Why does my baseline drop?

[Mike H.]

I would like to know exactly what's going on, but in the end I'll probably go with Tom's approach and select another column. FWIW the shark fin occurs in blank injections, 0 volume injections and in each of the four compounds I was trying to separate. It's too much hassle to develop a separation where the chromatography is atypical.

[danko]

I would like to know exactly what's going on

You can come closer to that state if you run the gradient without a column, as HW Mueller suggested.
If nothing else the bleeding notion will be dismissed - I presume.

Picking up another column will help but make sure it's a conventional reversed phase column (i.e. no polar groups embedded)

Best Regards

[tom jupille]

shark fin occurs in blank injections, 0 volume injections and in each of the four compounds I was trying to separate.

That pretty much confirms that it's bleed!

danko: you are correct about the relationship between refractive index and wavelength, but the effect on baseline does seem more pronounced at longer wavelength. Rather than hijack this thread, I'll start a new one on the topic.

[Gilson_PLC2020]

Most effects happen more at shorter wavelength. Because for a given light path length a shorter wavelength having consequently a higher frequency will have more complete waves. Or put another way, it has shorter waves, hence more of them for the light path length.

An effect will occur per complete wave, so the effect will be more pronounced where there are more waves. Given that the path length remains constant through the detector no matter what wavelength is chosen. Grating angles should add minimal if any length.

I've been working with the remote flow cell option for the Gilson 171 and 172 Diode Array Detectors which uses fibre optic cable to connect the flow cell to the detector box. Light loss and therefore detector sensitivity is greatest at the lower wavelengths, so the maximum 'usable' optical cable length is dependent on the wavelength used.