Chromeleon - manual integration / split peak

[Woodstock]

Hi everyone,

I'm hoping to get some help from here for my problem. I'm working with Chromeleon 6.70 and I usually integrate my chromatograms manually. Works fine. But: When I insert a 'split peak', the baseline gets somehow 'kinked' where the peak split is. This is somewhat annoying, especially if you split your peak a dozen times in one chromatogram and you need to straighten your baseline the same number of times...

The manual states that splitting peaks does NOT change the baseline - are there any hidden settings I can or need to change? Im kind of lost here, any help is appreciated.
Thanks in advance,
Woodstock

[dblux_]

Hi everyone,

...
The manual states that splitting peaks does NOT change the baseline -
...

Peak splitting doesn't change baseline.

Try to check whether baseline is 'kinked' with higher zoom (may be it's an optical distorsion of your monitor ??).

[unmgvar]

you can probably solve this using peak shoulder variable and the correct sensitivity
also there is the peak rider ratio that can be used
if the peaks are named you can also play around the baseline settings in the peak table

send me a CMD i can look at it in the week-end

[WillyOne]

You could use "lock baseline" integration event (Chromeleon) from before starting of the peaks to after end of splitted peaks.
Chromeleon will integrate at the time of the valley point but to the baseline level.
Hope it helps

[DR]

Or you could strive to improve the method (good data ssytems are no match for bad chromatography)

[Woodstock]

Thanks a lot for the suggestions (the lock baseline parameter didn't help though), I think I found a workaround - if you don't change the start or endpoint of your baseline after inserting peak splits, your baseline is just fine. If you do change e.g. the start point, you get kinks in the baseline where your peaks are split.

@DR: True, indeed. If you could see my chromatograms, you would run away anyway, I guess. It's protein analytics, and not much room for improvement here...

[unmgvar]

you are doing cation exchange?

[Woodstock]

I am indeed.

[unmgvar]

what type of column?
what type of gradient?

we are currently working on 3 things to improve separations of this sort especially for Mabs
1. we are trying more pH gradients than buffer gradients. we see also that the use of Na or Li can influence the results. we are still trying to figure out why. we saw at pittcon that others are also doing the same as we in this manner
2. we now screen 3 type of columns- WCX, antibodix from Sepax which is a modified WCX and also SCX
3. we go from 10u to 5u, more pressure but the system can handle it. and in some cases it is simply worth it.
the problem is that there are not too many vendors doing the 5u columns right now

[wormwoody]

Have you tried setting the QNT detection parameter to valley to valley?

[Woodstock]

unmgvar, thanks for sharing your experiences.
I work with WCX colums (Dionex), the gradient increasing the salt concentration (NaCl) and with that lowering the pH, but only slightly.
I'd love to try a lot of things to improve that method, but unfortunately we simply don't have the time to play around much further. It's all completely GMP-regulated, and no one seems to see an urgency for method optimisation when the FDA approves the results obtained so far... It's a disgrace.

wormwoody, yes, valley to valley is set in the QNT.

[unmgvar]

woodstock

i have seen it several times already. FDA says nothing about the chromatography until the end of phase 3. then suddenly you get a notice saying:"improve the application"

also in your case if the blank does not show a change on the base line then you cannot do valley to valley that follows the gradient in the run. you must do drop downs. this is also something i have seen too often. in one case the reason was "very logical". the separation is very bad and we want to give a resolution number. so valley to valley get's the job done. but the integration is completely wrong.