Negative peaks like the one in the image are always present when samples (and mobile phase as blank) are injected into columns with cross-linked polystyrene-divinylbenzene matrix. RI is used as a detector. It always happens at the same times according to the method; it has varied temperature, flow and the peaks run equally. Dirt from the columns has already been discarded. The columns were tested in another HPLC using the same method and in this other equipment the negative peaks do not appear, therefore damage to the columns is ruled out. It is also not about the initial injection peak.
The methods are operated correctly, parameters and suitable mobile phase (deionized water).
Thanks!!