We study surface-mediated, diffusion-controlled reactive processes on particles whose overall geometry is homeomorphic to a sphere. Rather than assuming that a coreactant can diffuse freely over the surface of the particle to a target site (reaction center), we consider the case where the coreactant can migrate only among N -1 satellite sites that are networked to the reaction site by means of a number of pathways or reaction channels. Five distinct lattice topologies are considered and we study the reaction efficiency both for the case where the satellite sites are passive and for the case where reaction may occur with rinite probability at these sites. The results obtained for this class ofsurface problems are compared with those obtained by assuming that the reaction-diffusion process takes place on a planar, two-dimensional surface (lattice). The applicability of our results to surface-mediated processes on "organizates" (cells, vesicles, micelles) and on colloidally dispersed catalyst particles is brought out in the Introduction, and the correspondence between the latticebased, Markovian approach developed here and Fickian models of surface diffusion, particularly with regard to the exponentiality of the decay, is discussed in the concluding section.
Section 1. IntroductionTheoretical efforts to understand diffusion-controlled reactive events on the surface of cellular, vesicular, or micellar systems or on the surface of catalyst particles dispersed in a fluid phase have usually invoked reaction-diffusion models solved by assuming that one reactant can diffuse freely over the surface of the system. In a typical application, a rotational diffusion equation is written down and solved for a concentration variable C(0, t)-namely,at sinO aO ae (where 0 is an angle defined with respect to an underlying spherical polar coordinate system and D is an associated rotational diffusion coefficient)-subject to a variety ofinitial and boundary conditions, C(0, t = 0) and C(0 = 00, t), respectively. There exists an extensive literature on this general class of problems, ranging from the classic review of Noyes (1) in 1961 to the more recent monographs of Aris (2), Nicolis and Prigogine (3), and Haken (4). The specific study of Bloomfield and Prager (5) was the motivation, in part, of subsequent work on the role of dimensionality and spatial extent in influencing intramicellar kinetic processes (6).The particular problem we should like to address derives from the observation that in many diffusion-reaction problems in which the reactants are confined to the surface of a particle, the species are not free to diffuse freely over the surface of the system. In cellular systems, for example, whereas it is known that lateral diffusion on the surface of the encompassing membrane may be quite free, a diffusing species must nonetheless negotiate transmembrane (and other) proteins whose presence bifurcates the surface into channels and breaks the rotational symmetry of the problem [for a recent review here, see Sackmann (7)]....