Practice with national-level exam (FACT, FACT Plus, NET, CUET, etc.) mocks, learn from structured notes, and get your doubts solved in one place.
Density as a forensic comparison parameter: physical basis (mass per unit volume, temperature dependence), measurement classes — sink-float (qualitative match-or-mismatch using bromoform + bromobenzene mixtures), density-gradient column (continuous liquid column with neutral-buoyancy zone), pycnometry (precise volumetric flask), hydrostatic weighing and modern oscillating-U-tube digital densitometers; the casework arc for soil, glass and polymer comparison and the role of the Stokes settling formula in density-gradient analysis.
Last updated:
Density is one of the oldest physical properties exploited in forensic comparison work, and it remains relevant precisely because it is difficult to fake. A glass fragment found in a suspect's clothing either matches the density of the source window or it does not. That match, when combined with refractive index and elemental fingerprinting, moves the casework from a qualitative observation to a statistically defensible opinion on source attribution. The measurement is not glamorous, but its place in the trace-evidence toolkit is secure.
The physical basis is uncomplicated: density is mass divided by volume, expressed in grams per cubic centimetre or kilograms per cubic metre. What makes it a useful forensic discriminator is its sensitivity to composition. A soda-lime float glass from a residential window sits near 2.49 g/cm³. Borosilicate glass from a laboratory flask sits closer to 2.23 g/cm³. Tempered automotive side glass and laminated windshield glass occupy overlapping but distinguishable ranges, and the density of the parent batch is remarkably consistent within a single manufacturing run. That manufacturing consistency, combined with natural variation between batches and manufacturers, creates the discriminating power that forensic scientists exploit in fragment comparison.
Four measurement classes cover the forensic laboratory's needs. The sink-float test is the rapid qualitative screen, placing a fragment into a liquid whose density can be adjusted until the fragment neither sinks nor floats. The density-gradient column provides a semi-quantitative continuous display of density across a population of fragments. Pycnometry and hydrostatic weighing give absolute density values with sufficient precision for scientific reports. The oscillating-U-tube digital densitometer, originally developed for pharmaceutical quality control, now appears in the more equipped forensic-physics laboratories as a high-throughput alternative for liquids and dissolved materials.
The casework arc runs from glass and soil in trace-evidence comparison through polymer identification in packaging and product tampering. Each application has its own precision requirement and its own reference dataset, from the ASTM E2365 glass-comparison standard used in US FBI and NIST laboratories to the ENFSI ENG3-2013 European guideline and the DFSS protocols under India's Directorate of Forensic Science Services. The underlying measurement principles, though, are the same across all these jurisdictions.
*Density is mass divided by volume, but in a forensic context, the real challenge is measuring volume accurately enough to matter.*
Density is formally defined as mass per unit volume, symbolised ρ = m/V. For solids used in trace-evidence comparison (glass fragments, mineral grains, polymer pellets), mass is measured on an analytical balance to four decimal places without difficulty. Volume is the problem. The volume of an irregular glass shard cannot be measured geometrically; it must be inferred from fluid displacement or from the geometry of a precisely calibrated vessel.
Temperature dependence is an often-underappreciated source of measurement error in density work. The density of a liquid changes with temperature, so any density measurement that relies on matching a solid fragment to a liquid of known density must be performed at a controlled temperature. A change of 1 degree Celsius shifts the density of a bromoform-benzene mixture by approximately 0.002 g/cm³. At the precision levels needed to discriminate glass batches, where the between-batch spread may be only 0.001-0.003 g/cm³, a 1 °C temperature fluctuation during measurement can produce a false positive or a false exclusion. The ASTM E2365 standard for glass-fragment comparison specifies that all density measurements should be performed in a temperature-controlled room at 25 ± 0.1 °C, and that the reference liquid's temperature should be monitored throughout the measurement.
The forensic precision requirement varies by application. For the qualitative sink-float screen, reproducibility to ±0.01 g/cm³ is acceptable. For the density-gradient column used in a casework comparison, ±0.001 g/cm³ is expected. For pycnometry used to establish a reference value, ±0.0005 g/cm³ is achievable with careful technique. Modern oscillating-U-tube densitometers (the Anton Paar DMA series is the instrument most widely cited in published forensic-liquid-density work) routinely achieve ±0.0001 g/cm³ for liquids, though not for solid fragments.
The Stokes settling formula governs the rate at which a particle migrates through a liquid column under gravity and is essential to understanding density-gradient behaviour. Stokes' law states that the settling velocity v of a spherical particle of density ρₛ in a fluid of density ρ_f and viscosity η is: v = 2r²g(ρₛ - ρ_f) / 9η, where r is the particle radius and g is gravitational acceleration. A heavier particle sinks; a lighter one rises; a particle exactly matching the liquid density at a given depth remains stationary. In a gradient column, the liquid density increases continuously from top to bottom, so each particle migrates until it reaches the depth where liquid and particle density are equal. The time required to reach equilibrium depends on particle size through the Stokes equation.
*The simplest density test has been in forensic use since the 1920s and still earns its place as the first filter in glass-fragment examination.*
The sink-float test uses a single liquid of known density, or a mixture whose density can be adjusted, to provide a binary include-or-exclude result for a solid fragment. The fragment either sinks (fragment denser than liquid), floats (fragment less dense than liquid), or remains suspended at neutral buoyancy (fragment density matches liquid density).
The two liquids most commonly used in glass and mineral comparison are bromoform (CHBr₃, density 2.890 g/cm³ at 20 °C) and benzene (C₆H₆, density 0.879 g/cm³) or its less hazardous modern substitute, bromobenzene (C₆H₅Br, density 1.495 g/cm³). Mixing these in varying proportions produces reference liquids at any density between roughly 0.88 g/cm³ and 2.89 g/cm³, bracketing the entire range of glass, mineral, and most polymer densities encountered in forensic work.
The practical protocol is straightforward. The questioned fragment is dropped into the test liquid at a measured temperature. If the fragment sinks, the liquid's density is increased by adding denser liquid until neutral buoyancy is achieved. If it floats, lighter liquid is added. A calibrated density float or a digital densitometer measures the liquid density at the equilibrium point. The ASTM E2365 protocol, adopted by FBI laboratories across the US, prescribes this procedure with explicit temperature monitoring and calibration checks against reference glass fragments of known density. The UK Forensic Science Regulator's Codes of Practice (Annex H on glass) similarly require calibration verification before each analytical session.
Limitations are well understood. The method is destructive of the reference liquid if repeated many times without recovery. The precision is limited by the sensitivity of the liquid-density measurement, not by the principle itself. For very small fragments (below approximately 0.1 mg), achieving neutral buoyancy is uncertain because surface tension effects become significant relative to buoyancy. For routine casework screening, these limitations are acceptable.
In India, DFSS-accredited laboratories performing glass comparison under the guidelines issued by the Directorate of Forensic Science Services (Government of India, 2019 Forensic Laboratory Manual) routinely use the bromoform-benzene sink-float test as the primary density screen before proceeding to the gradient column. The UK Home Office Forensic Science Regulator and Australia's ANZFSS member laboratories follow analogous two-stage protocols: sink-float screen, then gradient column for any fragment pairs that survive the initial exclude screen.
*A gradient column does something no single-liquid test can: it shows the density distribution of an entire population of fragments simultaneously.*
The density-gradient column is the workhorse of glass and soil density comparison because it converts a one-at-a-time binary test into a visual, population-level comparison. The column is a tall glass cylinder (typically 25-50 cm long, 2-3 cm internal diameter) filled with a liquid whose density increases continuously from top to bottom. Glass fragments dropped into the column migrate to their neutral-buoyancy depth and stop, producing a visible band at a characteristic depth. Known-density calibration floats, placed in the column alongside the fragments, establish the density scale at each depth.
The column is prepared by layered filling with miscible liquids of differing densities, then allowing diffusion to smooth the boundaries into a true gradient. Bromoform-bromobenzene and bromoform-benzene are again the standard pairs. The gradient stability is checked by verifying that calibration floats remain at their expected positions over the observation period, typically 24-48 hours. Temperature control is mandatory for the same reasons described in Section 1.
A key advantage over the sink-float test is population characterisation. If a questioned glass sample contains ten fragments from the same source, all ten should band at the same depth within a narrow range reflecting measurement uncertainty and true within-sample variation. A source-glass control sample produces its own band. If the questioned and control bands overlap within the measurement uncertainty, that is evidence consistent with a common source. If they are displaced, the difference in density is grounds for exclusion. The ENFSI ENG3-2013 guideline for glass comparison (published by the European Network of Forensic Science Institutes) specifically endorses the density-gradient column for population-level comparison and provides inter-laboratory precision data from proficiency trials across European forensic laboratories.
The Stokes settling formula discussed in Section 1 governs how quickly the fragments reach equilibrium. Small fragments (< 0.5 mm) may require overnight equilibration before the band position is stable. For expedited casework, centrifugal gradient columns (prepared in a centrifuge tube and spun) are described in the research literature and are used by some RCMP (Royal Canadian Mounted Police) forensic laboratories to compress the equilibration time.
Soil comparison using the density-gradient column yields a density-fraction profile rather than a single density value because soil is a heterogeneous mixture of particles with different densities. Quartz grains (density 2.65 g/cm³), iron oxide particles (density 4-5 g/cm³), feldspar (density 2.56-2.62 g/cm³), and organic matter (density 0.9-1.5 g/cm³) all band at different depths. The entire soil sample produces a density-fraction fingerprint, a set of band positions and relative band intensities that can be compared between questioned and control samples. The FBI Soil Examination Unit protocols describe this density-profile comparison approach in detail. Module 6 of this subject covers soil casework; the density-gradient method here is the common physical tool.
*When a casework opinion needs a specific numerical density value rather than a comparative band position, pycnometry delivers it.*
The density-gradient column provides relative comparison with high visual clarity but does not give an absolute density value. When a casework report requires a precise density measurement to be cited, or when no suitable reference material is available for a gradient comparison, pycnometry or hydrostatic weighing provides the absolute value.
A pycnometer is a glass flask with a stopper that has a capillary hole, calibrated to contain a precisely known volume of liquid at a specified temperature. The procedure for solid density uses three measurements: the dry mass of the solid fragment, the mass of the pycnometer full of liquid (water or a non-reactive organic liquid), and the mass of the pycnometer containing both the solid fragment and the liquid. The displaced-liquid volume equals the difference between the mass of the pycnometer full of liquid alone and the mass when partially filled by the solid. Dividing the solid's mass by this displaced volume gives its density. The calculation is rigorous and accurate to approximately ±0.002 g/cm³ for fragment masses above 0.5 g, but becomes progressively less precise for smaller fragments because the mass difference used to calculate displaced volume is a small number divided by another small number.
Hydrostatic weighing (Archimedes weighing) uses the buoyant force on a solid submerged in a liquid of known density. The fragment is suspended on a fine wire from an analytical balance and weighed first in air and then fully immersed in a liquid. The difference in apparent weight equals the weight of the liquid displaced, and from that, the volume is derived. Modern hydrostatic weighing kits, offered as accessories by Mettler-Toledo, Sartorius, and Ohaus, convert any analytical balance into a hydrostatic weighing station with a precision of ±0.001 g/cm³ for fragments above approximately 100 mg. The method is particularly suited to irregularly shaped fragments that cannot be measured by any other volumetric technique.
Both pycnometry and hydrostatic weighing require correction for the buoyancy of air on the mass measurements (the Siebert correction), which at atmospheric pressure introduces an error of approximately 0.0012 g/cm³ if neglected. The ASTM E2365 glass standard requires this correction to be applied. NIST (National Institute of Standards and Technology, US) Standard Reference Material 8092 (glass fragments of certified density) provides the calibration reference used by FBI-accredited laboratories for both pycnometry and hydrostatic weighing.
The UK Forensic Science Service (before its 2012 closure) published proficiency-trial data showing that pycnometry measurements of glass density across its participating laboratories had a between-laboratory standard deviation of approximately 0.0015 g/cm³, confirming that the method achieves the precision needed to discriminate between glass batches whose densities differ by 0.003-0.005 g/cm³. The current Forensic Science Regulator's Codes of Practice, inherited from that FSS work, specify the minimum precision requirement for glass-density measurement used in comparative reporting.
*The same instrument that controls sugarcane liquor quality in a distillery is now measuring the density of dissolved drug samples in a forensic laboratory.*
The oscillating-U-tube digital densitometer works on the principle that the resonant frequency of a U-shaped glass tube filled with a liquid depends on the tube's effective mass, which includes the sample mass. Measuring the resonant frequency therefore gives the density of the sample. The relationship is calibrated with two liquids of known density (typically air and deionised water) and is accurate to ±0.0001 g/cm³ across a wide density range without any operator-dependent judgment about neutral buoyancy.
The instrument's primary forensic application is the measurement of liquid-sample density: sugar solutions in food-fraud investigations, ethanol-water mixtures in alcohol adulteration cases, and dissolved-drug solutions in pharmaceutical counterfeiting. In a sugar-adulteration investigation, the density of the questioned syrup can be compared against the expected density at a given concentration and temperature; a discrepancy suggests either dilution or substitution of a non-sucrose sweetener, a preliminary finding that the FSSAI (Food Safety and Standards Authority of India) instructs its accredited laboratories to confirm by polarimetry and HPLC. US FDA forensic food-testing protocols reference oscillating-U-tube densitometry for the same purpose.
One significant limitation: the oscillating-U-tube instrument measures the density of the liquid filling the U-tube, not a solid fragment. It cannot directly measure the density of a glass shard. Some laboratories dissolve small glass fragments in hydrofluoric acid and then measure the density of the resulting solution, but this approach is not a direct comparison with fragment density and requires correction for the acid contribution. For solid glass fragments, the gradient column and hydrostatic weighing remain preferred.
The Anton Paar DMA 4100 M is the instrument most commonly cited in peer-reviewed forensic density work; the Mettler-Toledo Densito 30PX is a portable variant used in the field for liquid samples. Both instruments appear in equipment lists for US FDA-registered food-testing laboratories, and both are mentioned in the FSSAI laboratory accreditation technical guidelines for food analysis.
*No forensic scientist calls a glass match on density alone, but density is always the first filter.*
Glass comparison follows a hierarchy of discrimination. Physical features (colour, surface texture, thickness, fluorescence under UV) are assessed first. Density by gradient column provides the next filter. Refractive index (covered in the next topic in this module) is the primary discriminator in most casework. Elemental fingerprinting by LA-ICP-MS (Laser Ablation Inductively Coupled Plasma Mass Spectrometry, covered in Module 5) is the final, most discriminating layer.
The ENFSI ENG3-2013 guideline, the FBI Glass Examination Unit protocol (published in the SWGMAT guidelines), and the ASTM E2927 glass-comparison standard all structure casework as a sequential filter. Density by gradient column is used to screen out clear non-matches before the more time-consuming refractive-index measurement is performed. If density is measured on both a questioned fragment from a suspect's clothing and a control fragment from the crime-scene window, and the gradient column shows the fragments banding at the same depth within the column's precision, the analyst proceeds to RI. If they band at different depths by more than the analytical uncertainty, the analyst may exclude on density alone without performing RI measurement.
The probability of coincidental density match between two unrelated glass sources depends on the density distribution of the glass population. For ordinary soda-lime float glass, the population density has been estimated from several large databases to have a standard deviation of approximately 0.0025 g/cm³ around the mean of 2.500 g/cm³ (ASTM E2927 Appendix; FBI proficiency-trial data published in the Journal of Forensic Sciences, 1994). This means the density measurement alone provides moderate but not strong discrimination. The combination of density and refractive index reduces the coincidental-match probability substantially, which is why the ENFSI guideline requires both measurements before a comparison opinion is reached.
In an Indian casework context, glass comparison under the DFSS guidelines proceeds through the same filter hierarchy, with density by gradient column as standard. The CFSL (Central Forensic Science Laboratory) New Delhi has published its methodology in the CFSL Bulletin (2008, Vol. 58), confirming that bromoform-based density gradient is the primary physical comparison step. Australia's ANZFSS Guidelines for Forensic Glass Examination similarly mandate density and RI as the minimum comparison parameters.
*For polymers, density does more of the discriminating work than it does for glass, because the density spread between common polymer types is larger.*
The density ranges of common polymers are far more spread than those of glass types, making density a stronger first-pass discriminator in polymer comparison casework. Polyethylene (LDPE: 0.910-0.940 g/cm³; HDPE: 0.941-0.965 g/cm³), polypropylene (0.895-0.920 g/cm³), polyvinyl chloride (1.16-1.58 g/cm³), polyethylene terephthalate (PET, 1.29-1.40 g/cm³), and polycarbonate (1.18-1.24 g/cm³) occupy distinct, largely non-overlapping density ranges. A fragment of translucent packaging material found at a crime scene can be placed in a bromoform-bromobenzene gradient column, and the density band position often immediately identifies the polymer class before any spectroscopic analysis is performed.
For pharmaceutical-package tampering and food-fraud cases, density provides evidence that a package component has been substituted with a different-class polymer. A PET bottle whose cap is polypropylene is as manufactured; a PET bottle whose body shard has the density of polycarbonate has been replaced or adulterated. This preliminary finding would then be confirmed by FTIR spectroscopy, but the density measurement is faster and consumes no reagent.
The SWGMAT guidelines (now under OSAC auspices in the US) and the ENFSI polymer evidence guidelines both list density as the primary physical characterisation step for polymer fragments. In the UK, the Forensic Science Regulator's notes on polymer evidence (issued 2016) indicate that density measurement by gradient column or pycnometry should precede chemical characterisation. Canadian RCMP protocols for packaging-fraud casework mirror this sequence.
| Method | Output type | Precision (typical) | Fragment size limit | Key standard |
|---|---|---|---|---|
| Sink-float test | Qualitative (match/exclude) | ±0.01 g/cm³ | Above ~1 mg | ASTM E2365; DFSS Lab Manual |
| Density-gradient column | Semi-quantitative band position | ±0.001 g/cm³ | Above ~0.1 mg (surface-tension limit) | ENFSI ENG3-2013; FBI SWGMAT |
| Pycnometry | Absolute density value | ±0.002 g/cm³ | Above ~500 mg preferred | ASTM E2365; NIST SRM 8092 calibration |
| Hydrostatic weighing | Absolute density value | ±0.001 g/cm³ | Above ~100 mg | ASTM E2365; Mettler-Toledo kit protocol |
| Oscillating-U-tube | Absolute density (liquids) | ±0.0001 g/cm³ | Liquids only (or dissolved solids) | ISO 15212-1; Anton Paar DMA series |
A glass fragment from a suspect's jacket is placed in a density-gradient column at 25 °C. It bands at a depth corresponding to a density of 2.490 g/cm³. A control fragment from the crime-scene window bands at 2.493 g/cm³. The column's analytical uncertainty is ±0.001 g/cm³. The correct interpretation is:
Test yourself on Forensic Physics with free, timed mocks.
Practice Forensic Physics questions