Coagulation-fragmentation equations are mathematical models describing systems where particles clump together (coagulation) and break apart (fragmentation) while also spreading through space (diffusion). A central concern in these models is whether the total mass of particles is conserved over time, or whether a runaway clumping process called "gelation" can occur, where finite mass accumulates into an infinitely large cluster in finite time. Understanding when gelation is absent is important for applications in aerosol physics, polymer chemistry, and biological clustering processes.
The key challenge addressed here involves a gap between two previously studied settings. When the system is spatially uniform (no diffusion), researchers already knew that mass is conserved as long as the rate at which particles of sizes i and j combine grows no faster than proportional to their combined size (i plus j). But when particles also diffuse through space, proving the same result under the same condition had remained an open problem in low dimensions. This paper closes that gap: in one and two spatial dimensions, the authors prove that mass is indeed conserved under exactly that same critical growth condition on the coagulation rate. In up to four dimensions, they also prove that solutions remain well-behaved and regular, meaning no singularities develop, provided the coagulation and fragmentation rates decay appropriately with cluster size.
The technical significance is that diffusion makes the problem harder because it creates local concentrations of particles that can accelerate clumping beyond what spatially uniform models suggest. The authors develop new analytical tools to control these concentrations and show they cannot trigger gelation. By matching the spatially uniform threshold exactly in low dimensions, the result is essentially sharp, meaning the condition used cannot be easily weakened. This work advances the mathematical foundation for a broad class of physical and biological models involving both spatial transport and particle interactions.