Hi Sunny. An idea could be to mount the powder particles in resin and to grind/polish the sample enough to generate the cross section of one or more particles.
I noticed this question since Davide had give his aswer.
I've made cross sections from gas atomised powders that are very dense, simply with hot mounting method and with conductive resin. I've used sieved hot mounting resin particles with small grain size and I've first pre-mixed small amount (