MAI-Alchemy · Analytical Chemistry, Taught by Practice

The Analytical Column

where ideas resolve
Calibration · No. 17The Detector

Calibration — the hardware we calibrate

The detector is the constant

Of all the hardware, the detector is the most tightly specified and tested part of an instrument. Overlay one method’s calibrations and the shape holds while the level moves, so when calibration is lost, look upstream first.

By Medrado Analytical Innovations · Foundations, No. 17 · for the working analyst

In this piece · 8 min read
  1. A detector is built to be trusted
  2. What a detector's specifications mean
  3. How detectors are tested
  4. What the price buys
  5. Reading the stack of calibrations
  6. Calibrate less, verify more
  7. When calibration is lost: grab the detective hat
  8. The judgement

Pull up every calibration ever run for one method on one of your instruments and lay them on top of each other. Most analysts never do this; a calibration is judged once, when it is made, and then replaced by the next. But the stack tells a story a single curve cannot. The curves are not identical, and they are not random either. Their shape barely changes. What moves is their level. That pattern says something important about where an instrument's variation comes from, and it starts with the part whose signal we actually calibrate: the detector.

This piece is written for the detectors where the argument is cleanest: a flame ionization detector (FID) on a gas chromatograph, and a UV detector on a liquid chromatograph. Mass spectrometers and some selective detectors behave differently, and are noted where it matters.

A detector is built to be trusted

An instrument maker cannot sell a detector that cannot be trusted. Everything the instrument reports passes through it, so it is the component that is specified, tested and documented most carefully.

Check yourself

Answer in your head first, then open the answer. Any question can go into your Quiz me.

  1. Why can a detector's noise, drift and linear range be known before it reaches your lab?

    Show the answer
    “Detector test practices measure the detector on its own, separate from the column and the inlet, so its noise, drift, sensitivity and linear range are characterised before it reaches your lab.”

    See it in the article ·

  2. Overlay one method's calibrations on log-log axes. What sets the slope, and what sets the level?

    Show the answer
    “On log–log axes the detector's response law sets the slope, and everything upstream sets the level.”

    See it in the article ·

  3. The morning check standard passes. What should you do with the calibration?

    Show the answer
    “A check that passes is verification, not a reason to recalibrate.”

    See it in the article ·

Glossary

Noise
How much the baseline flickers when nothing is eluting.
Drift
How far the baseline wanders over time, usually stated per hour.
Minimum detectability
The smallest amount of analyte the detector can still see above its noise.
Operational qualification
Testing the instrument as installed, and at intervals, against stated limits; detector noise and drift are one part of it.

Sources

  1. ASTM E594-96(2026), Standard Practice for Testing Flame Ionization Detectors Used in Gas or Supercritical Fluid Chromatography (store.astm.org).
  2. ASTM E516-95a(2021), Standard Practice for Testing Thermal Conductivity Detectors Used in Gas Chromatography; ASTM E697, Standard Practice for Use of Electron-Capture Detectors in Gas Chromatography; ASTM E840, Standard Practice for Using Flame Photometric Detectors in Gas Chromatography.
  3. ASTM E685-93(2021), Standard Practice for Testing Fixed-Wavelength Photometric Detectors Used in Liquid Chromatography; ASTM E1657, Standard Practice for Testing Variable-Wavelength Photometric Detectors Used in Liquid Chromatography.
  4. USP General Chapter ⟨1058⟩ Analytical Instrument Qualification.
  5. Agilent CrossLab, GC Hardware Operational Qualification, test list and limits (https://www.agilent.com/cs/library/service/Public/GC%20Hardware_Operational%20Qual-Col2.pdf).
  6. EDQM / European OMCL Network, PA/PH/OMCL (16) 17 R3, Qualification of Equipment, Annex 2: Qualification of GC Equipment (FID noise and drift check).
  7. JCGM 200:2012, International Vocabulary of Metrology (VIM), 2.39 calibration and its notes; 2.44 verification.
  8. LCGC, "Ongoing care and feeding of your gas chromatograph: preventing problems before they start" (chromatographyonline.com/view/ongoing-care-and-feeding-of-your-gas-chromatograph-preventing-problems-before-they-start).
  9. M. Leal, "Why Sample Introduction Matters", "Why the Sample Point Matters" and "Why Variability Matters", Analytically Speaking, The Analytical Column (column.mai-alchemy.com).