You could do a couple of things; Try oxidizing the crucible prior to melting.  Use an oxygen rich environment (95% plus) at 1000 C.  The other thing you could do is add a slag forming agent prior to pouring.

Dear Rita Nedeljkovic,

Having no experience at all in melting Mg alloys, I can only make some remarks based on common sense:

1, Would increase the oxygen content of the environment of your melting pot imply an undesired oxidation of your Mg alloy under study?

2. Apparently the Fe content of your solidified Mg alloy stems from the mild steel crucible applied. So skip the mild steel crucible and replace it by a ceramic one.

I second , using graphite or ceramic crucible will help a lot. Furthermore,  you can cut off the bottom portion of the casting where heavy elements may accumulate. 

Dear Rita Nedeljkovic,

Presently, I'm working on the development of Mg-Li alloys. What I understood from the earlier researchers that unlike Al, Magnesium has less affinity towards Fe so the chances of Fe contamination is very less in Mg, so using of steel/inconel crucible is still under practice in many places.

In our case we are employing inconel crucible coated with the thin layer of Boron Nitride since it acts as a protective layer between the molten magnesium and the crucible. We have found no Fe contamination in any of the alloys we have fabricated. It was confirmed through various chemical analysis. Also, we purge high purity argon gas for the protective cover.   

The other better way is purchase a very high purity magnesium and use graphite crucible. This rules out any chances of Fe contamination. Also coat the mould with any of the high temperature lubricant if the mould is mild steel or any Fe based crucibles.