Nuclear reaction network unveils novel reaction patterns based on stellar energies

Despite the advances in discovering new nuclei, modeling microscopic nuclear structure, reactors, and stellar nucleosynthesis, we still lack a systemic tool, such as network approach, to understand structure dynamics of over 70 thousands reactions compiled JINA REACLIB. To this end, develop an analysis framework, under which it is simple know generally are possible not, by counting neutrons protons incoming outgoing from any target nucleus. Specifically, assemble here reaction node represents nuclide, link direct between nuclides. Interestingly, degree distribution exhibits bimodal that significantly deviates common power-law scale-free networks Poisson random networks. Based on cross section parameterizations REACLIB, surprisingly find universal for with rate below threshold, , where T temperature γ ≈ 1.05. Moreover, discover three rules govern pattern network: (i) reaction-type determined linking choices, (ii) distances reacting nuclides 2D grid Z vs N short, (iii) each in- out-degrees close other. By incorporating these rules, our model universally unveils underlying patterns hidden large dense regardless nuclide chart expansions. It enables us predict missing links represent not yet discovered.