It really depends on the crystallographic distortion and the amount of doping that you intend to do.

I think the following papers might help you identify potential outcomes.

http://www.sciencedirect.com/science/article/pii/S0955221914005512

Hi Sandeep,.... in addition to previous answer, to construct a perovskite structure, you have to concern the tolerance factor, t , in the range of  0.75 - 1 and the total charge is always neutral. It is meant if you replace the smaller cation with the the bigger one ( or higher oxidation number), of course, by equal-neutrality charge principle, it will form holes (of cation position-neighbors). However, concerning on the properties, we have to consider other factors such as transition metal, orbital d, etc.

If you replace the cation with a smaller cation, at the same reason, to neutralize the charge, it means that double smaller cation will place the replaced cation ( A4+ was changed by  B2+), and of course, there is a distance movement to settle its cation 

(position-rearrangement). So both replacement will lead to defect certainly. 

Dear Kumar,

(a) For each 4 Ti4+ ions replaced by Nb5+ ions (q = 4e), one Ti4+ vacancy (q = -4e) is created. In this case, the formula would be ATi1-x-(x/4)NbxO3 = ATi1-(5x/4)NbxO3 ;

(b) For each 2 Ti4+ ions replaced by Nb5+ ions (q = 2e), one A2+ vacancy (q = -2e) is created. In this case, the formula would be A1-(x/2)Ti1-xNbxO3 ;

(c) For each Ti4+ ions replaced by Nb5+ (q = e), one titanium ion is reduced from 4+ to 3+ (q = -e). This can be thought as 1 electron localized at Ti. In this case, the formula would be ATi1-xNbxO3 = ATi4+1-x-yTi3+yNbxO3 ;

(d) For each 2 Ti4+ ions replaced by Nb5+ ions (q = 2e), one oxygen interstitial (q = -2e) is created. In this case, the formula would be ATi1-xNbxO3+(x/2) ;

(e) An alternative to mechanism (d) is that the charge-compensating extra oxygen is not allocated as an interstitial in the bulk, but is placed at the surface, forming new unit cells using cations from bulk as building material. Consequently, cation vacancies are formed in the bulk. In this case, the formula would be (ATi1-xNbx)1-[x/(6+x)]O3. Notice that this is the same formula that we get when we multiply the formula from (d) by 3/[3+(x/2)] to "normalize" it and get the correct amount of oxygen per unit formula in this new picture, (e).

(f) In all previous cases, we assumed that the doping was successful. This may not be the case, and the niobium may not substitute Ti at all. If this is the case, the synthesis reaction is going to be AO + (1-x)TiO2 + (x/2)Nb2O5 --> ATi1-xO3-2x + (x/2)Nb2O5. This is, the Nb will be present as a secondary phase instead of as dopant. Consequently, Ti4+ and O2- vacancies in the ratio 1:2 will be present in the bulk.

Each case has its own "signature" and consequences to transport properties. In cases (a) and (b), secondary phases of TiO2 and AO, respectively, are expected, and I can't think of any obvious effects on either electronic or ionic conduction. In (c), no secondary phase is expected, but electronic conductivity should increase, due to presence of extra electrons localized in the titanium ions. The case (d) is not very likely, due to the close-packed arrangement in perovskite structure. But if it does occur, it is expected an increase in ionic conduction due to O2- ions, as long as the migration energy of oxygen interstitials is low enough in order to this transport mechanism to be favorable. Also, no secondary phase is expected. In case (e), no secondary phase is expected and I can't think of any obvious consequences to either ionic of electronic transport. In (f), a secondary phase of Nb2O5 is expected, as well as an increase in ionic conductivity due to oxygen vacancies.

For a low concentration of dopant, it may be hard to detect secondary phases, but for high concentrations (> 5%, I guess), it may be possible to detect them with X-ray diffraction, neutron diffraction, or maybe Raman scattering. A possible change in titanium's valence may be detectable by XANES, XEOL, XPS, etc. Increases in conduction relative to undoped ATiO3 may be detected with Electrochemical Impedance Spectroscopy. The behavior of conductivity with temperature can give a hint on the mechanism: electronic conduction with a fixed number of charge carriers (which is probably the case in highly doped samples if electrons are created) tends to decrease if temperature is raised, while ionic conductivity tends to increase exponentially when temperature is raised (assuming oxygen vacancies or interstitials as the charge carriers).

I hope the above considerations help you somehow. Best regards!

Dear Kumar,

(a) For each Ti4+ replaced by Zn2+, two holes are created (two oxygen ions change their oxidation state from 2- to 1-). In this case, the formula would be ATi1-xZnxO2-3-2xO-2x ;

(b) For each Ti4+ replaced by Zn2+, one O2- vacancy is created. In this case, the formula would be ATi1-xZnxO3-x ;

(c) An alternative to mechanism (b) is to assume that there's no oxygen deficiency. Instead, there is cations excess, and the excess cations are in interstitial positions.  I'd consider this mechanism very unlikely, due to close-packaging in perovskites. The formula would be (ATi1-xZnx)1+xO3;

(d) There is also the possibility that the doping was unsuccessful and zinc did not substitute titanium at all. If this is the case, the synthesis reaction is going to be AO + (1-x)TiO2 + xZnO --> ATi1-xO3-2x + xZnO. This is, the Zn will be present as a secondary phase instead of as dopant. Consequently, Ti4+ and O2- vacancies in the ratio 1:2 will be present in the bulk.

Again each case has its own "signature" and consequences to transport properties. In cases (a), (b) and (c), no secondary phases are expected, while in case (d) a ZnO phase is expected. In (a) and (d), ionic conductivity should increase, due to oxygen vacancies. In (b), electronic conductivity should increase, due to presence of holes. In (c), I can't think of any obvious consequences to either ionic of electronic transport.

As you can see, for each dopant, there are several possibilities of charge-compensating mechanisms. There are atomistic simulation techniques that can elucidate which is the most favorable one by comparing their formation energy. See, for example, the attached paper.

You can also try to force the choice of a specific charge-compensating mechanism (with a specific intent, such as increasing ionic conductivity, for example) through synthesis. If you want more information about it, please let me know.

Best regards!

Article Mechanism of luminescent enhancement in Ba2GdNbO6:Eu3+ perov...

"What will be the effect on dielectric properties if we replace small valence cation is replaced by large valence cation?"

Is not easy to answer. Just pointing out of that to get the single phase itself not only the tolerance factor (RA+R0)/sqrt(2)*(RB+R0), the octahedral factor RB/R0 (0.425) also play the important role to formalism the Perovskite structure.

If it single crystal than u cud calculate the displacement due to Jahn Taller effect. Conduction cud be by the exchange mechanism or charge defect.

But in real is not that simple, many thing has to consider. It is way better to synthesize the material and check; single phase? substitution successful? equally distribute or localized? balance of charges or a constrain? etc etc.