A sample band is injected at the top of a column packed with a stationary phase. A liquid mobile phase carries the mixture downward; each compound continually partitions between the mobile phase (moving) and stationary phase (stuck), so the fraction of time it spends "stuck" — its retention factor k — sets how slowly it migrates. A compound that prefers the stationary phase elutes late and narrow-looking on paper, but broadened by diffusion; one that prefers the mobile phase shoots straight through.
v(analyte) = v(mobile phase) / (1 + k)
k(reverse-phase) ∝ nonpolarity (C18 retains nonpolar longest)
k(normal-phase) ∝ polarity (silica retains polar longest)
k ∝ exp(−2.2·(eluent strength − 0.5))
- Reverse-phase vs normal-phase — swaps which end of the polarity scale is retained longest. C18 (nonpolar stationary phase) holds onto nonpolar solutes; bare silica (polar stationary phase) holds onto polar ones. Flip it and watch the elution order reverse.
- Eluent strength — a stronger (more organic, for reverse-phase) mobile phase competes harder for every analyte, shortening every retention time but compressing the peaks closer together — the core speed-vs-resolution trade-off in real HPLC method development.
- Flow rate — scales the pump speed uniformly; it changes how long the whole run takes without changing relative separation.
- Sample mixture — toggle which of the four analytes are present in the next injection.
The strip-chart below the column is the detector trace: each time a molecule reaches the bottom it adds one count to its analyte's peak, so narrow tight bands become tall sharp peaks and spread-out bands become broad shallow ones — exactly what a real UV or FID detector reports as compounds elute.