Chromatography Forum

Hello,
Iam currently involved in developement of a gradient method for analysis of a drug substance . I found that varying the organic solvent (M.phase B) from 100% Methanol to 4:1 (Methanol:THF) gives good change in selectivity with appreciable seperation for critical pair of peaks, but i observed that in the blank injected using this chromatographic condition there is appreciable drift (upwards) as the organic concentration increases during the gradient and the working Wavelength is above 260nm. My question is whether THF can be used in gradient elution and whether this upward drift observed can be minimised (or Absorbance due to THF can be matched) by addition of any additive to M.phase A.
Thank You,
Santosh Gandhi.

What is the mobile Phase A composition?

What is the sample injection solvent?


Currently, we are trying to "guess the picture on a jigsaw" with only a coule of pieces available.

I have read in the past that fresh THF has lower UV cutoff than old THF.

Is your THF batch recently purchased / opened or it has been used for a longer time in the past?

Regards

Check whether your THF contains BHT inhibitor. This will certainly cause significant absorbance at much higher wavelengths than THF itself. Spectroscopic or chromatographic grades of THF should be inhibitor free. BTW, the listed UV cutoff for THF (1AU!) is 215nm while for methanol it is 205nm.

At 260 nm, you're well above the UV cutoff. If the drift is non-linear, you may be seeing a shift in refractive index (to a certain extent, it's normal, but if it's excessive, it could be a symptom of an out-of-alignment flow cell in the detector).

THF will probably still be absorbing at 260nm, about 0.1AU, as it has a very broad absorption spectrum. I'd expect you to see a baseline shift of about 20 mAU, which is quite significant. Methanol has a much sharper cutoff with about 0.1AU at 230nm, and 0.01AU at 260nm...

It is not a good idea to try an flatten the baseline drift by increasing the absorbance of Mobile Phase component A, as that will mask impurities and minor peaks, as well as making the method difficult to reproduce.

The solution may be to subtract a blank baseline, but I'd actually just live with the drift, possibly increasing the sample concentration/injection volumes if you really want a visually flat baseline.

Bruce Hamilton

just try to be isocsratic