I am working on Zr based metallic glass which reacted with the Alumina crucible in which I was performing the DSC experiment. Kindly suggest possible reasons for this anamoly.
is probably right. Zr is a great oxygen scavenger, so you may have some ZrO2 formed on the surface of your glass (especially if it's in powder form).
Depending on the composition of your glass, Pt may not be inert (e.g. if it contains Cu, Ni,...). A possible solution would then be to use a zirconia crucible. If you can't get one, you can try to use a bed of coarse ZrO2 grains at the bottom of your alumina crucible. This may prevent the glassy sample from contacting the alumina during the (presumably short) part of the run spent substantially above the glass transition. You'll need to run "blank" runs with the ZrO2 grains alone to see if the DSC can "see" a reaction between them and the crucible. (My guess: probably not, if the liquid phase doesn't touch the crucible itself.)