Electron Microscopy: SEM, TEM and EDS

A scanning electron microscope focuses a finely collimated electron beam onto the specimen surface using a series of electromagnetic condenser and objective lenses, then deflects the beam in a raster pattern to scan point by point across the specimen. At each scan position, the primary electron beam interacts with the specimen material and generates multiple signal types simultaneously. Two are most commonly used in forensic work: secondary electrons and back-scattered electrons.

Secondary electrons (SE). Secondary electrons are low-energy electrons (typically less than 50 eV) ejected from the outermost 1-10 nm of the specimen surface by inelastic collisions with the primary beam. Because they are low energy, they are strongly influenced by surface topography: a surface asperity that points toward the SE detector ejects more SEs than a flat surface, which ejects more than a surface facing away from the detector. The SE image is therefore a topographic image: it shows the three-dimensional surface relief of the specimen with strong contrast from surface texture, cracks, edges, and elevated features. SE imaging is the standard mode for visualising particle morphology, fracture surfaces, fibre cross-section textures, and tool-mark surface profiles.

Back-scattered electrons (BSE). Back-scattered electrons are primary beam electrons that have been elastically or quasi-elastically scattered backward through a large angle by atomic nuclei in the specimen. The probability of back-scattering is proportional to the atomic number Z of the scattering element: heavy elements (high Z) scatter more strongly and appear brighter in the BSE image; light elements (low Z) appear darker. BSE imaging is therefore a Z-contrast or compositional image. On a cross-section of an automotive paint sample with multiple layers of different chemistry, the BSE image differentiates a titanium-dioxide-pigmented white topcoat (Ti, Z = 22, bright) from a carbon-black-pigmented black basecoat (C, Z = 6, very dark) from an iron-oxide-pigmented red primer (Fe, Z = 26, intermediate brightness). This visual chemical mapping is not quantitative (EDS provides quantification) but allows the analyst to orient the examination and identify which layers require elemental analysis.

The interaction volume. The primary electron beam does not simply bounce off the surface; it penetrates into the specimen and generates a pear-shaped interaction volume. At 20 kV accelerating voltage in a medium-Z material (silicon or iron), the interaction volume extends roughly 1-2 micrometres below the surface and is approximately 1-2 micrometres in diameter. SE imaging resolves features at the beam spot diameter (which can be sub-nanometre in the best instruments), but X-ray signals used in EDS are generated from the full interaction volume depth, giving an EDS lateral resolution of approximately 1-2 micrometres at 20 kV. Reducing the accelerating voltage to 5-10 kV shrinks the interaction volume and improves EDS spatial resolution at the cost of lower X-ray intensity and excitation of only low-energy X-ray lines.

Specimen preparation. Unlike optical microscopy, SEM requires the specimen to be in a vacuum environment (because electrons are scattered by air molecules). Non-conductive specimens (organic fibres, biological tissue, polymer paint layers) must be made electrically conductive to prevent charge accumulation on the surface that would deflect the primary beam and distort the image. The standard preparation is to coat the specimen with a thin (2-10 nm) layer of a conductive material, typically gold-palladium alloy deposited by sputter coating. For GSR examination, the aluminium stubs on which particles were collected by tape lift are usually conductive enough without further coating. For EDS elemental analysis, carbon coating rather than gold-palladium is preferred when the element of interest includes gold, silver, or palladium, to avoid interference from the coating material.

When the primary electron beam excites an inner-shell electron out of an atom (ionisation), a vacancy is created in the inner shell. An outer-shell electron drops down to fill the vacancy, releasing energy as an X-ray photon. The energy of this photon equals the difference between the binding energies of the two shells, and this energy difference is unique for each element (Moseley's law). By measuring the energies of X-ray photons emitted from the specimen, the EDS detector simultaneously identifies all elements present.

EDS detector architecture. The standard EDS detector is a silicon drift detector (SDD) cooled to approximately -20 to -40 degrees Celsius by a Peltier element. Incident X-ray photons create electron-hole pairs in the silicon; the number of pairs is proportional to the photon energy, and each photon event is counted as a pulse in the appropriate energy channel. A modern SDD detector can process 50,000 to 500,000 counts per second, collecting a usable spectrum in 30-60 seconds at typical forensic beam conditions. The energy resolution of an SDD is approximately 125-135 eV (full-width at half-maximum at the Mn K-alpha line at 5.89 keV), which is sufficient to separate the K-alpha lines of adjacent elements in most of the periodic table, though some overlaps (Ti and Ba, As and Pb) require care.

EDS spectrum interpretation. A forensic EDS spectrum displays intensity (counts) versus X-ray energy (keV). Each element present in the interaction volume appears as a peak at its characteristic energy. The analyst identifies elements from peak positions and confirms them by checking that all expected series lines for that element (K-alpha, K-beta; L-alpha, L-beta; M series for heavy elements) are present in the expected intensity ratios. The SDD detector's output is quantified by stoichiometric deconvolution software (ZAF or phi-rho-Z correction methods) to give elemental concentrations in weight percent or atomic percent.

Mapping mode. In EDS mapping mode, the beam is scanned across the specimen in the raster pattern and at each point the EDS spectrum is recorded. From this data cube (X position, Y position, and full spectrum at each point), the analyst extracts element-specific maps showing the spatial distribution of each element across the scanned area. For automotive paint cross-sections, a BSE image combined with EDS maps for carbon, oxygen, titanium, iron, and barium allows the analyst to assign chemical compositions to each layer and compare the composition to the RCMP PDQ database entry for the suspected vehicle make and model.

Gunshot residue (GSR) consists of the fine particles expelled from a firearm during discharge, drawn from the partially combusted primer, propellant, and bullet/jacket material. These particles are ejected in a plume from the barrel muzzle, the cylinder gap (in revolvers), and various gaps between the breech face, the extractor, and the frame. They deposit on the hands, forearms, face, and clothing of the shooter, and to a lesser extent on bystanders and surfaces in the vicinity.

The three-element GSR criterion (ASTM E1588). The standard ASTM E1588 test method for GSR examination by SEM-EDS specifies that a particle is confirmed GSR if it contains all three elements lead (Pb), antimony (Sb), and barium (Ba) simultaneously in a single morphologically rounded particle (typically 0.5-10 micrometres in diameter). This three-element criterion exploits the composition of the lead styphnate / antimony sulphide / barium nitrate primer formulation used in the vast majority of conventional centre-fire primers. Lead alone, or lead plus barium without antimony, may originate from many other sources (traffic pollution, some paints, some soils). The simultaneous presence of all three in a single spherical particle, with the specific composition ratios from a thermally melted and re-solidified primer residue, is highly characteristic of conventional GSR.

Limitations: lead-free and environmentally controlled ammunition. Beginning in the 1990s, manufacturers introduced "green" or "lead-free" primer formulations that substitute Diazol (DNBT-based, tetrazene-based) or SINTOX (zinc peroxide-based) or other formulations for the conventional lead styphnate mixture. These produce GSR particles that lack lead, antimony, or both. The ASTM E1588 standard includes a revised section (2011 update) addressing alternative GSR composition criteria for lead-free primer residues, typically relying on particles containing Zn, K, and N (for SINTOX primers), or Ba, Ca, and Si (for some European formulations), or Sn and Sb combinations. European forensic laboratories, particularly those in Germany (BKA), the Netherlands (NFI), and Scandinavia (NFC Sweden), see a higher proportion of lead-free primer cases because of EU environmental restrictions on lead in ammunition (under the EU Regulation on Hazardous Substances directive). The ENFSI Firearms Working Group published updated GSR interpretation guidelines in 2023 specifically addressing lead-free primer GSR.

SEM-EDS GSR examination workflow. The examination begins with automated SEM-EDS scanning. The specimen (an aluminium stub on which hand swabs have been deposited) is scanned in an automated particle-finding mode: the SEM rapidly acquires BSE images at low magnification, identifies particles above a size and Z-contrast threshold, automatically positions the beam on each candidate particle, collects a full EDS spectrum, and uses decision-tree software to classify the particle as characteristic, consistent with, or not characteristic of GSR. The elemental criteria that distinguish characteristic GSR from lead-free primer alternatives, and the analytical workflow for confirming three-component particles by instrumental GSR methods including SEM-EDS and AAS, are detailed in the forensic-ballistics module. Commercial automated GSR analysis systems (Thermo Fisher Scientific MorphoTopo, Bruker ESPRIT-Feature, Oxford Instruments AZtec) can screen 500-1000 stubs per day unattended. The analyst reviews the automated classifications, confirms borderline cases manually, and reports the number of confirmed GSR particles per stub.

Transfer and persistence. GSR particles transfer from the shooter's hands to surfaces and other persons by contact, and they are lost from hands by normal activity (washing, rubbing, pocket insertion). Persistence studies (reviewed by Blakey et al. 2018 in the Journal of Forensic Sciences) show that the majority of GSR on hands is lost within 4-6 hours for most individuals in normal activity. In a forensic case, the time between the shooting and the hand sample collection is therefore critical context for interpreting GSR particle counts. The FBI GSR protocol (as published in the FBI Laboratory Handbook), the UK FSR guidance on GSR evidence (2022), and the German BKA GSR standard operating procedure all specify that the collection time and activity history of the individual must be documented alongside the SEM-EDS result. For the sampling and persistence framework that governs how stubs are collected in the field, see the companion topics in the forensic-ballistics subject on GSR sampling protocols, persistence and secondary transfer.

Automotive paint examination by SEM-EDS complements optical microscopy (colour, thickness, layer count) and FTIR (polymer binder identification) by providing elemental composition data for each layer. This elemental data is essential when the questioned and known paint samples have the same optical properties (similar colour, similar layer count) but the elemental profiles allow discrimination, as happens frequently when multiple vehicle colours use the same pigment system with different extender formulations.

Sample preparation for paint cross-section SEM-EDS. A microtomed transverse section (as described in the paint examination topic) is mounted on an SEM stub with carbon adhesive, and carbon-coated to prevent charging. At 20 kV accelerating voltage and 10 mm working distance, the BSE image clearly shows layer interfaces by Z-contrast. The analyst measures each layer's thickness from the calibrated BSE image (cross-referenced against the optical comparison microscope measurement), then acquires EDS spectra at multiple points within each layer to establish the elemental profile.

RCMP PDQ database integration. The RCMP Paint Data Query (PDQ) database contains spectral and layer-sequence profiles for over 23,000 known automotive paint systems. The PDQ entry for a specific vehicle make, model, and model year includes the expected elemental profiles for each layer as determined by SEM-EDS. A forensic comparison matches the questioned paint's elemental profiles, layer by layer, against the PDQ entries for vehicles matching the class characteristics (colour, layer count) found optically. A match supports the conclusion that the questioned paint is consistent with originating from a vehicle in the matched PDQ category. In hit-and-run casework, this conclusion is typically phrased probabilistically using a likelihood-ratio framework, as specified in the ENFSI EPG Guidelines for Quality Assurance in Automotive Paint Examination.

Primer layer chemistry as a discriminator. The primer layer in an automotive paint system carries the highest elemental discriminating power because different manufacturers use different corrosion-inhibiting pigments: zinc phosphate, strontium chromate, basic lead silicate, or micaceous iron oxide. The combination of a specific primer elemental profile with a specific topcoat colour is a strong discriminating combination. This is why the RCMP PDQ examination protocol requires EDS data for every layer, not just the topcoat. Real-world hit-and-run casework from the FBI (reported in the ASTM E1610 paint-examination symposium proceedings) and the UK LGC Forensics automotive-paint database (a European Commercial Partner in the PDQ network) demonstrate cases where the topcoat colour alone was non-discriminating but the primer elemental profile narrowed the candidate vehicle list to a single make and model year. The full PDQ-database and casework workflow that contextualises these elemental results is covered in the paint databases, PDQ and EUCAP topic.

The transmission electron microscope passes a high-energy electron beam (typically 80-300 kV) through a thinly prepared specimen and detects the transmitted electrons. Because the specimen must be thin enough for electrons to penetrate (typically less than 100-200 nm for conventional TEM, and less than 50 nm for high-resolution work), sample preparation is more demanding than for SEM. The payoff is imaging capability at the lattice fringe level: individual atomic planes spaced 0.2-0.3 nm apart are resolvable in the best instruments.

TEM specimen preparation. For forensic work, TEM specimens are typically prepared by ultramicrotome sectioning (cutting sections of embedded material with a diamond knife to 60-100 nm thickness, floating them on water, and collecting on copper TEM grids). Alternatively, focused ion beam (FIB) milling (available in combined SEM-FIB instruments) can prepare site-specific cross-sections from specific regions of interest in a bulk specimen with nanometre precision. FIB preparation has become the preferred approach in forensic fibre and pigment work because it allows the analyst to select the exact cross-sectional location, whereas ultramicrotome sectioning produces sections at random positions along the material.

Forensic applications of TEM. Three forensic application areas justify TEM in a well-equipped laboratory. First, asbestos fibre identification below the optical resolution limit: chrysotile asbestos fibrils have a diameter of 20-40 nm, far below what PLM can measure. TEM images the tubular structure of chrysotile fibrils directly and can image the electron diffraction pattern, which is the definitive crystallographic identification. This is governed by EPA AHERA (40 CFR Part 763, Appendix A) and NIOSH Method 7402 for TEM asbestos analysis in air samples. Second, pigment particle characterisation in questioned paint and ink: TiO2 particles (100-300 nm) can be distinguished as anatase vs rutile polymorph by their electron diffraction spacing, providing more discriminating paint comparison data than SEM alone. Third, synthetic fibre ultrastructure: the crystal domain structure within drawn polymer fibres (the degree of crystallinity, crystal orientation, and amorphous-crystalline lamellar structure) is a forensic discriminator that FTIR and optical microscopy cannot access. These structural features correlate with the specific drawing and annealing conditions used in fibre manufacture, potentially identifying a specific production batch.

TEM-EDS and EELS. A TEM equipped with an EDS detector performs X-ray elemental analysis with spatial resolution determined by the beam spot size (as small as 0.1 nm in a field-emission TEM), far better than SEM-EDS. Electron energy-loss spectroscopy (EELS), a related technique, measures the energy lost by transmitted electrons to specific excitation events in the specimen, providing both elemental analysis (with higher sensitivity than EDS for light elements such as carbon, nitrogen, oxygen) and chemical bonding state information (distinguishing metallic iron from iron oxide from iron sulphide). EELS is not routine in forensic casework but has been applied in specialist manufacturing-defect cases where the bonding chemistry of a surface treatment was in dispute.

Interpreting GSR evidence in court requires addressing three questions: whether the GSR originated from a recent firearm discharge; whether it could have been acquired by secondary transfer; and what the particle count indicates about the probability of shooting versus innocent exposure.

Secondary transfer. GSR can transfer from a shooter's hands to a bystander's hands by handshake, to a seat surface by contact, to a police officer's hands by handling the suspect, or to an interview room surface. Studies of secondary transfer rates (Gassner and Stoecklein 2013, Gelsthorpe et al. 2019, and the ENFSI 2023 transfer position paper) show that secondary transfer rates are low (typically 1-5 characteristic particles per transfer event) but not negligible. High particle counts (greater than 10-20 confirmed Pb-Sb-Ba particles) are much less likely to arise from secondary transfer than low counts (1-3 particles). The analyst should report both the count and its context.

US GSR evidence. US federal and state courts treat GSR evidence under the Daubert standard (or Frye in non-Daubert states). Courts have admitted GSR-SEM-EDS evidence routinely. United States v. Llera Plaza (E.D. Pa. 2002) was a notable Daubert challenge, though it concerned latent fingerprint (ACE-V) methodology rather than SEM-EDS; GSR-SEM-EDS has been admitted as reliable under both Daubert and Frye standards in numerous federal and state proceedings. The FBI GSR protocol and OSAC Firearms and Toolmarks standards provide the baseline methodology that expert witnesses reference when challenged. US prosecutors typically present GSR evidence as a corroborative rather than standalone element.

UK GSR evidence. The UK Forensic Science Regulator's 2022 GSR evidence guidance required that expert reports include: particle count, method used (automated vs manual), comparison against population data (background GSR prevalence studies), and an explicit statement that GSR can be acquired by secondary transfer. R v. Barry George (2007 EWCA Crim 2722) is the landmark UK case on GSR evidence: George's 1999 murder conviction was quashed in 2008 partly because the GSR evidence (a single particle found in a coat pocket) was later assessed as consistent with background contamination. The case led to a revision of UK GSR reporting standards requiring population prevalence data for single-particle findings.

India. Indian forensic laboratories (CFSL Central, CFSL Hyderabad, CBI Forensic Science Laboratory) perform GSR examination by SEM-EDS, following a protocol broadly aligned with ASTM E1588 and the ENFSI guidelines. Case reports are submitted to court as part of the forensic expert's written opinion under BSA 2023 Chapter VIII (Expert Opinion). Indian courts have not yet had a GSR-specific Daubert-equivalent challenge, but the Supreme Court's general guidance on scientific evidence credibility (State of Haryana v. Ram Singh, 2002; and subsequent judgments on DNA and technical evidence) establishes that the examiner must disclose the method, its limitations, and any alternative explanations for the finding.

Germany, Netherlands and EU. The German BKA and the Dutch Netherlands Forensic Institute (NFI) are among the most active European centres for GSR research. The NFI's probabilistic likelihood-ratio framework for GSR interpretation (Ploum et al. 2017, Forensic Science International) expresses the GSR finding as a likelihood ratio comparing the hypothesis that the person fired a weapon versus the hypothesis that the person did not fire a weapon but had normal occupational exposure. This probabilistic approach, also recommended by the ENFSI Firearms Working Group 2023 update, is more conservative and scientifically rigorous than a simple presence/absence statement, and is the direction in which UK and EU GSR reporting is moving.

US: ASTM E1588 and OSAC. ASTM E1588 is the primary US standard for GSR examination by SEM-EDS. It specifies: the stub collection protocol (conductive adhesive on aluminium stub, specific swabbing technique), the SEM operating conditions (accelerating voltage, working distance, minimum magnification for particle search), the EDS spectrum quality requirements (minimum count rate, energy calibration tolerance), the three-element criterion for confirmation, and the particle-count reporting format. The OSAC Firearms and Toolmarks Subcommittee has adopted E1588 as a Referenced Standard and is developing a supplementary document on lead-free GSR analysis.

UK: FSR GSR guidance. The UK Forensic Science Regulator's 2022 GSR evidence guidance (FSR-EFG-01) specifies ISO 17025-accredited methodology, automated particle-finding as the required primary search method (manual-only scanning is accepted only as a complement for specific particle types), and a minimum of 100 fields or 1 square centimetre scan area per stub. The guide also specifies the background prevalence data that must be referenced when reporting low particle counts: the UK population prevalence studies (Hofstetter 2015, Rettler 2018) established background rates for the UK population that context a single-particle finding.

India: NABL accreditation and DFSS manual. Indian forensic laboratories performing SEM-EDS are required to hold NABL accreditation under ISO 17025 to submit results as scientific expert evidence in court. The DFSS Forensic Science Manual (2019 edition) includes a GSR-SEM-EDS chapter specifying stub collection, instrument operating parameters, the three-element criterion, and report format. Under BSA 2023 Section 63, SEM-EDS result files stored digitally require a hash-based integrity certification before they can be submitted as electronic records in court proceedings.

Canada and Australia. The RCMP National Forensic Laboratory Services operates an accredited SEM-EDS GSR programme aligned with ASTM E1588, with a published internal protocol that specifies both conventional and lead-free GSR criteria. The Australian Centre for Forensic Sciences (part of the NSW Police Forensic Services Group) and the Victoria Police Forensic Services operate ISO 17025-accredited SEM-EDS GSR programmes; both have adopted the probabilistic likelihood-ratio reporting framework for low-count GSR results, following the NFI Dutch model.