“…It turns out that within MFT, this issue can be fully resolved in terms of a modified DA for fictitious (+, o) BCs if the location and the curvature of the actual interface at x = 0 are known. Note, however, that this mapping onto a fictitious ersatz colloid with Dirichlet BC is not expected to hold beyond MFT, because the intrinsic order parameter profile vanishes at z 0 (x) linearly ∝ z − z 0 (x), whereas at a Dirichlet wall it vanishes ∝ (z − z 0 (x)) (β 1 −β )/ν , where β 1 is a surface critical exponent (β 1 (d = 4) = 1, β 1 (d = 3) ≃ 0.8 [11]). Thus, within MFT (β 1 − β )/ν happens to be equal to 1, too, whereas in d = 3 (β 1 − β )/ν ≃ 0.8.…”