Hi,

Performing BSE-analysis is possible for well-studied samples. I'd recommend to use it along with WDX or less preferably EDX for better understanding of a nature of BSE brightness difference.

Brightness of BSE image is relative characteristics showing a number of backscattered electrons leaving a sample in every point. Sometimes an absolute values could be obtained and compared with Monte-Carlo simulation. This number depends on sample composition in that point and crystal structure (last one usually gives a weak difference).

Also, these images are very sensitive to surface contamination and geometry so you need an ideal flat and clean surface perpendicular to the beam.

If you need a deeper view - ask for details.

Regards.

Could you tell more about the influence of the melting point on the BSE signal? Where did you read this?

You cannot use BSE for elemental mapping, you need EDS (or WDS) for that. 

 Choose a region of interest and EDS will produce a spectrum of various elements in terms of counts vs keV,

Thank you all for your valuable comments. I will use EDS for my elemental mapping

BSE would give you elemental contrast in a way that the average atomic number of the grain, phase or zone is analyzed.you would not know what element it is only if it is a high atomic number (Bright on the image) or low atomic number (darker on the image) material. as others have mentioned EDS or WDS is required for element mapping as well as identification. The crystalline contrast differences could also be visible in the image (however you would need very low mechanical damage on the surface of the samples by colloidal silica polishing or some other lapping compound). BSE would only give you an idea of where you might want to start the EDS mapping.