suppose my mobile phase is acetate buffer pH 4 : acetonitrile (30: 70) . in this condition degradation is seen in acidic and basic condition. But in H2O2 and in thermal condition there is reduction in peak area. We have tried other wavelength but no degradation peak is seen. We may say that may be in other mobile phase degradation may be seen. How can I try other mobile phase and how this can be applied in gradient.
Good think to do is perform at least spectral purity of the peak. Different detection principle ideally MS (it should be compatible with mobile phase) - to have different type of detection to discover potentially not UV/fluorescence active substances (you had not mentioned the detection type). The change in gradient can help in separating peaks but will not help with detection problems (that's why the peak purity is interesting). Ideal is to confront method development documentation. If you want to try the alterations you get to sphere of method development - so try to read basics - changing the mobile phase composition, pH, column, temperature all can lead to better separation. And separation of individual impurities and the main peak is what you want in stability indicating methods. The available potential degradation products are big help in method development.
You are pumping more organic 70% from intial run, the peroxide degradation may forming highly polar species that will not seperate in 70% ACN , try gradient , choose different column chemistry , may be helpful. Otherwise your degardent may be non-chromophoric so ELSD or CAD detection can be used.