I believe you are asking if you can switch buffers instantaneously so that you can monitor association of analyte and ligand in one buffer and monitor dissociation in another buffer. If so, then the answer is yes, I have done this many times. The simplest way to do this is to use whatever buffer you want to have for the dissociation phase of the experiment as the running buffer, and inject your analyte using whatever the second buffer is that you are interested in. The instrument/software will not know there are different buffers and will treat this like any other kinetic type injection. You can also take advantage of the 4 different buffer lines if you want to try various buffer combinations in the same experiment.
This sort of experimental design works particularly well when using the same buffer components, but changing pH of the buffer from association to dissociation. The same can be done with different buffering agents as well as different salt concentrations. Keep in mind with large changes in salt concentrations, you will have some bulk refractive index (RI) changes to deal with, but they can largely be subtracted out via standard reference subtraction in the software. I would caution against doing this with different DMSO concentrations as the RI changes can be massive and difficult to subtract.
Finally, keep in mind that this sort of assay can be useful for many reasons, but I would not recommend reporting KD's from this type of experiment. The KD is simply the kd/ka, and since you would have an association/dissociation rate during sample injection with one buffer (keep in mind both binding AND dissociation are happening during sample injection) but dissociation of analyte from ligand potentially at a different rate during the dissociation phase of the experiment, it would not be suitable to report a KD from that experiment.
In my experience, reporting off-rate differences and also overlaying sensorgrams to show qualitative differences between different buffers or different interactions can be particularly impactful with this type of experimental design.
Yes, in addition to the below answers you can use the "dual inject" function in the T200 software. I use this, for example, when doing FcRn binding experiments. I will use PBS pH 7.4 as the running buffer and PBS 6.0 as the injection buffer (proteins diluted into PBS pH6).
Using the dual injection I can have PBS 7.4 running though the system, then it injects FcRn which is diluted into PBS pH 6.0 as the first injection, the second injection is PBS 6.0 alone which will be the dissociation of the FcRn at pH6, after this "dissociation" injection the buffer will switch to the running buffer of pH7.4 and the FcRn will dissociate rapidly.