You have a nice set of sensorgrams there. Nice Rmax.
You state that the analyte is an antibody from hybridoma's. Although in some situations this can be fit by a 1:1 interaction an antibody should be fitted with a bivalent model.
Comments on the sensorgrams:
P1S1 --> no binding / bad curve, do not analyze the dissociation
P1S17 --> as said the bivalent model will probably fit better
P1S25 --> nice
P1S33 --> the fitting algorith got stuck: try different initial values for the (bivalent) fitting
P1S49 --> as P1S33
As an additional remark: for a better estimation of the kinetic constants I would do an analyte dilution series in stead of a single curve.
You are right, in your set-up a 1:1 interaction is the model to use.
Second thing I noticed that your curves are red and the fit is black, my mistake that I did not see that earlier.
What are you using as a reference? Is this the anti-mouse IgG Fc alone, without hybridoma supernatant? Maybe there is some background binding to the capture antibody that can explain the non-ideal curves.
What also is missing in your post is the amount of hybridoma IgG captured and the size (Mw) of the analyte. I can imagine that different amounts of antibody are captured between different clones. If you have a high density surface then the dissociation may suffer from rebinding of the analyte during dissociation (e.g. P1S17 and P1S49).
P1S33 has a very straight association, possibly mass transfer during association but the second part of the association is also straight. In addition the dissociation is strange by the fact that it first looks fast and then stops. No real good explanation for this curve. Could this maybe not be a true monoclonal?