A consideration of the instability strip in the H-R diagram suggests red giants don't last long in astronomical terms, but in actual years I don't really know...my totally-unsupported-by-real-data guess is at most several tens of millions of years, probably much less. To produce an iron world (simply as an exercise for the science fiction writer) from an earthlike predecessor would require extreme heat for the stony crust and mantle to boil off, and perhaps even part of the core, yet not so much or for so long that the entire planet is vaporized. Presumably, we can adjust the parameters of our thought experiment so that happens. The iron world, if possible at all, is poised between the extremes of total vaporization and temporary partial surface melting. They are probably very rare, but not necessarily impossible.

The core of this planet would already be hot, perhaps even partially molten, so when the insulating outer layers and the external heat source are removed the core would rapidly cool due to the thermal conductivity of the metal. If the core was highly differentiated when the planet was originally formed, the surface might be pure nickel iron, with little difference to the core in chemical composition. It would probably resemble the surface of a glacier, broken and uneven, due to radiative cooling at the surface, although at what characteristic scale I can only guess. However, there is always a possibility that a relatively thin regolith might form from non-metallic minerals precipitating out of solution and floating up to the surface from the liquid parts of the cooling core. The near-stellar environment might also conceivably supply enough non metals, perhaps derived from the star's solar winds (the star would now be well into a planetary nebula stage) to provide a regolith of sorts.

If these planets are possible, they might well be very strange places.