Although the interactions between proteins and various types of macromolecules are of considerable technological interest, 1 little theoretical work has been done that makes use of the fact that proteins are often comparatively small particles. Thus, our interest is in a regime opposite to that focused on in the usual depletion theories. 2 Here, a protein will be viewed, perhaps naively, as a small hard sphere whose dielectric permittivity is negligible compared to that of water. Its interaction otherwise with some macromolecular segment will be inert. Some time ago, de Gennes already presented several, not so widely known, preliminary calculations of a spherical particle interacting with a semidilute solution of polymers. 3 One supposition he made concerning the irrelevance of a certain scale, will be proved here. My aim is to present a scaling analysis of the interaction between a small sphere and a macromolecule, particularly in dilute solution. Though obviously of restricted validity, the expressions derived may prove helpful in qualitatively understanding phase separation phenomena occurring in nondilute suspensions.We first consider a protein sphere of radius a immersed in an aqueous solution containing a semidilute polymer which is well soluble. Its Kuhn length is A K , and the excluded volume between two segments is ) A K 3 . In a self-consistent field approximation, the polymer segment density ψ 2 (r b) at position r b is given by 2 where the origin is at the center of the sphere, is an eigenvalue, and ψ must tend to zero at the protein surface (r ) a). Without solving eq 1, I wish to investigate the nature of the depletion layer surrounding the protein. Far from the sphere, the concentration ψ 2 asymptotes toward a constant c o , the bulk concentration of Kuhn segments, so we conveniently introduce ψ ≡ c o 1/2 f and eq 1 becomes Here, the correlation length ≡ A K (3 c o ) -1/2 is supposed to be larger than the radius a, a condition easily realizable in practice. Assuming spherical symmetry and setting r ≡ aR, one is faced with finding the solution f ) f(R,a/ ) to with boundary conditions f ) 0 at R ) 1 and f ) 1 at R ) ∞. For an infinitesimally small sphere, we have simply which implies the depletion layer around the sphere is approximately of size a. Moreover, the second and third terms in eq 3 are straightforward regular perturbations for a , . Hence, eq 4 remains valid within a zero-order approximation, even when a > 0 provided a , . We conclude that the scale of the depletion layer for a small sphere is given solely by its radius a and does not involve at all, at least to the leading order. The irrelevancy of (assumed earlier 3 ) is nontrivial for it is circumstantial. Note that an improved theorysa hybrid approach combining scaling and self-consistent arguments as in the theory of polymer adsorption 4 swould not alter this conclusion. One would require a correlation length ∼ c o -3/4 from scaling theory instead of ∼ c o -1/2 and hence replace f 3 in eq 3 by f 4 .We now rederive a previous result 3...