Peak splitting can have several reasons. First, the application of several radiation components, commonly K alpha1 and K alpha two which represent two characteristic energies of x-ray photons during transition within the electron shell. There are several others which can be reduced by absorber foils (for K beta) or if you use monochromators. Another option is that you have two or more phases after syntheses. The thrd, and this is the perhaps most probable, is a reduction of symmetry of your phase. Symmetry-equivalent planes, e.g. cubic cube-planes {001} split into {001} and {100} (tetragonal, hexagonal, trigonal) or even into {001}, {100} and {010} (orthorhombic, monoclinic, triclinc). These planes ave then different distances and cause different Bragg angles. This becomes even more challenging for more general planes like {123} which have only a single reflex for cubic but 24 different reflexes for triclinic crystals.

The geometrical concept is quite easy, and in order to collect a bit experience, you should have a look at programs like PowderCell which offers in a simply way nearly all fundamental parameters which you can change and study (including the peak splitting caused by the radiation doublet mentioned above). PowderCell also offers a simple full pattern refinement for simple phase combinations. The term rietveld is a) not properly used since it described originally a crystal structure refinement, i.e. it deals with single phases. Moreover, the guy who invented it was Hugo Rietveld, i.e. one should certainly write Rietveld refinement. But you already recognize that ignorance does neither prevent a wrong use nor a wrong spelling. But, YES, it is not a big deal for these kind of programs to solve your specific problem. If there is no correct solution, you will see that specific reflections cannot be described by Laue indices hkl of the assumed phase, i.e. either the phase is wrong, or you need another phase to describe the remaining peaks. Another hint: the symmetry reduction is usually not simple for non-crystallographers. Here also PowderCell helps a lot since it offers all maximal subgroups descriptions of the International Tables of 1995. These are not all, but first obviously only a small amount of possible are missing, and second, there is perhaps still no other software available which is doint this automatically. Short example: Enter you perhaps cubic phase. Then select the subgroup button and select one of the offered lower symmetry groups. Another structure descriptions will appear with the same signal, but now you can change e.g. the lattice parameters a and c if you selected a tetragonal (or rhombohedral) subgroup. And you will immediately see the peak splitting caused by your lattice parameter variation... Good luck!     

Conference Paper Introduction to PowderCell

Chapter PowderCell : a mixture between crystal structure visualizer,...

Peak splitting can have several reasons. First, the application of several radiation components, commonly K alpha1 and K alpha two which represent two characteristic energies of x-ray photons during transition within the electron shell. There are several others which can be reduced by absorber foils (for K beta) or if you use monochromators. Another option is that you have two or more phases after syntheses. The thrd, and this is the perhaps most probable, is a reduction of symmetry of your phase. Symmetry-equivalent planes, e.g. cubic cube-planes {001} split into {001} and {100} (tetragonal, hexagonal, trigonal) or even into {001}, {100} and {010} (orthorhombic, monoclinic, triclinc). These planes ave then different distances and cause different Bragg angles. This becomes even more challenging for more general planes like {123} which have only a single reflex for cubic but 24 different reflexes for triclinic crystals.

The geometrical concept is quite easy, and in order to collect a bit experience, you should have a look at programs like PowderCell which offers in a simply way nearly all fundamental parameters which you can change and study (including the peak splitting caused by the radiation doublet mentioned above). PowderCell also offers a simple full pattern refinement for simple phase combinations. The term rietveld is a) not properly used since it described originally a crystal structure refinement, i.e. it deals with single phases. Moreover, the guy who invented it was Hugo Rietveld, i.e. one should certainly write Rietveld refinement. But you already recognize that ignorance does neither prevent a wrong use nor a wrong spelling. But, YES, it is not a big deal for these kind of programs to solve your specific problem. If there is no correct solution, you will see that specific reflections cannot be described by Laue indices hkl of the assumed phase, i.e. either the phase is wrong, or you need another phase to describe the remaining peaks. Another hint: the symmetry reduction is usually not simple for non-crystallographers. Here also PowderCell helps a lot since it offers all maximal subgroups descriptions of the International Tables of 1995. These are not all, but first obviously only a small amount of possible are missing, and second, there is perhaps still no other software available which is doint this automatically. Short example: Enter you perhaps cubic phase. Then select the subgroup button and select one of the offered lower symmetry groups. Another structure descriptions will appear with the same signal, but now you can change e.g. the lattice parameters a and c if you selected a tetragonal (or rhombohedral) subgroup. And you will immediately see the peak splitting caused by your lattice parameter variation... Good luck!     

Conference Paper Introduction to PowderCell

Chapter PowderCell : a mixture between crystal structure visualizer,...

Please Check is there any phase separation during the synthesis. You may have chance to get separately Fe3O4 as well as Co3O4 or CoFe2O4 or FeCo2O4 something like that instead of getting the expected product Fe1.5Co1.5O4. You can use different probabilities and based on trial and error you will have to find the actual status of your compound. If your material shows more than one phase that could answer your question if the splitting is related to different phases..

Thanks Gertz and Bhange Sir for your answers. 

I'll do as per your answers and come back to you.

I have another question:  Can addition of any other material compound or impurity also split the peaks?

Impurities give you additional peaks which might look like splitting, but maybe you got phase with lower symmetry e.g. tetragonal with cell parameters close to cubic?

https://www.flickr.com/photos/85210325@N04/14348388500/in/album-72157645018820696/

I agree with Gert Nolze and Bhange D.S.

It would be useful to see your XRD pattern. Is your Fe1.5Co1.5O4 really cubic. Have I understand right? BCC or FCC ..??

@ Bartłomiej Gaweł: Could you shortly explain how impurities can give additional peaks which look like splitting in XRD?

I agree with Gert and think that such peaks are a consequence of phase separation or transformation

Sorry for the grammar mistake...on the top.

If your compound does not include separate phases ....if the additional peaks are not related to these phases (Fe, Сo and the related ones)....if crystal lattice of your compound is cubic...... keep in mind distortions of cubic crystal lattice which can be uniform and non-uniform. I cannot see your XRD pattern, but if all peaks are not split I expect non-uniform distortion. You can find some information in my papers in 2015 and 2016.

Peak splitting may also due to formation of new phase. for example cubic phase have only few sharp peaks, where as the tetragonal phase will have their peak splitted.

Sorry for replying so late.

It was synthesized by solution combustion method using metal nitrates and urea as fuel. Please find the attached XRD.

They look like K-alpha_2 to me.

Also, unrelated, but worth considering for attention - you have a rising background, probably due to Fe fluorescence causing interference.

 Shishonok Sir

I just compared the data with one available from NIMS website where it is matching with Fe1.5Co1.5O4 BCC fd-3m.

I don't have access to the JCPDS source. If anyone can provide me with the data for different phases of the material?

I think Ian may be right, because k-alpha_2 radiation has the half intensity of k-alpha_1 and at first sight it seems to be the case.