Forensic chemistry

Forensic scientists use chemistry to analyse evidence from crime scenes. Many of their tests are ones you meet at school, carried out very carefully.

Chromatography separates mixtures. To check whether a threatening letter was written with a suspect's pen, ink from the letter and from the pen are spotted on chromatography paper and a solvent soaks up through them. Each dye travels its own distance, measured as

Matching R_f values in the same solvent suggest the same dyes.

Flame tests identify metal ions: a lilac flame points to potassium, a crimson one to lithium.

Fingerprints on paper can be revealed with ninhydrin, which reacts with amino acids in sweat to give a purple colour.

Some tests are presumptive: a quick colour change that hints at a drug or at blood. These can give false positives, so samples go to a laboratory for confirmatory tests on instruments such as mass spectrometers, which can identify tiny amounts of a substance.

Every piece of evidence is sealed, labelled and signed for each time it changes hands, so nobody can claim it was swapped or contaminated.

Presumptive colour tests

Colour tests are fast and portable but respond to a structural feature, not to one compound.

• Marquis reagent (methanal in concentrated sulfuric acid) turns purple-black with MDMA, orange-brown with amphetamine and purple-red with morphine or heroin.
• Scott test (cobalt(II) thiocyanate) gives a blue product with cocaine, but also with lidocaine, a common cutting agent.
• Kastle-Meyer test for blood: colourless reduced phenolphthalein is oxidised by hydrogen peroxide to the pink form, a reaction catalysed by the haem group of haemoglobin. Plant peroxidases, such as those in horseradish, can also trigger it.

Confirmation

Thin-layer chromatography compares a sample's R_f with a reference standard run on the same plate. Infrared spectroscopy shows functional groups (cocaine has strong ester C=O absorptions near 1700 to 1750 cm^{-1}), and the fingerprint region below 1500 cm^{-1} can be matched against a library.

GC-MS is the workhorse. The gas chromatograph separates the mixture and records each component's retention time; the mass spectrometer then gives its molecular ion and fragmentation pattern. A match on both retention time and spectrum against a standard is very strong evidence of identity.

Alcohol

Roadside breath devices use a fuel cell in which ethanol is oxidised at a platinum electrode, producing a current proportional to the ethanol present. Evidential instruments measure infrared absorption at more than one wavelength so that other compounds, such as propanone in a diabetic's breath, are not mistaken for ethanol. Blood samples are confirmed by headspace gas chromatography with an internal standard.

Justifying an identification

International guidance (SWGDRUG) ranks techniques by how much structural information they give. Category A methods (IR, mass spectrometry, NMR, Raman, X-ray diffraction) are the most discriminating; chromatographic methods are category B; colour tests are category C. A defensible drug identification uses a category A method plus at least one other, so a single false match cannot decide a case.

Why colour tests mislead

In concentrated acid, protonated methanal is a strong electrophile. In the Marquis test it condenses with electron-rich aromatic rings, such as the methylenedioxyphenyl ring of MDMA or the phenol ring of morphine, to give highly conjugated cations that absorb visible light. Any sufficiently electron-rich arene can do the same, so the colour reports a structural motif, not a molecule. The Scott test similarly responds to many protonated tertiary amines, including several local anaesthetics and antihistamines. Unconfirmed field tests have led to wrongful arrests.

GC-MS in practice

Electron ionisation at 70 eV gives reproducible fragmentation, so spectra can be searched against large libraries. Cocaine (M = 303) shows a base peak at m/z 82 from the tropane ring system and a strong ion at m/z 182 from loss of a benzoyloxy radical (303 - 121). Quantitation uses an internal standard, ideally a deuterated analogue that co-elutes but differs in mass. Because the analyte and standard suffer the same losses, the ratio of their peak areas cancels variations in recovery and injection volume.

Toxicology

Drugs are metabolised before sampling, so toxicologists target metabolites. Heroin is deacetylated within minutes to 6-monoacetylmorphine (6-MAM), then to morphine. Morphine alone could come from codeine, prescribed medicine or poppy seeds, but 6-MAM can only come from heroin. Cocaine hydrolyses to benzoylecgonine, detectable in urine for days. These polar metabolites are measured by LC-MS/MS: a precursor ion is selected, fragmented and a specific product ion monitored, which gives selectivity even in blood.

Fire debris

Debris is sealed in clean metal cans or special polymer bags. In the lab, the can is heated with an activated charcoal strip inside; volatile hydrocarbons adsorb onto it and are then eluted with a little solvent for GC-MS. Burning carpet and plastics make hydrocarbons of their own, so analysts use extracted ion profiles: m/z 57 and 71 for alkanes, 91 and 105 for alkylbenzenes, 128 for naphthalene. Petrol shows a distinctive cluster of C_3-alkylbenzenes (trimethylbenzenes and ethylmethylbenzenes) in ratios that pyrolysis rarely reproduces. ASTM E1618 sorts ignitable liquids into classes such as gasoline, petroleum distillates, isoparaffinic and aromatic products.