A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Calibration relates the instrument signal to analyte concentration using prepared standards. The external standard method builds a calibration curve from independently prepared standards and reads unknown concentrations by interpolation; it assumes the sample matrix does not affect the response. Standard addition overcomes matrix effects by spiking known amounts of analyte into the sample itself. Internal standards — chemically similar compounds added at a constant concentration — correct for instrumental drift and variable injection volumes in chromatography. Limits of detection (LOD) and quantification (LOQ) are derived from the calibration regression statistics.
Determine a metal concentration in a complex environmental water sample using all three calibration approaches and compare results. Observing that external and standard addition methods disagree (but standard addition is reliable) makes matrix effects tangible.
Every analytical instrument converts a physical property — absorbance, current, ion count — into a signal. Calibration is the process of translating that signal back into a concentration. The core procedure is always the same: prepare solutions of known concentration (standards), measure their signals, fit a line through the data, and use that line to predict unknown concentrations by interpolation. The differences between calibration strategies come down to controlling specific sources of error that the basic approach cannot handle.
The external standard method is the default. You prepare a series of standards in a clean solvent, build a calibration curve, and read off unknown concentrations. It is fast and simple but rests on a critical assumption: the sample and the standards behave identically in the instrument. When that assumption breaks down — because the sample contains dissolved salts, organic matter, or other species that suppress or enhance the analyte signal — you get systematic error. This is the matrix effect, and it is the central practical challenge in real-world analytical chemistry.
Standard addition is the remedy for matrix effects. Instead of comparing your sample to standards prepared in clean solvent, you spike known quantities of the analyte directly into your sample. Because the spiked analyte and the native analyte sit in the same matrix, both experience the same suppression or enhancement. The signal increases linearly with the amount spiked; extrapolating that line back to zero signal gives the original concentration. The trade-off is more sample and more measurements per unknown, so standard addition is reserved for cases where matrix effects are confirmed to be significant.
Internal standards solve a different problem: random instrumental variation that causes signals to drift from injection to injection even at the same concentration. In chromatography, for example, injection volume can vary slightly between runs. Adding a fixed amount of a chemically similar compound (the internal standard) to every sample and every calibration standard means it fluctuates by the same factor as the analyte. Dividing the analyte signal by the internal standard signal cancels that factor, producing a ratio that is stable even when absolute signals are not.
Limits of detection and quantification are often misunderstood. The LOD (3σ/slope) and LOQ (typically 10σ/slope) are calculated from blank precision and calibration sensitivity — they are statistical properties of the method, not arbitrary choices about the range of the calibration curve. A method can have a low LOD with only five calibration points, or a high LOD with twenty, depending on the instrument noise and the steepness of the calibration slope.