We demonstrate the use of maximum entropy based deconvolution to recon
struct boron spatial distribution from the secondary ion mass spectrom
etry (SIMS) depth profiles on a system of variously spaced boron delta
layers grown in silicon. sample-independent response functions are ob
tained using a new method that reduces the danger of incorporating rea
l sample behavior in the response. Although the original profiles of d
ifferent primary ion energies appear quite differently, the reconstruc
ted distributions agree well with each other. The depth resolution in
the reconstructed data is increased significantly and segregation of b
oron at the near surface side of the delta layers is clearly shown.