In High Performance Liquid Chromatography (HPLC), the actual gradient is a bit different from the true gradient programmed because of dwell volume and other effects. For example, in the figure added, we can see the programmed gradient in red and the actual gradient in blue.

I'm wondering if there are formulas I can use to approximate/estimate the actual gradient from my programmed gradient or gradient steps (given dwell volume, flow-rate, etc...)?

Approximations are ok, I don't expect this to be computable exactly. Any suggestions are welcome.

Programmed gradient (red) vs actual gradient (blue)

Figure just to illustrate the real gradient being different. source

  • 3
    $\begingroup$ There is no simple formula to calculate the actual gradients, as you can see yourself. The functional output is quite complicated (RC filter like output). I am sure if someone had a very detailed knowledge of fluid mechanics they could have easily predicted that. However there are empirical ways: Search this title. “Measure Your Gradient”: A New Way to Measure Gradients in High Performance Liquid Chromatography by Mass Spectrometric or Absorbance Detection $\endgroup$
    – AChem
    May 13, 2023 at 13:50
  • 2
    $\begingroup$ I assume the exponential response with transport delay is good enough. $\endgroup$
    – Poutnik
    May 13, 2023 at 14:40

1 Answer 1


Based on the comments of @AChem and @Poutnik the "exponential response with transport delay" seems to be ok for my use use-case.

The Paper “Measure Your Gradient”: A New Way to Measure Gradients in High Performance Liquid Chromatography by Mass Spectrometric or Absorbance Detection, cited by @AChem. Summarizes it nicely:

The gradient produced by an HPLC is never the same as the one it is programmed to produce. Gradient non-idealities are usually categorized into three types:

  • Gradient delay time (dwell time)
    • Delay from the time the gradient is programmed to be produced to when it actually reaches the point where the sample is injected.
  • Gradient dispersion
    • Gradient dispersion is the rounding out of the gradient, resulting in more gradual changes in slope as if a low-pass filter were applied to the gradient profile.
  • Solvent misproportioning
    • Any other gradient non-ideality that is not described by the former two categories

Gradient delay and dispersion

So I ended up modelling:

  • Gradient dispersion with exponential smoothing ("exponential response" mentioned by @Poutnik, "RC filter" mentioned by @AChem)
  • Gradient delay by adding a linear lag to my time domain ("transport delay" mentioned by @Poutnik)
  • 1
    $\begingroup$ it is good to know that you were able to model the gradient! Very nice work. $\endgroup$
    – AChem
    May 17, 2023 at 18:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.