[I] Firstly, about the very high nuclelation efficiency of TiO2/ZrO2:

The nucleation efficiency is judged from the of activation energy of nucleation and Avrami constants* (k, n; especially n).The combination of these two oxides have quite low Eact (255.5 kJ/ mole) but quite high value of n(4.38 ) because two these oxides form a single homogeneous ZrTiO4 spinel. Of couse, it is the TiO2 which first promotes a metastable liquid-liquid phase separation which starts internal crystallization. The other metal oxides do not form single homogeneous phase.

[II] Now about how it is caused in the bulk of glass:

Suppose, we add TeO2, ZrO2 into the ordinary silica glass containing extended O-Si-O network forming regular SiO4 tetrahedra possessing bridged oxygens[BO].Glass being a condensed liquid is porous,i.e. have vacancies.The ions like Ti (IV), Zr (IV) enter these vacant sites. The oxygen having strong affinity for these metal ions would be attracted toward them to bring about distortion in a regular tetrahedral SiO4 in two steps. First it changes this regular tetrahedron to an asymmetric one by elongating three Si-O bonds and shortning one Si-O bond. Then, one Si-O breaks to form SiO3 [trigonal pyramid having non bridging oxygens (NBO)].This increase in NBO causes an increase in refractive index because NBO are more polarizeable than bridging oxygens (BO) and thus this effect is transmitted in whole of the bulk of the glass.

[III] Lastly,*the Avrami equation [given by Avrami, M (1939); containing two constants k, n] describes how solids transform from one phase (state of matter) to another at constant temperature.

Dr. Sehgal, you have discussed how Ti or Zr promotes the formation of NBO..

But, the formation of NBO also caused by the addition of alkali oxide like Na2O, K2O or alkaline earth oxides..But these oxides do not act as nucleation agent . Can you please address this issue?

Secondly, You have mentioned "The oxygen having strong affinity for these metal ions would be attracted toward them" in comparison to other metal ions....Why it is so?

For the nucleation agent, it is not necessary that both TiO2 and ZrO2 have to present in the glass. both the oxides alone may act as nucleating agent. so, the idea of spinel formation will not be valid in this case.

Thanks for your reply. The possibility of liquid liquid phase separation in the presence of nucletaing agent is interesting.....Do have any idea why these oxides promote L_L phase separation?

Once again, thanks...

It gives me some satisfaction to know that your self have agreed with my answer of your second question.I ,now, try to answer your queries for this question in my own humble way:

[1]There are three factors which may make Na/K/ Mg unsuitable.

(a)We are adding Ti/Zr (+4) for a specific purpose, i.e. to change the properties- especially the OPTICAL ones by changing their refractive index(RE).What good will the addition of more of Na/K/Mg bring to the ordinary glass which have already these ions present in it?

(b) The presence of the NON-STOICHIOMETRIC oxides (TiO2/ZrO2) have an additionl advantage because of their greater tendency to occupy interstial sites (empty spaces) while Na/K/Mg ions will prefer substitution.So there should be lesser chances of formation of NBO in presence of Na/K/Mg ions.

(c) Moreover, NBO increases covalent character .Ti/Zr (+4)—O (-2) bonds should be much covalent (according to Fazan's rule, i.e. higher charge/ radius ratio) than that of Na/K (+)—O (-2) or Mg (+2)---O(-2). So Na/K/Mg ion addition would hardly change the optical properties.In a way, there should be less change inNBO in Na/K/Mg ion addition than that of Ti/Zr (+4) addition.

[II] It is reported in the literature that TiO2 is one among the hardest substances; implying very high stability of Ti-O bond or very great affinity of O for Ti (+4).

[III] I agree that both TiO2 and ZrO2 separately can be added to bring about changes in the glass. But your self wanted answer in the presence of both these oxides.

[IV] Lastly, the formation of spinel [ZrTiO4], its homogeneous nature and the favorable Avrami parameters are already reported in the literature.