It entirely depends on the annealing temperature. However, for any appreciable temperature (>500 C), impurities in the crystal will start diffusing depending on their diffusion constant. If you have defect centers in the wafer (e.g., damage from your FIB milling), the impurities will diffuse to those defect centers as they provide a low energy "sink". When this process is intentional, it's referred to as "gettering" (see these notes:http://dunham.ee.washington.edu/ee528/notes/getter.pdf), to trap unwanted metallic impurities away from the active regions of the wafer, typically at the surface. Phosphorous has a lower diffusivity than other impurities, so it will move slowly, but it will definitely reorganize. Given that you have a moderate doping level and it's uniform, this process shouldn't affect the uniformity of doping in the wafer (unless you reach temperatures much higher than 950C, where the dopants will start moving faster).

If you're using Ga ion milling, you may introduce Ga ions as unwanted impurities at your surface. Because your FIB damage is at the surface, you may trap Ga and other metallic impurities there by "gettering" during annealing. Some people introduce intentional damage at the backside of the wafers by scratching or lapping, to trap impurities away from the active surface at the top. You may want to look into it.

Hope this helps.

Hi Tanuj,

Thanks a lot for your reply and useful info. I am using Ne-FIB milling silicon. In principle Ne should have less chance of getting trapped and damaging silicon. I haven't found much information/data about Ne milling damage of silicon to compare. Could please let me know if you have any info regarding damage caused by Ne milling.

Thanks again,

Salahuddin