When the External Method Lets You Down
The external standard method rests on one fragile promise: that the exact amount you measure of the sample is the same as the amount you measured of each standard. For a clean pipetted volume into a spectrophotometer, fine. But picture injecting a microscopic droplet into a gas chromatograph by hand — the volume jitters from injection to injection. If your standard's injection was a hair larger than your sample's, the unknown looks falsely dilute, and nothing in the data warns you.
The same trouble appears whenever the sample passes through messy preparation steps — extraction, evaporation, filtering — that might lose a little of your analyte along the way. The signal shrinks, but you have no idea by how much. External standards, measured separately, never experienced those losses, so they can't account for them.
Adding a Companion: The Internal Standard
The clever fix is to add a known, fixed amount of a second substance — an internal standard — to every sample and every standard alike. You pick a companion that behaves much like your analyte but can be told apart from it (a different peak, a different wavelength). Now, whatever misfortune befalls the sample, it befalls the companion too. Inject a smaller droplet, and both signals shrink together. Lose 10% during extraction, and you lose 10% of both.
Here is the key move: instead of reading the analyte's signal alone, you look at the ratio of the analyte's signal to the internal standard's signal. Because both rise and fall together, the ratio stays steady no matter how much was injected or lost. You have converted a slippery absolute measurement into a robust relative one.
The Response Factor: Comparing Apples to Oranges Fairly
There is a catch: the instrument rarely responds equally to your analyte and your companion. Maybe one molecule gives twice the peak the other does at the same concentration. So a raw signal ratio of 1 doesn't mean equal amounts. You need a conversion factor that captures "how much more (or less) signal the analyte gives per unit, compared with the internal standard." That number is the response factor.
You measure the response factor once, from standards that contain known amounts of both the analyte and the internal standard. With it in hand, the calculation for an unknown is clean: measure the signal ratio, correct it with the response factor, and multiply by the known internal-standard amount to recover the analyte amount.
- Add the same known amount of internal standard to every standard and every sample.
- From the standards, measure the response factor — how the analyte's signal compares to the internal standard's, per unit.
- For the unknown, record the ratio of analyte signal to internal-standard signal.
- Combine that ratio, the response factor, and the known internal-standard amount to get the analyte concentration.
What It Fixes, and What It Doesn't
Internal standards shine against problems that hit the analyte and the companion equally: injection-volume scatter, partial losses during prep, gradual signal drift between morning and afternoon. Because both ride the same boat, the ratio cancels the wobble. This is also how you can honestly report recovery — how much of what you started with actually survived to the detector.
But be honest about its blind spot. If something affects only the analyte and not the companion — a contaminant that overlaps just your analyte's peak — the ratio won't save you, because the two no longer move together. An internal standard corrects for shared troubles, not selective ones. Choosing a companion that genuinely shadows the analyte is therefore everything.