3
$\begingroup$

I was wondering how the form and curvature of HPLC peaks come to be. As I understand it, the mobile phase which contains the analyte, runs through the column. Many times, the distribution equilibrium is established inside the column, so the analyte is held back in the column. Finally, in the detector, peaks like a gaussian curve like this are recorded:

enter image description here

What causes the form of a peak to look like a gauss curve? I suppose more analyte is detected at the halfway point of the peak, but why?

Improve this question
$\endgroup$
7
  • 4
    $\begingroup$ This must be described in many textbooks and internet sites focusing on chromatography, as it is foundation of all chromatographic techniques. // Simulate multistage extraction with shifting mobile phase steps. It turns out it follows binomial distribution, converging to Gauss curve. $\endgroup$
    – Poutnik
    Commented Apr 27, 2023 at 17:19
  • 1
    $\begingroup$ You are asking 2 different questions. "What causes the form of a peak to look like a gauss curve?" : as Poutnik said, this is well explained by many sources, but in essence it's down to diffusion. Imagine the initial sample as a very short cylinder on top of the column, which would result in a narrow ~rectangle when detected. As the solvent carries this cylinder through the column, the molecules in it also diffuse in all directions, resulting in a random distribution, which translates to a ~gaussian peak. $\endgroup$
    – user6376297
    Commented Apr 27, 2023 at 18:00
  • $\begingroup$ "I suppose more analyte is detected at the halfway point of the peak, but why?". That's an approximation, but is justified for the same reasons as above. If the shape resulting from the above mentioned processes is ~gaussian, the mode and mean ~coincide. The mean occurs at 'half time', and so ~does the highest concentration. $\endgroup$
    – user6376297
    Commented Apr 27, 2023 at 18:03
  • 2
    $\begingroup$ Even in this picture it's quite visible it's not Gauss curve. $\endgroup$
    – Mithoron
    Commented Apr 27, 2023 at 20:09
  • 1
    $\begingroup$ @Mithoron "Why to spoil a good story by the truth?" :-} We know the world is not ideal and there are always some effects not involved in education models. // There are 3 factors, Diffusion, prevalent at slow linear mobile phase speed, analyte transfer delay, prevalent at high speeds, and speed independent factor due dispersion of mobile phase passed distance. // Peak distorsion is often caused by nonlinear distribution between phases. $\endgroup$
    – Poutnik
    Commented Apr 28, 2023 at 7:02

0

Reset to default