Random sensorgram drift in dissociation phase

Hello, Im SPR user and I have problem in my sensorgram.

I performed SPR analysis (Receptor-Ligand binding) during several weeks, and I have difficult problem in my experiment.

This is my experiment condition

Instrument : Biacore T200
Chip type : CM5
Immobilization method : Amine-coupling
expected KD : 1~5nM
Binding MOA : 1:1 binding
Dilution factor : 1/2 serial dilution, 5nM concentration

Below is summary of my problem.

1. In dissociation phase, sensorgram drift was occured randomly. Sensorgram drift means that random sensorgram drops below the sensorgram of low concentration at dissociation phase (often it go down to the baseline) .
I think sensorgram of association phase (Analyte injection step) is perfect so I don't know why this phenomenon occured.

2. To confirm stability of experiment, samples of the same concentration are analyzed at the front and at the last. In this case, the result of the later analyzed sample always has a higher RU than the result of the first analyzed sample. I think that the results of the samples analyzed later should be the same or lower because of loss of ligand activity.

The baseline (Fc 2) appears to be maintained in the range of 1 to 5 RU, and there is no non-specific signal on the reference channel.

I read that the small air bubble could affect sensorgram drift at this homepage, so I tried to re-run the experiment by degassing the running buffer, but I did not see much effect.

The fitting of the sensorgram does not seem to fit well at the dissociation phase for above two reasons. I am worried that KD values will be affected by these problems.

I would be grateful if you could give me some advice on why.

I always appreciate you kindness,

Thanks for reading,

I have not a good idea why this is happening. Degassing is a good practise when you use this buffer for making the samples too. The flow buffer in the T200 goes through a degasser thus air bubbles should not be a problem but it is good you tried.

The analyte vials should be capped since uncapped vials could explain the increase in response of the last injections (buffer evaporation = higher concentration analyte = …).

Kind regards
Arnoud