I am not sure what the reason is for your yttria depletion. But if it is removed e.g. by any type of chemical transport, there is not reason to stop. This means you run through the tetragonal phase field and the 2-phase field "M+T" into the monoclinic field, until you have monoclinic ZrO2 only. I do not know literature on this specific system; but this is just from a general consideration of phase equilibria. But another point is, whether your phase diagram is really correct!? In "Phase Diagrams for Ceramists" I found a very recent one (see below), the original reference is: O. Fabrichnaya, Ch. Wang, M. Zinkevich, C. G. Levi, and F. Aldinger, J. Phase Equilib. Diff., 26 [6] 591-604 (2005). And according to this one you never reach the "monoclinic only" phase field below 1000°C.

I am not sure what the reason is for your yttria depletion. But if it is removed e.g. by any type of chemical transport, there is not reason to stop. This means you run through the tetragonal phase field and the 2-phase field "M+T" into the monoclinic field, until you have monoclinic ZrO2 only. I do not know literature on this specific system; but this is just from a general consideration of phase equilibria. But another point is, whether your phase diagram is really correct!? In "Phase Diagrams for Ceramists" I found a very recent one (see below), the original reference is: O. Fabrichnaya, Ch. Wang, M. Zinkevich, C. G. Levi, and F. Aldinger, J. Phase Equilib. Diff., 26 [6] 591-604 (2005). And according to this one you never reach the "monoclinic only" phase field below 1000°C.

The term yttria depletion is midleading, it implies that the yttrium actually disappears. The current opinion is that in contact with water the yttrium oxide kept in solid solution is rather hydrolysed to form yttrium hydroxide. Thus it cannot stabilize the oxygen vacancies keeping the zirconia tetragonal any more. The procees proceeds starting from the grain boundaries. Probably this process happens in both cubic and tetragonal phase but as the cubic phase contains much more yttria than the tetragonal and as the lattice constants of cubic and tetragonal are quite similar only the loss of yttria from the tetragonal will contribute to the ageing phenomenon.

A further problem comes from the fact that coprecipitated 3Y-TZP zirconia contains an excess of ~20-25 % of yttria over the amount present at the t/t+c boundary this excess initially does not form cubic phase but is kept in the lattice provided that the sintering temperature is low. Only upon prolonged thermal treatment excessive phase separation occurs. therefore ageing is much less pronounced in Y-TZP sintered at low temperatures.

In case of TZP from yttria coated powders having a gradient structure the ageing is much less detrimental.

(see: Matsui, J Am Ceram Soc 2009 and related works; Burger J. Mat. Sci. 1997, Ohnishi, J Ceram Soc. Jap. 2008)

It depends what is the driving force for yttria depletion. At high temperature, the metastable tetragonal phase decomposes into a rich cubic and poor tetragonal/monoclinic phase. In this case the end compositons are on the T/(T+F) and (T+F)/F separation lines at high temperature.

There are quite a few processes that can lead to yttria depletion at high temperature, with vanadates forming YVO4, with sulfates forming Y2(SO4)3, possibly also by reaction with fluorine forming YF3, in these cases the yttria / zirconia phase diagram is no more relevant because of the presence of other elements that have to be taken inot account.

I am not sure if we are talking about he same subject.

In terms of structural ceramics i was thinking about the low temperature degradation phenomenon (at ~ 30-40°C) which leads to the formation of monoclinic phase from tetragonal in Y-TZP of typically 3 mol-% Y2O3 stabilizer. this is something that happens to e.g. Y-TZP in implants for hip, knee or dental and finally has led to the ban of Y_TZP from Hip-implants and the shift to zirconia toughened alumina materials where this effect is much less pronounced. Ce-TZP which should not contain many (if any) oxygen vacancies shows ageing but at much slower extent. Thermodynamic considerations do not help very much, this is a question of kinetics. The typical compositions (~3mol-% Y2O3) are thermodynamically unstable. the metastability of the t-phase at ambient temperature is the reason for the exceptional toughness of partially stabilized zirconia materials (transformation toughening). As there is at low temperature definitely no yttria redistribution by diffusion a composition m+f will never be reached. Most probable hydrolyzed yttria is still there and not transported away but becomes useless and only monoclinic is formed, the only cubic phase that can be found is the one that was already there before the sample was cooled down in the furnace after sintering.

It is known that e.g. alumina addition drastically reduces ageing and that the aluminum ions are incorporated into the grain boundaries replacing yttria. Alumina is much more difficult to hydrolyze and can stabilize the t-phase in the boundary more efficiently than yttria (again kinetics !!).

(Extensive literature on this subject e.g. by Jerome Chevalier)

Yttria stabilized zirconia materials with ~ 5 mol-% yttria which are used in e.g. gas turbine thermal barrier coatings are long term stable the t'-phase does not transform. Here chemical attack by compounds contained e.g. in the engine fuel may lead to yttria depletion (see Gregoire's comments).

@JP Xu which penomenon exactly are you referring to ?