Entropy-driven phase behaviour of all-DNA associative polymers

Associative polymers (APs) with reversible, specific interactions between ``sticker'' sites exhibit a phase behavior that depends on a delicate balance between distinct contributions controlling the binding. For highly-bonded systems, it is entropy that mostly determines if, on increasing concentration, the network forms progressively or \textit{via} a first-order transition. With the aim of introducing an experimentally-viable system tailored to test the subtle dependence of the phase behavior on the binding site topology, here we numerically investigate AP polymers made of DNA, where ``sticker'' sites made by short DNA sequences are interspersed in a flexible backbone of poly-T spacers. Due to their self-complementarity, each binding sequence can associate with another identical sticky sequence. We compare two architectures: one with a single sticker type, $(AA)_6$, and one with two distinct alternating types, $(AB)_6$. At low temperature, when most of the stickers are involved in a bond, the $(AA)_6$ system remains homogeneous, while the $(AB)_6$ system exhibits phase separation, driven primarily by entropic factors, mirroring predictions from simpler bead-spring models. Analysis of bond distributions and polymer conformations confirms that the predominantly entropic driving force behind this separation arises from the different topological constraints associated with intra- versus inter-molecular bonding. Our results establish DNA APs as a controllable, realistic platform for studying in the laboratory how the thermodynamics of associative polymer networks depends on the bonding site architecture in a clean and controlled way.