Hi!

                  I have experience with other perovskites like manganites and generally high temperature treatments (in steps like mixing and heating 2-3 times) are able to decompose and remove the carbonates from the samples. To obtain the correct perovskite phase, after each heat treatment it is preferable to have an XRD investigation and analysis.

I quite agree with the first comment. If you sinter the samples at high temperatures, the formation of carbonates during sintering is excluded, if you are afraid of some admixtures of carbonates on the surface during cooling, you can quench the samples in an inert atmosfere

Dear Sebastian, Avoiding the carbonate route may not yield purer phase in the shortest time. The carbonate phase may decomposes at temperatures above 700 C. Like  the first two comments have suggested, first monitor the phases with XRD from say 300 C upwards and you'll be able to identify the temperature at which the carbonate phase disintegrated. Previous researchers have shown the carbonate route as the most economical sol-gel route. You may read this paper for more info read-www.researchgate.net/publication/233741913_A_new_stoichiometry_of_cuprate_nanowires

Article A new stoichiometry of cuprate nanowires

Hi, carbonate contents depends on your preparation method (e.g., urea decomposition, alkoxides, peroxydes, etc), so I'd need more data to answer properly. However, it's not a bad idea to work with acac derivatives, but you'll also need to get rid of the organics. I'd suggest to perform FTIR and TG/DSC analysis in order to determine a) carbonate presence, b) which is the best thermal treatment to eliminate carbonate. I hope this helps! Feel free to ask and discuss

I would have suggest the same as M. Soler-Illia. FTIR & TG/DSC could be a good beginning (XRD is essential but maybe much longer ; TG/DSC gives somes hinches to where to look. Also take careful attention to the heating rate : There's a lot of work on the temperature profile based on this...

The method I used is so-called diol-based sol-gel route. And one more important aspect is that the compound is alkali tantalate (perovskite oxide).

As I mentioned already, two distinct alkali precursors were used: acetate and acetylacetonate (beta-diketonate). In both cases I have performed FTIR, DTA/TG and in situ XRD. Phase evolution is completely different in both cases: i) acetate route forms firstly intermediate pyrochlore, and ii) acetylacetonate route not (intermediate (bi)carbonate structures are detected and at higher temperature perovskite phase). IR spectra in both cases show CO32- bands in powders calcined even at 1000C.

Can you comment more knowing the described details? Is decomposition of alkali metal precursors so different that it changes so much the phase formation path?

Thanks!

I have also had two different phase evolution routes for the synthesis of Pr123 using two different biopolymers. In one the ceramic phase appeared below 800 C while the other above 800 C. An intermediate was completely absent in the first instance but was the prominent intermediate in the second. So don't be surprised at such result. I would suggest that you try a sol-gel route where the gel will serve as a sacrificial template. This would control the nucleation of the CO32- in nano pockets that could melt earlier than you're getting now. 

Hi, it is possible to avoid formation of bulk carbonate species in BaTiO3, but it is practically impossible to hapmer its formation on the surface. Recently, we have applied the citrate method to prepare a gel of Ba and Ti citrate. After drying, we have got the dried gel TG/DTA analyzed. Accordingly, we calcined the gel at 800 C for 3 h to get microcrystalline BaTiO3 free of bulk carbonate with 6 squared meter/g surface area. However, a post-IR analysis of the material revealed the presence of minority surface carbonate species.

How crucial is formation of surface carbonate/bicarbonate on the properties etc.?

Are alkali or alkaline-earth metals more prompt to carbonate formation?

If you are planning to measure surface properties by techniques like XPS, etc., then it is important to remove the carbonate/bicarbonate layer in order to get good signal from the material. Otherwise, the surface layer does not influence transport/magnetic/structural etc. properties.