Despite the advances in discovering new nuclei, modeling microscopic nuclear structure, nuclear reactors, and stellar nucleosynthesis, we lack a systemic tool, in the form of a network framework,
to understand the structure and dynamics of 70 thousands reactions discovered until now.
We assemble here a nuclear reaction network in which a node represents a nuclide, and a link represents a direct reaction between nuclides.
Interestingly, the degree distribution of nuclear network exhibits
a bimodal distribution that significantly deviates from
the power-law distribution of scale-free networks and
Poisson distribution of random networks.
The distribution is universal for reactions with a rate below the threshold,
λ-Tγ, where T is the temperature and γ≈1.05.
We discovered three rules that govern the structure pattern of nuclear reaction network: (i) reaction-type is determined by linking choices, (ii) spatial distances between the reacting nuclides are short,
and (iii) each node in- and out- degrees are close to each other.
By incorporating these three rules, our model
unveils the underlying nuclear reaction patterns hidden
in a large and dense nuclear reaction network.
It enables us
to predict missing links that
represent possible new nuclear reactions not yet discovered.