I have rarely seen neutralized silica in the manner you described used with success. Typically if the acidity of the silica gel is an issue due to peak broadening because it is interacting with basic groups on the molecule, you can use a low percentage of triethylamine in the eluant to improve the separation, but in your case, where the issue is decomposition on the column, that may not be helpful. It is easy enough to check - run a 2-D TLC. Cut a small square TLC plate, spot your sample in one corner, and elute it in one direction using an eluant with a small amount of triethylamine. Do the same with the eluant you used previously. Let the plates stand for a few minutes to dry, then put the plates in sideways and let them elute the other direction. Spots off of the diagonal indicate decomposition of material on the plates.
Usually when people find that silica gel is acidic enough to decompose their compounds and adding TEA to the eluant doesn't work, the switch is to neutral or basic alumina columns.