VisualChem The interactive chemistry textbook
中文

Blood alcohol by headspace gas chromatography with an internal standard

Separate ethanol from the 1-propanol internal standard on a polyethylene glycol capillary column, build an area-ratio calibration line from five calibrators, measure the ethanol in an unknown blood sample and decide whether it crosses the drink-driving and drunk-driving thresholds.

Difficulty ●●○ ≈ 25 min

Aim Determine the ethanol concentration (mg/100 mL) of an unknown blood sample by the internal-standard method, and compare the precision of calibrating with the ethanol area alone and with the area ratio.
PEG column 15 m · 70 °C · FID · internal standard 1-propanol

Blood alcohol by headspace GC: the internal-standard method

Every vial: 0.50 mL of blood (or calibrator) + 2.00 mL of 1-propanol internal-standard solution (0.100 g/L), sealed; the headspace gas is injected.

① Calibrators (ethanol, mg/100 mL)

② Unknown blood sample #1

Chromatogram of the last injection

No injection yet.

Ethanol peak area alone
050100150200Ethanol / (mg/100 mL)00.51
Area ratio A(ethanol)/A(propanol)
050100150200Ethanol / (mg/100 mL)00.050.1

At least 3 calibrator levels are needed to fit the line.

Retention times come from this unit’s GC model (PEG phase; retention indices ethanol 926, 1-propanol 1026, acetone 798, all estimates). Peak area = relative FID response (effective carbon number/molar mass: ethanol 1.4/46.07, propanol 2.4/60.10) × vial concentration × the injection’s volume factor; that factor is shared by every peak of one injection (6 % relative SD), plus 0.5 % independent integration scatter, all from a deterministic hash of the sample number. Headspace partitioning, matrix effects and evaporative loss of ethanol are not modelled.

Procedure

  1. Prepare the vials: into each headspace vial put 0.50 mL of calibrator (20–200 mg/100 mL ethanol) or blood and 2.00 mL of 1-propanol internal-standard solution (0.100 g/L), and seal at once.
  2. Inject each of the five calibrators and check the chromatograms: ethanol and 1-propanol must be fully resolved (and acetone, if present, must be clear of ethanol).
  3. Record the ethanol and propanol areas of every injection and compare two calibration plots: ethanol area alone, and the area ratio.
  4. Inject the unknown blood sample at least twice, read the ethanol concentration from the area-ratio line and report the mean.
  5. Reveal the true concentration and compute the relative error; then repeat with another blood sample.

What to notice

  • Peak areas differ by a few per cent between injections — the injection volume varies — while the ethanol-to-propanol ratio barely moves.
  • The calibration with ethanol area alone has a visibly lower r²; the area-ratio line is almost perfect.
  • Per gram, the FID responds to ethanol only about 76 % as strongly as to propanol, so the area ratio is not the concentration ratio — the response factor must be measured with calibrators.
  • Some blood samples contain acetone (in diabetic ketosis, for instance); on the polar column it elutes before ethanol and does not interfere.

In a real lab

  • Blood is a biohazard: wear gloves and dispose of waste as clinical waste; keep ethanol calibrators tightly sealed to avoid evaporative loss.
  • Headspace equilibration needs a fixed temperature (for example 15 min at 60 °C), identical for every sample and calibrator.
  • Forensic labs usually run two columns of different polarity in parallel and validate every batch with control and blank samples.
  • Report the measurement uncertainty; when a result is near the 20 or 80 mg/100 mL thresholds, the uncertainty decides the conclusion.

Related lessons

All experiments