making it difficult to discern ridges in potassium maps. The Na + peak and respective ion counts on mapping were consistently higher in samples where the ngerprint was deposited over laser printed inks, even in poor donor samples. This implied that determining the ngermark chronology over these inks was still possible (Fig. 3).
Increasing the donor pool, experimenting with different printing styles (photocopies, text and dra printing), and testing aged inks and ngerprints were used to establish relative sensitivity and potential limitations of this ToF-SIMS technique. Following a broad, varied sample study of ngermarks on/under laser-printed ink, factors appearing to affect interpretation were generally a combination of ngerprint donor and printed ink characteristics. Compromising factors  included lower ion counts resulting from decreased quantities of natural secretions in collected ngermarks, either as a result of them being signicant depletions in a depletion series, and/ or because the deposit was acquired from a poor donor, and/or on a cold day. Furthermore, as the paper provided similar ion fractions to the distinguishing secondary ions in the ngermarks, identifying ridges in FOI samples was found to depend on the amount of ink deposited on the paper surface, with text and dra-printed ink making data analysis more difficult. Fingerprint age, ink age, as well as ink colour did not appear to particularly affect interpretation, as was demonstrated through the successful identication of FOI/FUI sequences where ngermarks and inks had been aged for over a year. This indicated that ion mobility through a stratum of ink with time did not compromise FOI/FUI determination in this type of printed ink. Nevertheless, aged ngermarks on dra-printed 300 dpi inks were more difficult to identify, especially aer the samples had been aged for a substantial time period (S60 Fig. 8). Furthermore, although ngermarks over coloured laser-printed inks could not be distinguished in respective K + maps; it was always possible to ascertain an FOI ngermark from the elevated ion counts in Na + maps.

Inkjet printing
Inkjet printing is an alternative deposition technique used for liquid phase inks that consist of an organic colorant solute, dissolved or dispersed in a solvent. A xed quantity of ink is ejected from the printer nozzle, where it spreads under momentum and surface tension along a surface (Fig. 6). Studies have shown that drop spread is dependant on the viscosity of the polymer, the height of the nozzle, as well as the structure of the receiving surface. [26][27][28] This wetting-penetration effect was reected in results obtained, where SEM-EDX analysis generally did not allow for the discernment of printed areas on paper as there was no particular difference in height, composition, or in the appearance of paper bres, presence of voids and crystallinity of ller aggregates within the paper (Fig. 4). Interface edges of coloured ink and paper in effect required light marking with a ball point pen to identify printed and non-printed paper. A variation in colour was only observed in black ink-jet printed area; though EDX results did not identify any particular difference between this black ink and paper components. Compositional analysis in all samples merely indicated that calcium was dominant in the paper, whether or not it was inked over, and that the inks were organic. This organic nature was conrmed by ToF-SIMS analysis, with most ink fragments observed characteristically denoting long-chain hydrocarbon fractions, with similar major mass spectral peaks obtained from cyan, magenta and yellow inks.
Sequencing trials of ngermarks over and under printed inkjet inks were executed on fresh ('wet') and older ('dry') samples. A slight difference in the strata height and composition of the paper and inked areas was observed in the maps, allowing for identication of paper-ink interfaces, and of printed areas. Initial ion maps analysed from fresh FOI/FUI samples with freshly printed inks gave poor results, mainly owing to outgassing effects experienced on UHV conditions, which resulted in poor ion counts. Chemical images designated swollen printed paper bres, indicative of the absorption of these 'wet' inks into the paper cellular structure, which expanded as a result of this ink application. Fingermark ridges were near-invisible in this fresh-sequencing setup, although an increased amount of salts were observed in a few FOI samples; typically those obtained from the thumb and index ngers from a better donor. This was not a consistent result and therefore implied the methodology was unsuitable for ridge identication on freshly printed ink-jet paper surfaces.
Observation of aged samples showed that printed paper bres had dried and returned to their original shape and structure. Fibre edges in the black printed areas were somewhat less dened than those in the coloured samples, suggesting that this black ink was deposited as a thicker layer. This could also account for the backscatter greyscale difference between the paper and black ink observed in the SEM images (Fig. 4). Although ion counts in aged samples were improved, sequencing determination of FOI and FUI ngermarks was still inconclusive. In black inkjet-ngermark examples, no ridges were observed in the FOI/FUI samples examined, making it impossible to ascertain whether a ngermark was at all present in a mapped location. In the coloured ink samples, ridges were observed in the majority of the sodium ion maps in the inkedover areas of the paper. These ridges, however, were not limited to FOI sequences but were also observed in the FUI samples as in Fig. 5c. Fig. 6 schematically demonstrates the differences between the two printing techniques. Toners in laser-jet printing effectively fuse onto a surface in a 'dry' process, forming the discrete layer of ink observed in SEM images and SIMS maps. Inkjet ink is conversely deposited in solution, therefore penetrates into the paper, colouring the paper bres and drying level to the surface (Fig. 6a). When a ngerprint is deposited onto a printed document, a mixture of ngermark components are brought into direct contact with the document surface. The isolating stratum of fused Si-based toner in laser-jet ink is assumed to decrease the porosity of the paper surface, implying that the ngermark components rest on top of this relatively 'impermeable barrier' formed by the ink layer. The altered porosity also explains why migration of components is not observed aer samples are aged for signicant periods of time, implying that ngerprint components are not allowed to interact with the porous paper surface and therefore remain on the ink surface and are identiable by ToF-SIMS. Conversely, inkjet ink does not form this isolating layer but is deposited by the printer as a 'wet' solution. This is demonstrated in the visible paper bres of the inked-over areas, even aer the ink has dried ( Fig. 4 and 5).

Ballpoint pens
Mass spectroscopic techniques including ToF-SIMS have previously been used to explore the composition of ballpoint pen inks, their drying and ageing characteristics; as well as the deposition order of ngermarks over/under these inks on paper. 11,23,[29][30][31] Work carried out in this study corroborated published results, with ToF-SIMS compositional analysis indicating that inks from the separate BIC pens examined in this study were both similar, organic-based complexes. Basic crystal violet dye [C + (Me) 6  Paper bres were still visible through the ink layer, especially when a single line of ink was drawn on the surface. Comparatively low sodium ion levels and the presence of basic violet peaks in the ink nonetheless sufficiently allowed for discrimination between paper and ink, and for the visualisation of a paper-ink interface. This also facilitated friction ridge discernment, generally from elevated Na + levels in FOI samples. Ridge detail was easier to ascertain on 'solid' inks, rather than those over inked lines, owing to the elevated ink coverage. Paper bres were, however, still mostly visible with this ink type.
Our FOI/FUI order of deposition experiments also supported previous deposition order experiments with ball point pen inks. 23 The investigations carried out in this study however looked at different mass fragments: Na + in preference to the ink peak intensities at m/z 372 and 358 Da. This was because these ions were endogenous to the ngerprints and were also the focus of our study, as well as being the base peaks in the  This journal is © The Royal Society of Chemistry 2014 Analyst majority of the mass spectra. The smaller eld of view provided by our instrument also meant that smaller, ink-dominant areas could be mapped. Results consistently presented elevated Na + levels in FOI samples, but were found to be similarly inconclusive in FUI samples, with some Na + accretions observed in FUI examples. Variation in ion levels was attributed to the variability observed between samples, with inuencing factors including quantity of material in the ngermark residue itself as well as ink thickness in the manually drawn ink lines on a paper surface. This in turn affected the overall thickness of the isolating layer of ink and its ability to mask sodium ions in the paper. Improved results were acquired with thumb and index prints, where ridge detail was consistently observed in FOI examples ( Fig. 7a and b). Line-drawn examples were more difficult to interpret. The protocol, therefore, although providing an improved result to inkjet FOI/FUI sequencing, still gave an inconclusive result.

Blind tests
A comprehensive blind test series that looked at 51 latent ngermark samples deposited over/under laser-printed ink resulted in a success rate of 100%. This corroborated our previous work and showed that this ToF-SIMS methodology is suitable for determining the chronology of a ngermark over/under laser-printed ink. Results tallied with observations given for known samples, where Na + ions were visibly identiable over the ink, only in FOI samples. These ions followed ngerprint ridge detail and demonstrated up to third level features.
Observations were generally corroborated in K + maps, with examples of approximately 6 month old 5 th depletion FOI sample and a 1 st depletion FUI over/under text provided in Fig. 2. A number of examples were provided to test those variables identied as limiting factors of our methodology, as identied previously (3.1). Aged samples, quantity (coverage) of ink on the surface as well as depletion prints were analysed, with the oldest, poorest coverage and 14 th depletion examples displayed in Fig. 8. S50 shows that ridge detail was still visible in even (4 th ) depletion samples le to age for over one year, only when the ngermark overlies the ink, as can be seen when compared to the FUI equivalent in S21. Furthermore, mobile ions did not appear to migrate through laser-printed ink from the paper, therefore quantity of sodium ions on an ink surface or a ngermark were not affected with time and interpretation was not compromised, even aer ageing for over 421 days (S21). The 14 th depletion over text (T117) was the maximum depletion sample tested, which still consistently showed salt accretions from pore areas over the lettered ink. Aged, depletion ngermarks over/under dra-printed ink looked at a combination of factors negatively affecting ngermark quality for interpretation. Dra-printed ink in particular reduced the quantity of isolating layer of ink on a much smaller scale. S60, a 3 rd depletion aged for just over a year, displayed mere traces of salts over the inked-over areas of this 300 dpi ink. Na + ions nevertheless visibly corresponded to equivalent K + accretions that followed faint ridge detail over ink in respective maps, therefore strongly suggesting a FOI. Amounts visible, however, were minimal, therefore interpretation, especially when purely based on visual examination, could be queried and attributed to possible contamination.

Development
Ninhydrin (NIN) development appeared to compromise our ability to discern ngermark components over ink in FOI samples with the certainty that was observed in the latent ngerprint equivalents. Contact application of NIN with a cotton swab also appeared to have a diluting, and a seemingly smearing effect on the surface residue, making maps difficult to interpret, ridges difficult to identify and results generally inconclusive. A good donor provided better ion contrast in FOI samples with visibly elevated sodium ion peaks over ink, though several areas needed to be mapped as images were less denitive, especially when text samples and/or h depletions over ink were examined. Furthermore, in certain FUIs Na + ions appeared to migrate over the ink, especially around ink-paper boundaries (Fig. 9). This was assumed to be a result of mobilisation of salt ions through contact application of the solventbased (wet) developer with a cotton swab over the surface. The presence of Ca 2+ ions, which were characteristically limited to paper in non-developed samples, over the ink in both FOI and FUI maps, further supported this material migration/smearing hypothesis, besides designating possible contamination effects. Samples developed through non-contact spray application of NIN still had a diluting effect, though secondary ion maps had improved contrast levels when compared to swab application. Na + ions, however, did not follow friction ridge detail and Ca 2+ accretions were still present over the ink in both FOI and FUI samples, albeit in lesser quantities. Furthermore, faint Na + accumulations indicative of salt migration/smearing were again observed over laser printed ink in certain FUI examples. This subsequently compromised data interpretation as poor FOIs acquired from either a depletion sample and/or a poor donor produced maps with similar levels of ion contrast. NIN  development therefore signicantly affected the ability to discriminate between a ngermark overlying or underlying the ink. Iodine (IOD) was explored as a non-contact alternative, solvent-free developer. Its vapours are assumed to be physically absorbed into the fatty acids in the ngermarks in a transient, reversible process. 32 This made it particularly desirable for developing the ngermarks in this study owing to a potentially less impinging and detrimental effect on ngermark constituents when compared to NIN development.
Initial maps generated from samples developed with IOD presented strong Na + and K + secondary ion signals over the ink in FOI samples. This observation was consistent, even with 5 th depletions. Results were corroborated through a six-donor sample study that looked at split prints, where one half of print was le undeveloped and the other half fumed with iodine. The simultaneous ToF-SIMS analysis and mapping of these samples allowed for effective comparison of data, where mass spectra were both stacked and overlaid for comparative purposes, and maps compared as shown in Fig. 10. There were no prominent differences in relative peak heights between non-fumed and fumed ngerprints. Minor peak-height variations observed were assumed to be a result of inter-sample differences, owing to the presence of more ridges/material/salts in one half of the print than in the other, within a particular area, despite comparisons being made from splits of the same ngermark. Fig. 10 shows the total, Na + , K + and Ca 2+ secondary ion maps, and compares undeveloped vs. fumed, FOI and FUI maps; from a known, relatively poor donor. The total and Ca 2+ maps indicate the location of the printed text, whereas Na + and K + maps show the difference in abundance of these ions over the ink in the FOI vs. FUI samples. FOI results clearly show that the ngermark overlies the ink, whether the sample was IOD fumed or le undeveloped. The latent FUI sample also presents an elevated Na + signature along ngermark ridges on the paper beneath the ink, indicating that in this particular example, it was also possible to identify the FUI on the paper. These observations were corroborated in maps acquired from the other 5 donors.
The nal blind sample test was executed to test the viability of the experimental protocol through analysis of a laser-printed document with ngermarks possibly deposited, before or aer the ink present. This therefore made the test quasi-operational, it being the scenario expected with samples acquired from forensic investigation cases. Five out of the six ngermarks deposited on the laser-printed text from this quasi-operational study were located through iodine development, and all samples were correctly identied as FOIs following ToF-SIMS analysis. Fingerprint samples were further developed with NIN, and developed ngermarks showed excellent contrast and ridge detail, therefore indicating that the methodology for FOI/FUI determination (IOD fuming and ToF-SIMS analysis) did not affect further (amino acid) development for ridge-pattern identication.

Inter-laboratory validation
Imaging latent ngermarks deposited over laser-printed ink on paper in the IONTOF GmbH ToF-SIMS in the Surface Analysis Laboratory at the University of Surrey provided similar results to those described above (3.1). An abundance of hydrocarbon fragments as well as Na + and K + ions at m/z 22.990 and 39.098 respectively were observed corresponding to ridge detail in secondary ion maps in FOI samples. The heavier Bi 3+ primary ion beam also resulted in increased levels of data from ngermark ion fragments in the higher mass range of the spectrum, such as those at m/z 65.973, 87.971 and 284.326 Da (Fig. 11 and  12). The IONTOF GmbH ToF-SIMS was additionally able to scan a wider area of the sample, with a maximum eld of view of 6000 Â 6000 mm 2 obtained. This produced images with increased amounts of ridge detail, but also implied that the Na + ions could not be used to distinguish between FOI/FUI samples because of their abundance in the paper substrate which consequently affected data analysis. Furthermore, hydrocarbon data present in the ngermark was also not as identiable over the ink lines ( Fig. 11 and 12). K + ions were, however, visibly elevated in ridge detail in FOI maps, and were therefore consistently used to distinguish the order of deposition of ngermarks on/under ink lines with this instrument.
Sample development by NIN immersion again compromised the ability to identify whether the sample was an FOI/FUI, with K + and other ion fragments appearing over the ink, even in FUI samples. The 'NIN effect' described previously was corroborated by the results obtained in this inter-laboratory validation study, from both known and blind-test samples analysed (Fig. 11).
1,2-Indanedione (IND) was preferentially selected owing to its frequency of use in the NFI as a developer for ngermarks on paper. Development with this reagent gave excellent results, although one sample demonstrated that the quality of donor residue affected interpretation condence. K + ion maps (m/z 39.098 Da) generally demonstrated clear ridge detail over the ink only when the ngermark had been deposited over this inked layer. Potassium ions could not be identied on the paper substrate using this ToF-SIMS, or over the inked line when an FUI sample was examined. These observations were corroborated in the blind tests studies as demonstrated in the example provided in Fig. 12 below.

Conclusions
This paper demonstrates that this ToF-SIMS protocol can be used to identify the chronological sequence of natural ngermarks over laser printed ink on paper, and is therefore suitable to verify whether a suspect's ngermark is actually associated with this type of printed evidence. Selected endogenous markers within the ngermark, particularly Na + and K + ions in secondary ion maps jointly show elevated ion counts and ridge detail over the laser-printed ink, only when a ngermark overlies an ink line. Our extensive sampling trials and blind tests indicate that several factors affect quality and subsequent interpretation of results, with reliability generally depending on (i) Na + and K + ion count of in a ngermark, (ii) amount of ink present on the surface and (iii) ink type. A depletion print acquired from a poor donor, for example, denotes a lower ion count, making FOI/FUI sequencing identication more complicated, but still possible, as long as the ngermark was deposited on laser-printed ink. Text and dra laser-printed ink reduced the quantity of isolating layer of ink, implying a smaller area for FOI/FUI determination; again making interpretation more difficult but possible. Ink colour and sample age did not appear to affect acquired data on laser-printed ink. Successful ngermark chronology determination with ball point pen ink depended on the type and amount of ball point pen ink deposited on the surface, with thinly applied and solvent-based inks drying level to the paper surface and therefore compromising sequencing determination. This was because successful identication of Na + and K + markers in a ngermark over the ink necessitated a relatively impermeable ink barrier over the paper, which restricted ion mobility and therefore enabled ridge identication. ToF-SIMS was therefore unsuitable for the identication of the deposition order of a ngerprint on inkjet ink. Development with ninhydrin to locate the latent ngermark prior to ToF-SIMS analysis also affected the reliability of the technique, but the successful results obtained with iodine fuming and indanedione indicated their operational effectiveness, as was demonstrated in the quasi-operational blind test studies and the inter-laboratory validation assessment.

Analyst
This journal is © The Royal Society of Chemistry 2014 Analyst Paper