Increasing signal during dissociation

Hi everyone.
I am having some trouble assessing the kinetics of my mAbs against the antigen.
In a Biacore T200, I have immobilized FGF in a CM5 sensor chip (~800 RU) by amine chemistry. The kinetic titration sensorgrams have a very strange profile, especially during the dissociation phase (figure). All solutions (HBS-EP) are filtered and I have performed desorb regularly and before starting the experiments... I don't think there is a problem in regeneration, once the baseline is back to zero after the cycle. I'm starting to suspect the immobilized antigen, but as the mAbs, it is pure (one band in SDS-PAGE). Could someone help me?

Hi,
How is your reference channel doing? It looks like the reference is not stable and by subtracting the reference you introduce 'drift' in your antibody channel.
Check raw data first before subtracting a reference.
In addition, run a blank (buffer injections only) just like a normal experiment and perform a double referencing.

Kind regards
Arnoud

Thank you, Arnoud!
The reference channel seems stable (look at the figures). I will try the double referencing.
Warm regards

Did you match the buffer of the sample to the flow buffer? It seems that the injection puls at each next injection is more going down. This could add to the drift during dissociation.

Thank you
That was my problem: the buffer!
Warm regards