No, you'll have to etch the aluminum then the steel, Al is simply too reactive to etch at the same rate as steel. You MAY be able to wax coat the steel, barker's etch the Al, strip, wax coat the Al for a nitol etch of the steel...but i'm sure that if the layers are thin enough for SEM, it would be difficult to get a good coating. 

I studied Al-Cu welded structures under optical microscopes. We used colloidal diamond solution to polish the samples. After using this, there was no need to etch the samples. We polished by holding the samples near the edge of the disc polisher. This gave us better control on the extent of polishing needed on different areas. 

You can see the images in my poster which is attached.

Fig 3. is of the polished samples and 5 onward are of the obtained micro-structures.

If you still want to go for enchant, there is no common solution to do this. You may consider soaking a cotton and rubbing it softly over the two different area or as Antonio suggested, use wax(it may damage the polished surface)

Al- Krolls reagent (10-15 seconds)

Fe-Picral (20 seconds-1 min)

Please do some trials on sparring samples so that you don't waste the important ones in cases this goes wrong.

Both has nitric acid (Krolls reagent also contains HF) and are hazardous so take necessary precautions in handling them. I have also attached a handbook for metal etchants.

Hope this helps.

http://vector.umd.edu/links_files/Handbook%20of%20Metal%20Etchants.pdf

Conference Paper Impact of Silver coating on Al-Cu welding in flexible

Thank you Mr. Antonio and Ankit.

In fact I had tried different combinations. The better result has given Nital and Kellers reagent. But, still at the interface (IMC layer exist and irregular boundary) nearer to the interface the details are not visualized. 5xxxx series contain Mg, may be that will be the problem? Also, nearer to the interface steel particles distributed in the Al matrix.

May, I ask, what results are you looking for? The grain size or the interface composition or soundness of the joint? See, the joints will be irregular. It depends on the type of welding and skill. 

Please refer to ASM Handbook Vol. 9 and 6. They have some micro-graphs along with preparation method in detail. Additional, as I mentioned earlier, try with diamond solution and polishing paper using automatic polishing machine 

The specification are:

Poly-crystalline diamond solutions. 3 micron for initial and 1 micron for later stages. In final run, polish it with colloidal silicon carbide solution for 30 seconds to 1 min. You can use papers of grit 400 and 600 grit. Observe it in as polished condition. It may not give out grain boundaries in a good way but you can certainly distinguish the phases, see dendrictes formation and so on.

Steel particles are distributed in Al- matrix as the melt solidified some Fe got captured in it. I think most probably you will have a zone with such behavior near the interface and you can always characterize these to know the extend by EDX.

P.S. I don't think presence of Mg should be a problem.

Thank you Mr. Ankit.

Let me try...