But this happens with TiO2 too, from which I don't expect the same gag as from iron oxides.
Also, oxygen evolves from the liquid, and I suppose the catalyst is immersed. Though, the pressure first increases, then drops, despite the catalyst looks inaccessible to the evolved oxygen.
Could something else, outside the liquid, scavenge the oxygen? A seal ring, a grease...?
Does the temperature drop between 115 kPa and 90 kPa, and enough to explain the pressure drop? Then, just a leak appearing only at 115 kPa and +65°C would suffice.