Soil Examination: Density Gradient, Mineralogy, Palynology, Diatomology

The density-gradient column is a tall glass cylinder containing a liquid medium that varies continuously in density from bottom to top. When a soil sample is introduced at the top of the column, individual mineral grains sink until they reach the layer where the liquid density equals their own grain density, and they float there. The result is a series of horizontal zones, each enriched in minerals of a particular density range, stacked from the bottom (densest minerals) to the top (least dense minerals). The visual pattern of these zones is the density-gradient profile.

Column construction. The standard forensic soil density-gradient column uses a mixture of bromoform (CHBr3, density 2.89 g/cm3) and bromobenzene (C6H5Br, density 1.50 g/cm3). By varying the ratio of the two liquids, the analyst can create any intermediate density between 1.50 and 2.89 g/cm3. For soil mineral analysis, a column covering the range 2.2-2.9 g/cm3 is adequate for most quartz-to-hornblende mineral assemblages. The gradient is established by a two-bulb mixing device that continuously adds the lighter liquid to the denser reservoir through a mixing chamber, producing a column with linearly increasing density from top to bottom. The gradient is stabilised by capillary action and by gravity, and remains usable for 48-72 hours at room temperature.

Sample preparation. The questioned and known soil samples are pre-treated identically. Organic matter is removed by low-temperature (30-35°C) hydrogen peroxide digestion, which destroys humus without dissolving carbonates or dissolving clay minerals. The sample is then washed with distilled water, dried, and disaggregated to individual grains. The sub-2 mm fraction is used. Grains are introduced as a dilute slurry onto the surface of the column. The column must be pre-wetted with the column liquid by adding a small rinse volume to prevent surface-tension artefacts at the introduction point.

Reading and comparison. After equilibration (typically 30-60 minutes for fine sand particles; several hours for clay-size particles, which are rarely used in density analysis), the column is photographed against a uniform light background. The depth-density calibration, established by introducing reference solutions of known density before the sample, converts the visual band positions to density values. The analyst records: the number of bands, their positions (density values), their relative widths, and their colours (brown bands indicate ferruginous minerals; colourless bands indicate quartz; black bands indicate magnetite or ilmenite).

Comparison between questioned and known samples uses the same column simultaneously (side-by-side comparison columns) or sequential columns run under identical conditions. A match requires the same number of bands at the same density positions within the measurement uncertainty, typically plus-or-minus 0.02 g/cm3. Differences in band position, width, or colour indicate a different mineral assemblage. The FBI Soil Examination Unit, as documented in Raymond Murray's case studies and in the OSAC Trace Evidence Subcommittee draft standard (2023), treats the density-gradient profile as a second-tier comparison parameter after Munsell colour and before XRD.

Boundary with Module 4. The density-gradient column technique as a physical-measurement method for trace-evidence comparison is introduced in the Module 4 topic on density measurement. The soil-specific protocol described here covers the mineral-assemblage interpretation, the bromoform-bromobenzene mixture rationale, and the forensic casework logic specific to soil comparison. There is no duplication of the basic physics of settling or the Stokes equation, which are treated in Module 4.

The texture-class determination described in Topic 1 gives a three-component description (% sand, silt, clay). A more detailed particle-size distribution (PSD) provides the full continuous grain-size spectrum from clay through fine sand, and this more detailed picture is a stronger discriminator than the three-class summary.

Wet sieving. The standard first step in particle-size analysis is wet sieving. The air-dried, disaggregated sample is washed through a stack of stainless-steel sieves with mesh sizes from 2 mm down to 0.063 mm (63 micrometres), typically at 2 mm, 1 mm, 0.5 mm, 0.25 mm, 0.125 mm, and 0.063 mm. The mass retained on each sieve is weighed, and the cumulative percentage finer-than curve is plotted. This curve, standardised by ASTM D422 in the US and by BS 1377-2 in the UK, describes the sand sub-fractions. The fraction passing the 0.063 mm sieve (silt and clay combined) is then subjected to further analysis.

Laser diffraction. Laser diffraction particle-size analysis (also called laser granulometry) measures the angular pattern of laser light scattered by particles passing through a beam. Particles scatter at angles inversely related to their size: large particles scatter at small angles, small particles scatter at large angles. The Mie scattering theory or the Fraunhofer diffraction approximation (for particles larger than 10 micrometres) relates the scattering pattern to a particle-size distribution in the range of 0.02-2000 micrometres. Instruments like the Malvern Mastersizer 3000 or the Beckman Coulter LS 13 320 produce a full continuous PSD in 3-5 minutes per sample. The ASTM C1070 and ISO 13320 standards define laser-diffraction PSD methodology.

The key forensic advantage of a full laser-diffraction PSD over a simple texture class is its statistical discriminating power. A soil with clay loam texture (35% clay, 35% silt, 30% sand) can be distinguished from a second clay loam from a different location by comparing their full laser-diffraction curves. The shape of the curve, its modality (unimodal vs bimodal vs multimodal), the D10/D50/D90 percentile grain sizes, and the sorting coefficient are all quantitative parameters that can be compared using statistical distance metrics (Mahalanobis distance, Kolmogorov-Smirnov test).

In routine forensic casework, the UK Home Office Forensic Science Service (when it existed) and its successor private providers (Eurofins Forensics, Cellmark, Key Forensic Sciences) use laser diffraction as a standard second-tier soil comparison tool. The RCMP National Forensic Laboratory Services uses a similar approach. In Australia, the ANZFSS Standards and Guidelines for Forensic Geological Analysis recommend laser diffraction as the preferred particle-size method, citing the Malvern Mastersizer platform as the reference instrument; several Australian state forensic laboratories have adopted it in casework since 2015. The Indian CFSL Hyderabad has Malvern Mastersizer instruments but their use in soil casework is not yet routine in all reported cases.

Polarising-light microscopy (PLM) of sand-fraction mineral grains is one of the most information-dense single techniques in forensic soil analysis. Each mineral species has a distinctive combination of optical properties, all measurable on the petrographic microscope: colour, pleochroism, cleavage angle, habit, refractive index relative to the immersion oil, birefringence (retardation colours under crossed polarising filters), optical sign (positive or negative), and extinction angle. Together these properties form a deterministic identification key for most rock-forming minerals.

Preparation of a PLM mount. The sand fraction (0.063-0.5 mm, obtained by sieving) is mounted in a calibrated immersion oil of known refractive index on a glass slide, covered with a thin coverslip, and examined under the petrographic microscope at 100-400x magnification. The analyst first surveys the slide in plane-polarised light (PPL), noting grain colour, shape, surface texture, inclusions, and cleavage. The analyser is then inserted to produce crossed polarising filters (XPL), and the birefringence colours of each grain are recorded. Grains are identified to mineral species using standard optical mineralogy references, particularly the Tröger tables or the Nesse "Introduction to Optical Mineralogy" reference.

The heavy-mineral fraction. The sub-fraction of grains with density above 2.85 g/cm3, the heavy-mineral fraction, is particularly forensically valuable because it reflects the trace-mineral suite of the parent rock, which varies geographically more than the dominant quartz-feldspar assemblage. Typical heavy minerals in different rock-types: basalt-derived soils are rich in pyroxene (augite) and hornblende; granite-derived soils are rich in zircon, apatite, and muscovite; metamorphic soils may contain garnet, staurolite, kyanite, and sillimanite; sedimentary soils from aeolian deposits may have a distinctive rounded, frosted zircon population. Heavy-mineral assemblage counting is a standard technique in the sedimentological literature (the HUSUM count, 200-300 grains per sample) and has been applied in forensic geology since the 1970s.

The FBI Soil Examination Unit, as described in Murray (2011) "The Forensic Examination of Soils and Sediments," counts a minimum of 200 heavy-mineral grains per sample to produce a statistically reliable mineral frequency distribution. The ENFSI ENG-FG1 guideline recommends a minimum count of 100 grains but acknowledges that 200 or more is preferable for statistical comparison. The Indian FSL soil examination procedures require PLM mineralogy when initial colour and texture comparison is inconclusive.

Powder X-ray diffraction (XRD) identifies crystalline minerals by their unique d-spacings, the distances between parallel planes of atoms in the crystal lattice. When a beam of monochromatic X-rays (typically CuKα radiation at 1.54 angstroms from a Cu X-ray tube) strikes a powdered sample, constructive interference occurs at specific angles (given by Bragg's law: nλ = 2d·sinθ) for each crystal plane family. The result is a diffractogram with peaks at positions that are diagnostic of mineral identity: quartz produces a major peak at 2θ = 26.65°; calcite at 29.4°; kaolinite at 12.4°; illite at 8.8°; smectite (montmorillonite) at 5-6° when air-dried and 16-17° after glycolation.

Sample preparation for bulk XRD. For bulk mineral analysis, the soil sample is ground in a ball mill to a 10-20 micrometre particle size to produce true powder diffraction. It is then packed into a flat aluminium or silicon sample holder and placed in the diffractometer. The ASTM D934 and ISO 29581-2 standards describe the required preparation procedures. Semi-quantitative mineral abundance can be estimated from peak heights or integrated areas using the Reference Intensity Ratio (RIR) method or full-profile Rietveld refinement, implemented in software packages such as TOPAS, HighScore Plus, or SIROQUANT.

Clay-mineral XRD analysis. Clay minerals require special preparation because their diffraction peaks are sensitive to swelling, interlayer composition, and crystallinity. Three oriented clay slides are prepared from the less-than-2-micrometre fraction: air-dried, glycolated (ethylene glycol solvated overnight to expand swelling clays), and heated at 550°C (to collapse kaolinite and distinguish it from chlorite). The presence of smectite, which expands from 14 Å to 17 Å on glycolation, is confirmed by comparing the three diffractograms. This three-state preparation is the standard procedure in the UK's British Geological Survey (BGS) laboratory and in USGS soil mineralogy protocols.

Forensic application. For a questioned soil versus a known soil comparison, a matched XRD profile, in which the peak positions, relative heights, and clay-mineral assemblage are the same within measurement uncertainty, is strong evidence of mineralogical consistency. XRD cannot distinguish between samples from different locations that happen to have the same mineralogy, but combined with particle-size and Munsell colour data, a matching XRD profile substantially narrows the geographic source range. The FBI Soil Examination Unit has used XRD as a routine third-tier comparison tool in major casework since at least the 1980s. The RCMP uses it in parallel with scanning electron microscopy / energy-dispersive X-ray spectroscopy (SEM-EDS) for soil elemental composition. In India, the CFSL New Delhi has an XRD facility used for soil and building-material analysis.

Forensic palynology is the application of pollen and spore analysis to legal investigations. Pollen grains and fungal spores are produced in enormous quantities by plants and fungi, dispersed through the air, water, and soil, and deposited in characteristic assemblages that reflect the local plant community at the time of deposition. Because each plant species produces morphologically distinctive pollen, and because plant species have defined geographic ranges and habitat preferences, the pollen assemblage in a soil sample is a biological map reference that constrains the sample's geographic origin with a resolution that physical and chemical methods alone cannot achieve.

Pollen morphology and identification. Pollen grains are identified by a combination of morphological characters: the overall outline (spherical, prolate, oblate, triangular in polar view), the number and type of apertures (colpi, pores, colporate), the surface texture (psilate, granulate, reticulate, echinate, striate), and the wall structure (intine and exine layers). The exine wall is composed of sporopollenin, one of the most chemically resistant biological polymers known, and it survives acid hydrolysis, organic-solvent extraction, and decades to centuries of burial in soil. Pollen grains are typically 10-100 micrometres in size; spores of ferns, mosses, and fungi range from 5 to 200 micrometres. Identification to genus is possible for most taxa; species-level identification is possible for many wind-pollinated trees and grasses.

The Bryant and Mildenhall framework. Vaughn Bryant (Texas A&M University) and Dallas Mildenhall (GNS Science, New Zealand) developed the systematic casework protocol for forensic palynology that is now the standard reference in most English-speaking forensic communities. Their framework, documented most comprehensively in Bryant and Mildenhall (1998) "Introduction to Forensic Palynology" and Mildenhall (2004) "Hypericum pollen determines the presence of burglars at the scene of a crime" (Forensic Science International), requires: (1) comparison of the questioned pollen assemblage from the evidence item against the known assemblage from the scene; (2) accounting for background pollen (the regional rain of airborne pollen that occurs everywhere); and (3) identification of indicator species with restricted geographic ranges or specific habitat associations that can distinguish scene-specific from background assemblages.

Bosnia mass-grave investigations. The most internationally prominent application of forensic palynology to grave-site investigation is the International Criminal Tribunal for the Former Yugoslavia (ICTY) casework involving the Srebrenica 1995 massacre graves, in which palynological analysis of soil samples from bodies and from burial sites contributed to the linkage of secondary graves to primary graves and to specific geographic locations. Dallas Mildenhall's analysis in the Mihaela Marincu proceedings (2009), which involved pollen assemblage comparison between soils adhering to human remains and soils from candidate sites in Bosnia-Herzegovina, demonstrated that the assemblage profiles were distinctive enough to support geographic sourcing opinions at the courtroom level. The case established a precedent in the international criminal-law community for accepting forensic palynology evidence. Similar methodology was used in the 2005-2010 Kosovo grave investigations by the International Commission on Missing Persons (ICMP).

UK casework. In the UK, forensic palynology has a long operational history. The Birmingham pub bombings (1974) generated early casework in which pollen from a suspect's clothing was compared against pollen from a known scene location. The Forensic Science Service employed specialist palynologists before its closure in 2012; their casework methodology, documented in Horrocks and Walsh (1999) "Forensic Palynology" (Journal of Forensic Sciences), provides the technical reference used by UK academic forensic palynologists who now provide expert witness services on a consultancy basis.

US casework. In the United States, Bryant's group at Texas A&M has provided palynology expert testimony in multiple federal and state criminal cases. The technique is admitted under Daubert in several circuits, with the most recent comprehensive review of its scientific basis provided by the OSAC Trace Evidence Subcommittee. The US lacks a centralised reference pollen library comparable to the New Zealand Pollen Reference Collection at GNS Science, but regional libraries maintained at university herbaria serve the same function.

Indian context. Forensic palynology is not yet an established routine technique in Indian SFSLs or the CFSL network. Academic research groups at the Birbal Sahni Institute of Palaeosciences (Lucknow) and at several botanical-survey institutions maintain pollen reference collections for paleoenvironmental research. These collections are relevant to forensic palynology, and there are peer-reviewed case reports from India applying pollen analysis to drug-trafficking (opium poppy pollen on drug shipments) and to environmental crime. The gap is at the operational-laboratory level: the CFSL and SFSL network does not yet have designated palynology analysts or accredited pollen comparison protocols, though this is beginning to change following the DFSS (Directorate of Forensic Science Services) capacity-building initiatives from 2020 onward.

Diatoms are unicellular photosynthetic algae (class Bacillariophyceae) that construct an intricate siliceous cell wall called the frustule from biogenic silica (opal). Frustules are species-specific in shape, size, raphe configuration, and striae pattern, and they survive in sediments for geological timescales. Two forensic applications are distinguished here: drowning attribution (covered in depth in the forensic-medicine subject, drowning module) and soil-source linkage, which is the Module 6 application.

Diatom assemblages in soil. Soils near water bodies, particularly those in floodplains, lakeshores, peatlands, wetlands, and paddy fields, contain diatom frustules that have been deposited from the overlying water by flooding, splash, or aeolian transport. The diatom assemblage in a soil sample reflects the ecology of the water body: acid, soft-water mountain lakes produce a different assemblage (acid-tolerant Tabellaria, Pinnularia, Frustulia species) than alkaline, hard-water lowland rivers (Cocconeis, Amphora, Denticula). Marine-influenced coastal soils contain marine and brackish-water diatom taxa absent from inland freshwater soils. These ecological constraints make diatom assemblage comparison a geographic tracer with spatial resolution in the range of a few to tens of kilometres, depending on the specificity of the indicator species.

Forensic diatomology for soil linkage. When a questioned soil sample contains diatoms, their assemblage comparison against a known soil can supplement or confirm the mineral and pollen data. A classic application is the investigation of clandestine graves in floodplain locations: the pedogenic diatom assemblage at the grave site is typically distinct from the regional background because it reflects the site-specific hydrology. Sgt. Chris Burgess of the South Yorkshire Police (UK) demonstrated in a 2003 case study that diatom assemblage comparison between soil adhering to a suspect's clothing and soil from a river-bank scene produced a match that supported the prosecution's scene-association argument.

Relationship to the drowning-attribution application. The forensic-medicine drowning topic covers the lung-diatom test (the Peabody test) in detail, including its disputed evidential value and the requirement for species-matched comparison between lung diatoms and water-source diatoms. This topic covers only the soil-linkage application: frustules present in the soil's mineral or organic fraction, compared between questioned and known samples for geographic sourcing. The underlying methodology (frustule extraction, acid digestion, slide preparation, counting, identification) is the same in both applications, but the interpretive framework is different. The drowning application asks "did this person drown here?"; the soil-linkage application asks "did this soil come from near a water body with this diatom assemblage?".

Standard procedure. Diatom extraction from soil follows the Hendey (1964) acid-digestion protocol or its modern equivalents: the sample is treated with cold HCl to dissolve carbonates, then with concentrated H2SO4 and HNO3 to destroy organic matter, washed repeatedly with distilled water, and the cleaned frustules are mounted in Naphrax (refractive index 1.73) or Pleurax high-refractive-index medium on a glass slide. Counts of 200-400 frustules per slide are standard, with identification to genus and species using the Krammer-Lange-Bertalot reference series ("Susswasserflora von Mitteleuropa") for European taxa or the Patrick and Reimer "Diatoms of the United States" for North American taxa. Indian diatom taxonomy is covered by the monograph series of the Botanical Survey of India, particularly the Freshwater Algae volumes.

The Murray-Tedrow comparison protocol, adopted by the FBI Soil Examination Unit and referenced in the ENFSI ENG-FG1 guideline, organises the soil examination techniques into a tiered workflow where each tier adds discriminating power and cost.

Tier 1: Colour and gross morphology. Munsell colour, macroscopic texture (sticky vs crumbly vs sandy), and the presence of obvious inclusions (charcoal fragments, root material, anthropogenic debris) are assessed first. These observations take minutes and can rule out a geographic match quickly when colour or texture is dramatically different. The FBI estimates this tier eliminates approximately 60-70 percent of no-match comparisons.

Tier 2: Particle size, organic matter, and density gradient. Laser-diffraction PSD, LOI organic matter, and the density-gradient column profile are performed when tier 1 does not exclude a match. Together, these three parameters provide a quantitative, instrument-based description of the physical-chemical composition of the sample. They take half a day and cost significantly more, but they add substantial discriminating power. The density-gradient profile alone, when distinctly characterised by multiple heavy-mineral bands, can rival the discriminating power of XRD in many cases.

Tier 3: Mineralogy. PLM of the sand and heavy-mineral fractions and powder XRD of the bulk and clay-mineral fractions are the most time-intensive and most discriminating physical-chemical analyses in the tier. When tiers 1 and 2 do not exclude a match, tier 3 mineralogy is required before a positive association opinion can be stated. In international casework, the combination of Munsell colour, particle-size distribution, and matched XRD profiles has supported source-association opinions admitted in US federal courts, UK Crown Courts, the ICTY, and the ICMP proceedings.

Tier 4: Biological analysis. Palynology and diatomology are applied when the physical-chemical data are inconclusive, or when the case context suggests that biological data (geographic seasonality, habitat specificity, site-specific exotic species) might provide discriminating information not available from mineralogy. They are the most expensive, most time-consuming, and potentially most site-specific techniques in the stack.

The reporting standard in the US (OSAC draft 2023), UK (FSR Code of Practice), and under ENFSI ENG-FG1 requires the forensic soil examiner to report: the specific techniques applied, the comparison criteria used, the direction of the opinion (consistent with / inconsistent with a common source), and the strength of the evidence in a Bayesian likelihood-ratio framework where the case permits such quantification. A simple assertion of "match" without methodology is not acceptable under any of these standards.